显示语言 中英对照 中文 English
点击段落跳转到对应时间
0:00 有的几十年里什么事都不发生，有的几周里就发生了几十年的变化。过去九个月里，我们就见证了几十年的进步。 There are decades where nothing happens and weeks where decades happen. And we have seen decades of progress happen in the last nine months. 0:12 如果你意识不到当下这一刻的分量，那才是妄想，那才是精神错乱。 If you're not recognizing the gravity of the moment, that's the delusion. That's the psychosis. 0:19 如果一个精灵突然从瓶子里蹦出来说：“你想要什么都行。”谁不会欣喜若狂呢？ Who would not get delirious if suddenly a genie pops out of the bottle and says, "You can have whatever you want. 0:26 你在操作系统里梦寐以求的每一个功能，我都能交付给你，大部分五分钟搞定，少数二十分钟，真要敞开了玩，也就花我两个小时”？ Every feature you've ever dreamed of in an operating system, I can deliver them to you, most of them in five minutes, a few in 20, and if we really go hog wild, it's gonna take me two hours"? 0:38 我想要刹车性能直逼极限的快车。我想要能下潜到马里亚纳海沟的潜水表。我想要60秒内就能装好的操作系统。 I want the fastest car braking to speed limits. I want the diver watch that can go down the Mariana Trench. I want the operating system that can install in less than 60 seconds. 0:48 赛车一直让我着迷的一点是：当我从车里爬出来、整个人累垮了、头都快抬不起来的时候，我会躺在车库的地板上想：“我去，我还活着。” One of the things I always loved about race cars was when I would stumble out of the car absolutely smashed and barely able to hold my head up, and I'd lay down on the garage floor and just think, "Holy fuck, I'm alive." 1:01 奥弗顿之窗不会自己打开，它是被敢于冒一点险的人，一次推动一点，慢慢推开的。 The overtone window does not open itself. It opens one nudge at a time by people risking a little. 1:09 对。 Yeah. 1:09 冒一点名声的险，承受一点反对、一点批评，或者有时候是很多名声、很多批评、很多反对。 A little reputation, a little pushback, a little criticism, or maybe sometimes a lot of reputation or a lot of criticism or a lot of pushback. 1:16 和你爱的人一起孕育新生命，真的就是活在这个星球上最极致的体验。 Creating life with another human that you love is literally the peak experience of being on the planet. 1:28 接下来是与 David Heinemeier Hansson 的对话，也就是大家熟知的 DHH——Ruby on Rails 的创始人，37signals 的 CTO，全新 Amachi Linux 操作系统的创造者， The following is a conversation with David Heinemeier Hansson, also known as DHH, creator of Ruby on Rails, CTO of 37signals, creator of the new Amachi Linux operating system, 1:39 畅销书作家、赛车手，也是世界上最敢说、最高产的程序员之一。 bestselling author, race car driver, and one of the most outspoken and prolific programmers in the world. 1:49 过去二十多年里，编程对他而言就是精心手工雕琢优雅的 Ruby 代码。 For the past 20-plus years, for him, programming meant meticulously handcrafting beautiful Ruby code. 1:56 但最近，自 2025 年底以来，DHH 拥抱了 AI 革命，并公开地把自己再一次转型为最敢说、最高产的智能体工程师之一， But recently, since late 2025, DHH has embraced the AI revolution and has openly transformed himself into, once again, one of the most outspoken and prolific agentic engineers, 2:08 尽管他很讨厌这个称呼。 even though he hates that term. 2:13 那就叫他们“某种新形态编程”的践行者吧——在这种编程里，AI 承担大部分实际编码工作，人类则靠高层设计、愿景和品味来掌舵。 So let's say practitioners of whatever programming is becoming where AI is doing most of the actual programming and the human steers the ship with high-level design, vision, and taste. 2:26 是的，没错。DHH 有时确实颇具争议，但他永远无所畏惧、才华横溢，而且聊起来很有意思。 Yes, it's true. DHH is at times controversial, but he's always fearless, brilliant, and fun to talk to. 2:34 这又是一场激烈、大开眼界、如过山车般狂野的对话。这里是 Lex Fridman 播客。 This was, once again, an intense, eye-opening, and a wild rollercoaster ride of a conversation. This is a Lex Fridman podcast. 2:43 如果想支持本节目，请查看简介里的赞助商信息，那里也有联系我、提问、反馈的链接。现在，亲爱的朋友们，有请 DHH。 To support it, please check out our sponsors in the description where you can also find links to contact me, ask questions, give feedback, and so on. And now, dear friends, here's DHH. 2:56 13 个月前，我们就坐在这里聊编程，然后，呃，一切都变了。 13 months ago, we sat down right here to talk about programming, and then, uh, everything changed. 3:03 那时候，你对 AI 在编程过程中的角色还有些怀疑，之后我们就经历了智能体工程的这场飞速演变。 At that time, you were a bit skeptical about the role of AI in the process of programming, and then we went through this rapid evolution of, uh, a- agentic engineering. 3:14 那么，先问个简单的问题：你对 AI 在编程中角色的看法发生了什么变化？你兴奋吗？害怕吗？还是在坐情绪过山车？ So, uh, simple question to start: How has your view on AI's role in programming changed? Are you excited? Are you terrified? Are you on an emotional rollercoaster ride? 3:26 我兴奋得不得了。 I am incredibly excited. 3:29 你一直在聊“乐趣”这个话题。 You've been talking about fun a lot. 3:31 没有什么生存威胁。对我来说，它在情感层面根本不存在。 There is none of the existential threat. That doesn't exist for me as an emotional component. 3:39 它只存在于理性层面，而在情感层面，对我来说是百分之百纯粹的、不掺任何杂质的喜悦—— It is only there as an intellectual component, and the emotional component for me is 100% pure, unadulterated joy- 3:47 嗯 Hmm 3:47 ……还有乐观和惊叹：我们竟然让计算机做到了这种事。 ... and optimism and amazement that we've made computers do this. 3:52 我觉得特别有意思，我们 13 个月前才聊过，可感觉就像在不同的宇宙、不同的时代里对话。 And I find it so interesting that we talked just 13 months ago because it's like we talked in different universes, different eras. 4:02 嗯。 Hmm. 4:03 我很喜欢一句话，我想是列宁说的：“有的几十年里什么事都不发生，有的几周里就发生了几十年的变化。” I love this quote. I think it's Lenin. "There are decades where nothing happens and weeks where decades happen." 4:11 过去九个月，我们就见证了几十年的进步。你想想，假如你就在莱特兄弟首飞的现场。 And we have seen decades of progress happen in the last nine months. I mean, imagine you're there when the Wright brothers take flight. 4:21 对。 Yeah. 4:21 前一天《纽约时报》还在写“人类要飞起来还得一万年”，第二天我们就上了天，没过几年，跨大西洋的飞机就有了。 Yesterday, The New York Times would write, "It's gonna be 10,000 years before we fly," and then the day after, we're up in the skies, and just a few years after that, there were cross-Atlantic planes. 4:33 整个世界彻底变了。能亲历那样的时刻，是多大的福气。 The whole world has completely changed. What a blessing to be there in that moment. 4:38 如果你拉开距离看整个人类历史，有多少人有幸一辈子就生活在世界剧变的时代里？ If you zoom out and look at all of human history, how many humans got to live within the same epoch that they were born in? 4:47 他们从没见过世界和社会的彻底翻篇。而我竟然有幸赶上了两次，真是莫大的幸运。 They never saw that complete change of the world and of society. And to have been blessed with two of those feels just such a privilege. 5:00 我亲历了互联网—— I got to see the internet- 5:01 嗯哼 Mm-hmm 5:01 ……从互联网时代之前到之后，现在又是 AI 之前、AI 之后。这一趟太精彩了。何其幸运。 ... from pre-internet to post-internet, and then now pre-AI, post-AI. What an amazing run. How fortunate. 5:12 是的，但关键是，这次的变化感觉比人类历史上任何一次都快。 Yeah, but the thing is, this feels like a thing that happened faster than anything else in human history. 5:18 没错。 Correct. 5:18 所以我会说，大概在十二月，也许是十一月下旬—— So I, I would say somewhere around December, maybe late November- 5:22 11 月 24 日。 November 24. 5:24 这就像……你知道，当人们谈论—— This is... You know, when they talk about- 5:27 就是那个确切的时刻 That's the exact moment 5:27 ……世界大战中一国入侵另一国之类的时刻——我们就是这样谈论 AI 改变一切的。对，它……它真的是从许多优秀开发者的……转变过来的。 ... when one nation invades another in a world war or something like this, it-- this is how we talk about AI changing everything. Yeah, the... And it really just w- shifted from many great developers. 5:40 它从 AI 做点基础的自动补全、大概写 5%、10%、15%、20% 的代码，变成了写 80% 的代码。 It shifted to where AI is doing some, uh, basic auto-complete, maybe writing 5, 10, 15, 20% of code to writing 80% of code. 5:49 或者 100%。 Or 100. 5:50 或者 100%，对，尤其是那种不直接面向用户的产品。 Or 100, yeah, especially if it's not a, a public-facing product. 5:54 对我来说，这就是为什么即使回看一年前的对话，我其实并没有不同的观点。我的观点没变。 To me, that's why even when I look back upon our conversation a year ago, I don't actually have different opinions. I have the same opinions. 6:05 一年前，我不喜欢当时提供给我们的 AI 形态。它是自动补全模式，或者是 AI 聊天机器人模式。聊天机器人我其实挺喜欢的，我们当时聊过。 A year ago, I did not like the mode of AI we were offered. It was the auto-complete mode, or it was the AI chatbot mode. Now, the chatbot I actually liked, as we talked about. 6:16 它从一开始就是个很棒的老师，也是在互联网上查东西的好工具。但它不会取代我亲手雕琢代码。可是……后来智能体出现了。 Great tutor right from the get-go, great way of looking things up on the internet. Not what was gonna replace me chiseling code. But- ... then we get the agents. 6:28 对。 Yeah. 6:28 智能体一开始大概只新鲜了五分钟，然后就变得很厉害，再后来就是：“我的天，这就是 AGI 吗？” And the agents start out being curiosities for about five minutes, and then they get amazing, and then they get, "Oh my God, is this AGI?" 6:40 对。 Yeah. 6:41 而这一切就发生在去年，甚至只是过去九个月之内。基本上有这么几个阶段。 And all of that happened since last year, even just in, within the last nine months. We have basically these few phases here. 6:50 先是前智能体时代的 AI。那时我也很兴奋，但它并没有从根本上为我重写游戏规则，没有彻底改变我的工作方式。 We have AI in the pre-agentic era. I was excited about that, but it was not fundamentally rewriting the rules of the game for me. It was not completely changing how I worked. 7:03 我还在亲手雕琢代码，只是多了个小帮手、小跟班—— I was still chiseling code, I just had a little helper, a little sidekick- 7:08 嗯哼 Mm-hmm 7:08 ……可以跟它碰撞想法，也能更高效地在网上查资料，诸如此类。 ... who could bounce ideas off, and I could look up this, uh, information online and so forth in a more efficient way. 7:15 它只是让我本来就在做的事效率更高了，并没有改变我和计算机之间的那种情感联结。然后到了 2025 年 11 月 24 日。 It was just a more efficient way to do what I was already doing, and it didn't change the emotional connection I had to, to the computer. Then we get to November 24th, 2025. 7:27 对我来说，Opus 4.5 就是那条分界线，突然之间……我 24 号当天甚至没试，大概是 26 号才试的。 Opus 4.5, to me, was the dividing line, where suddenly... I didn't even try it on the 24th. I think I tried it on the 26th. 7:38 我给了它几个任务，然后发现它输出的质量，跟我自己会写出来的东西接近得吓人。 I give it a couple of tasks, and I realize that the quality of the output is uncannily close to what I would've written. 7:49 我记得当时我往后一靠，心想：“刚才发生了什么？” And I, I remember just leaning back and thinking, "What just happened?" 7:55 嗯哼。 Mm-hmm. 7:55 “怎么、怎么我们夏天还在聊那个自动补全的烂摊子，短短几个月后就变成这样了？ "How, how did we go from this auto-complete mess that I was talking to you about in the summer to this just a few short months later? 8:05 智能程度怎么提升了，可用性怎么也一起提升了？”这就是智能体框架的问题——我不确定 Opus 4.5 是不是比 Opus 4 聪明多少， How did we get both the increase in intelligence and then also the increase in usability?" This agent harness question, where I don't know if Opus 4.5 was that much smarter than Opus 4, 8:16 Opus 4 是我们夏天用的版本，但它操控你的电脑、使用工具、检查自己工作成果的能力，它把智能用得让你能真正干出有意义的活， which is what we had in the summer, but its ability to instrument your computer, to use tools, to check its own work, to apply its intelligence in such a way that you could get real meaningful work out of it, 8:29 这是完全不一样的。 was completely different. 8:36 我认为这就是几乎所有在圣诞假期留心并开始上手玩的人都经历的大转变。我是这么听说的—— And I think this is then the big change that happens for almost anyone who paid attention and started playing with it over the Christmas break. This is what I heard- 8:44 对 Yeah 8:44 ……从 Shopify 和其他员工很多的地方听说的，他们突然能喘口气，突然有了几周时间可以靠下来看看正在发生什么， ... from Shopify and other places with lots of employees, who suddenly had a breather, suddenly had a couple of weeks to lean back and just look at what was going on, 8:54 试了试，然后同样被震撼了：这些智能体跟之前的东西完全不是一个物种。 gave it a try, and had the same mind-blowing experience that these agents were of a different genre than what we had before. 9:05 到这时候我已经很兴奋了。到了十二月，我已经在重新审视自己所有的先入之见—— And then what we get to is I'm already excited at this point. Like, by December, I'm already revisiting sort of all my priors- 9:15 嗯哼 Mm-hmm 9:15 ……想着：“哇，既然它连这个都能做，那个是不是也行？” ... and going, like, "Wow, if it can do this, can it also do that?" 9:18 嗯哼。 Mm-hmm. 9:18 “哦，对，它行。”而现在，Opus 4.5 看起来已经像个智障模型了。 "Oh, yeah, it can." And again, Opus 4.5 now looks like a retarded model. 9:26 对。 Yeah. 9:26 这就是这种进步的魔力——你以为自己到站了。我记得当时还想：“如果这就是我们拿到的最后一个模型，我也满足了。” And this is the magic of this progress, is you think you've reached something. I remember thinking at the moment, "If this is the last model we get, I'll be set." 9:38 对。 Yeah. 9:38 “我就很开心了。” "I'll be happy." 9:39 嗯。 Yep. 9:39 “未来二十年我就用 Opus 4.5—— "I could live with Opus 4.5 for the next 20 years- 9:42 嗯 Yep 9:42 ……你也听不到我一句抱怨，”因为看到一个智能体用我想要的方式把所有这些活都干了，实在太不可思议了。 ... and you would hear no complaints from me," because it was just so incredibly amazing to see an agent do all this work in the way that I wanted it done. 9:53 因为它不只是能完成任务，我还能看着它一路的轨迹说：“对。对。呃，这里可能不太行，但差不多了，”然后你给它两条意见。 Because it was not just about it being able to solve a task, it was also that I could look at its path- out there and go, "Yep. Yep. Uh, maybe not there, but almost," and it would give you two notes. 10:05 它就能按我想要的方式到达我想去的地方。它能写出我愿意合并的代码。如果是 Ruby，它写出来的代码还真的漂亮。Rust 就另当别论了。 You get to where I wanted to go in the way I wanted to go there. You could produce code I wanted to merge. You could produce code that actually looked beautiful if it was written in Ruby. Rust, different question. 10:19 但智能体这种真正成为我工作方式延伸的能力，是非常新鲜的。 But this ability for the agents to truly become an extension of how I wanted to work was very novel. 10:30 可接着刚到早春，我们就有了子智能体。 But then we wait just until early spring, and suddenly we get sub-agents. 10:37 我们有了能拆分任务的框架，本来 Opus 要干很久的活，突然只花五分之一、十分之一的时间，因为任务能被切碎， We get harnesses that can subdivide the task, and something that would take Opus quite a while suddenly took a fifth of the time, a tenth of the time because it could get chopped up, 10:47 一下子就有八个子智能体替你干活。 and suddenly you got eight sub-agents working for you. 10:54 但智能体时代的前两个阶段，对我来说，依然感觉要我自己开车。 But both of those two first phases of the agentic age, to me, still felt like I had to drive. 11:02 我可以告诉它我要什么、去哪儿找，它跑偏了我再稍微把把方向，然后就能到终点，而且快得多。 I could tell it what I wanted, where to look for it, and steer it a little bit when it went off, and then we'll get there, and I would go much faster. 11:10 嗯哼。 Mm-hmm. 11:11 但我必须坐在驾驶座上，必须告诉它我要什么，必须亲自审查、把关产出的东西。 But I had to be in the driver's seat. I had to tell it what I wanted, and I had to be the reviewer, the auditor of what was coming out. 11:21 然后到了现在，今年夏天，有了 Opus 5、Fable 和 Sol、GPT Sol，还有一些开放权重模型（程度稍逊）， And then finally now, this summer, with Opus 5, Fable, and Sol, GPT Sol, and to a lesser extent, some of the open weight models, 11:30 我们进入了一个新纪元：我不再告诉它我们要去哪儿，我只告诉它我遇到的问题。 we've arrived at a new era where I'm not telling it where we're going. I'm telling it the problem I have. 11:40 我告诉它我脑子里那个模糊、朦胧的想法，它来告诉我要去哪儿。 I'm telling it the fuzzy, vague idea I have. It tells me where we're going. 11:47 嗯哼。 Mm-hmm. 11:47 它告诉我该走哪条路。我还是会看一眼，因为我是个好奇的人，我喜欢计算机，也喜欢看结果，但我其实已经不太需要看了。 It tells me which path to take, and I will still look at it because I'm a curious person, and I like computers, and I like the outcome of it, but I really kinda don't have to. 11:58 在产出代码、挑选路线的环节里，我已经变成了可选项。还记得早期 GPS 刚出来的时候吗？ I've become optional in the part that produces the code that picks the route. Remember when early GPS systems came out? 12:07 嗯哼。 Mm-hmm. 12:09 跟看地图比，它们已经很神了，但你还是得盯着点：它会不会把你开进港口？我记得那些报纸文章。“啊，GPS 太坑了，因为人们不留神，直接开进港口里。” They were amazing compared to looking at a map, but you still wanted to pay attention. Is it gonna drive you in the harbor? I remember these newspaper articles. "Ah, GPS, they're terrible because people don't pay attention. They drive in the harbor." 12:21 上一次 GPS 把人开进港口是什么时候？这种事不再发生了。事实上，现在车都能自己开了，对吧？我们现在就到了这个地步：我可以信任它。 When was the last time GPS drove anyone in the harbor? Like, that just doesn't happen anymore. In fact, the cars now just drive themselves, right? And this is where we've arrived at now, that I can trust. 12:33 在我目前从事的领域里，我可以信任它，完全放心它会罩着我。它不会干蠢事，就算干了蠢事，也能自己挽回。 For the domain I'm working in right now, I can trust it and feel completely confident that it's gonna have my back. It's not gonna do something stupid, and if it does something stupid, it's gonna be able to recover. 12:45 嗯哼。我们应该提一句，程序员所在的领域五花八门—— Mm-hmm. We should mention that there's all kinds of domains that- 12:48 没错 Correct 12:48 ……我其实不确定，但我觉得开发中最常见的领域大概就是 Web 开发的 CRUD。你有一个数据库。 ... programmers operate in. There's... I actually don't know, but I think the most common domain in development is, like, web dev crud. You have a database. 13:00 对。 Yes. 13:00 有一个面向用户的界面，数据来来回回，可能只给一个人内部用，也可能给几个人、一小群人用，或者面向全世界。 You have a user-facing interface, and it does something back and forth, and it could be internal to just one person, to multiple people, to a small number of people, or to, to the world. 13:11 我觉得在这种领域，它…… And I think for that, it's... 13:14 我是说，你真的可以做到 100% 的代码都由 AI 来写，而且几乎——如、如果你是一个好程序员，对背后发生的事有很好的直觉，你完全可以不看代码， I mean, you could really get to the 100% of code written by AI and really almo- I- if you're a good programmer and you have a good intuition about what's happening behind the scenes, you can legitimately not look at the code, 13:25 至少——这、这一直是我的体验。 at least th- that's, that's been my experience. 13:30 我不知道需要多少编程方面的既有经验，才能直觉地知道系统在幕后运转正常，就靠观察系统的、 I don't know how much, uh, prior experience with programming you need to have to kinda know that things are working correctly behind the scenes, like intuitively by observing the, 13:40 呃，涟漪效应、症状来判断。 uh, the, the ripple effects, the symptoms of the system. 13:45 但我认为在那个领域，呃，已经接近 100% 了。然后还有你所在的领域，比如写一个 Linux 发行版。 But I think for that domain, uh, it's, uh, it's close to 100%, and then there's domains like you're also operating in, which is writing Linux distribution. 13:55 那种领域可能要多看一眼，因为你对速度之类的东西很痴迷。那种地方，你可能得多看看代码。 There, maybe there's more because you're obsessed with speed and all that kinda stuff. There, maybe you need to look at the code a little bit more. 14:03 再往下还有安全攸关的系统，一直到操作核电站或者自动驾驶汽车，可能得更仔细地检查代码。 Then there's maybe safety critical systems all the way down to operating a, a nuclear power plant or a self-driving car. Maybe you need to look at the code more carefully. 14:12 对，但话说回来，AI 在发现——这方面强得离谱—— Yes, but that said, AI is insanely capable at both finding- 14:17 对 Yeah 14:17 ……以及修复安全漏洞方面都强得离谱。这就是 Fable 引发的那场大风波。这个模型找漏洞的能力太强了，攻击者完全可以利用，所以根本不敢发布它。 ... and fixing security vulnerabilities. This was the whole blow-up about Fable. This model was so capable of finding holes that a attacker could exploit that it was simply not safe to release. 14:30 讽刺的是，在这个领域，我们似乎已经达到了几乎没人能企及的智能水平，因为很多这类安全漏洞都涉及把一连串组合拳串起来。 So the irony here is that when you look at that field, it seems like we've reached levels of intelligence that virtually no human can match because many of these security holes are about stringing combo moves together. 14:47 你在这儿找到一个小漏洞，单看它可能没什么大不了，但把它和另外四个组合起来，突然就有了 RCE——远程命令执行。 You find one little vulnerability here that by itself might not be the worst thing in the world, but then you combine it with four others, and suddenly you have RCE, Remote Command Execution. 15:00 能做到这一点的人类非常稀有，他们通常在国家级组织或者其他秘密机构里工作。 Humans who are able to do that are very rare. They usually work inside state-sponsored organizations or other clandestine operations. 15:11 他们不是随便在外面晃悠找漏洞的人。AI 在这方面已经强到这种程度了。 They're not just out and about finding things. So they've gotten so good at that. 15:16 正是 Linux 发行版让我彻底“吃了 AI 的药”。过去三个月我一直在做 Umachi，就是前几天刚发布的那个叫 Quatro 的版本。 The Linux distribution is why I've gotten 100% AI pilled because I've been working on Umachi for the last three months, this version that just dropped a few days ago called Quatro. 15:31 几乎从一开始做这个版本起，智能体加速就接近 100%，而过去两个月就是 100%。我没有亲手写—— And almost right from the beginning of working on that version, the agent acceleration neared 100%, and in the last two months it has been 100%. I have not written- 15:42 真的吗？ Really? 15:43 ……Quatro 里发布的任何一行代码。所有代码的整体结构我都审过。 ... any of the code that shipped in Quatro by hand. I've reviewed the shape of all of it. 15:52 对。 Yeah. 15:52 系统模型层里所有关键的部分，我都逐行审过，而一大堆 UI 代码我没看。 I've reviewed the individual lines of anything that's critical in the model layer of the system, and I've not looked at a bunch of the UI code. 16:02 一大堆辅助代码我也没看，而且所有新功能没有一个是完全手工写的。 I have not looked at the bunch of the auxiliary code, and I've not written any of the new functionality entirely by hand. 16:12 但 Web 这块，也就是 Basecamp 和 HEY 的实际演进——我们的商业产品，用户很多，代码库也相当大， But then the web part, actually evolving Basecamp and HEY, our professional products that have lots of users and are relatively large code bases, 16:21 想完全用智能体加速，结果出人意料地棘手。我们不久前刚发布了 Basecamp 5。 have proven surprisingly tricky to fully accelerate with agents. We just released Basecamp 5 not too long ago. 16:31 那是 37signals 第一个真正由智能体加速的产品。因为我们从二月左右就进入了最后冲刺阶段，那时智能体已经很强了。 That was the first product at 37signals that was really agent accelerated. Because we were in this final sprint phase from around February. By then, agents were already good. 16:43 嗯哼。 Mm-hmm. 16:44 当时我们有一阵早期的兴奋劲儿：成了，问题解决。可以让设计师直接编程了，他们知道自己要什么功能，知道想做成什么样。 And we had this early surge of, it's solved. We could just have the designers do the programming. They know what features they want. They know what shape they want it to take. 16:55 就让他们凭感觉来吧。我们真就让他们凭感觉编程了。 Let just, let them vibe. And we let them vibe. 16:59 对。 Yeah. 16:59 结果我们收到一大堆 PR，单个拎出来可能一时都说得过去，但加在一起，把整个系统的架构毁了。 And we ended up with a lot of PRs that individually perhaps could have been justified for a hot moment, taken all together, destroyed the architecture of the system. 17:10 对。 Yeah. 17:10 我们实际上不得不手动清理，用人手一点点收拾残局，才回到一个感觉统一、连贯的架构。所以那个领域里我们还有点那样的空间。 And we actually had to clean up manually, mop it up by hand, by human hand, to get back to an architecture that felt cohesive and coherent. So we still have a bit of that space. 17:22 顺便说一句，那是二月的事，现在已经大不一样了。 That was February, by the way. Things are quite different now. 17:24 等等，这里面的教训是什么？是不是其中一个教训是：现阶段你必须得是程序员才能凭感觉编程？ Well, hold on a second. What's the lesson from that? Is that, is- is one of the lessons from that that you have to be a programmer at this stage to be able to vibe code? 17:33 在现有的、有分量的——代码库上凭感觉编程， To be able to vibe code on existing substantial- 17:38 我懂了 I see 17:38 ……哪怕只是 CRUD 代码库，如果你想保留当初让系统走到今天的那套架构。 ... code bases, even if they are crud, if you wanna retain the element of architecture that got that system to where it was. 17:45 这也是我反复强调的一点：当人们指责凭感觉编程的人是垃圾制造机时，我直接反问他们， Now, that's also a point where I've stressed many times that when people accuse vibe coders of being slob generators, I go right back at them and say, 17:54 “你看过普通程序员的产出吗？”那也是另一坨垃圾。 "Have you looked at the average programmer's output?" That is some other slob too. 18:02 如果你看过很多伟大公司幕后的样子，看过他们的代码库经手过三千个程序员之后变成什么样——惨不忍睹，绝对惨不忍睹。 If you've looked at behind the scenes of many great companies and what their code bases look like after there's been 3,000 humans through them, they're awful. Absolutely awful. 18:14 能说说你的直觉吗？ Can you tell me the intuition you have? 18:17 我退后一步观察那些我们都依赖的各种应用，比如 Adobe Photoshop 之类的，感觉它们的开发进度好像并没有加速。 I seem to just, stepping back and observing the different apps that we all rely on, I don't know, Adobe Photoshop, all this kind of stuff, there seems to be the progress on development there has not accelerated. 18:29 那为什么？从 Basecamp 这种成熟的、用户量巨大的产品上，你能总结出什么经验？ So why is it, what lessons can you draw from Basecamp, like well-established, huge user base? 18:35 为什么我们没有看到这些成熟应用在新更新、新功能之类的地方出现超高速的增长？ Why aren't we seeing like super rapid increase in like new updates, features, all this kind of stuff in these well-established apps? 18:44 原因有好几个。先说最关键的第一个：一旦是一个人类团队协作做一件事，瓶颈很少在实现上。 Multiple reasons. I'll start with the first one that's the most critical. As soon as you're having human teams work together on something, the bottleneck is rarely implementation. 18:56 瓶颈在人的带宽和沟通上。 It's human bandwidth and communication. 18:59 当你有一个产品经理、几个设计师、上面一个 VP、再上面一个 CTO，而每个人都想参与产品塑造的过程——因为大家都在证明自己存在的价值—— When you have a product manager and a couple of designers and a VP above them and a CTO above them, and everyone wants to be part of the shaping process because we're all justifying why we're here, 19:11 所有的生产力就都死在那儿了。 that's where all the productivity goes to die. 19:19 过去三个月做 Amachi 让我领悟到：要拿到那种神奇的 10 倍、100 倍，极少数情况下甚至 The revelation I've had working on Amachi the last three months is that to get that magical 10X, 100X, in a few rare cases, 19:29 1000 倍的生产力提升，你必须直接和智能体交互，不能让另一个人夹在中间当带宽中介，因为那实在太慢了。 1,000X productivity boost, you have to interact with the agents directly, and you cannot intermediate that bandwidth with another human because it's simply too slow. 19:49 一方面，这有点扫兴。我是说，我喜欢人，一起协作很美好。 And on the one hand, that's a bit of a bummer. I mean, I like humans, and it's great to work together. 19:55 但这也意味着，对于这些智能体能做到什么，我们得调整预期——如果驱动它们的是人，还有三层审批，加上大公司的整套官僚机器， But it also means we need to temper our expectations with what these agents- Can do if it's humans driving it, and you have three layers of approval and all the other machinery of a large corporation, 20:07 实现环节只是其中一小段。 the implementation part is only a small segment of it. 20:14 我要说的另一点是：大多数组织不知道自己想要什么，也不知道怎么把东西做得更好。 The other thing I'd say is that most organizations don't know what they want. They don't know how to make it better. 20:23 他们的瓶颈不在实现，而在想法，在愿景，在品味。 They're not bottlenecked on implementation, they're bottlenecked on ideas. They're bottlenecked on vision. They're bottlenecked on taste. 20:33 如果这些要素的供给超不过你的实现能力，那再强的实现也没用。你可以把一大堆烂点子都实现出来，然后呢？你要发布那玩意儿吗？ And if you don't have those element in excess of your implementational capacity, it doesn't help. So you can make a lot of shitty ideas come true, then what? Are you gonna ship that? 20:46 那听起来就像，我不知道，微软会搞出来的东西。那不是我们想复制的东西。那个流程其实……因为我们早就有这种能力了。 That sounds like, I don't know, something coming out of Microsoft. That's not what we're trying to replicate here. That process is not actually-- Because we already had this. 20:55 你退后一步想想那些巨无霸组织，他们动辄有上万名程序员可用。对，我这是在拿微软开涮。 If you step back for a moment and think of the towering organizations we, we have who have had tens of thousands of programmers at their disposal. I mean, I'm picking on Microsoft here. 21:08 我有时候也挺喜欢微软的，但这次就拿他们说事，因为他们几十年来有用不完的资源、用不完的编程产能，对吧？ I love Microsoft some of the time, but I'll pick on them in this case because they have had endless resources, endless programming capacity for decades, right? 21:18 这就说明了：光能写很多代码，并不能做出伟大、有吸引力的软件。另一点是，我们拥有这种产能才六个月左右。 This is what's showing us that just being able to write a lot of code does not produce great, compelling software. Now, the other thing is we've had this capacity for about six months. 21:32 就人类消化、内化新事物的时间尺度来说，这根本不算长。 That's not very long in human life cycle capacity of internalizing what's going on. 21:40 嗯。 Mm. 21:40 我觉得特别好笑的是，对 AI 的批评竟然是：“怎么还没更快？” And I think it's actually funny that the critique of AI is, "Why isn't it going faster?" 21:46 嗯哼。 Mm-hmm. 21:47 你在说什么啊？人类历史上没有任何一种进步有 AI 这么快，而你居然因为过去三个月我们还没把整个世界重写成软件乌托邦就急了？ What are you talking about? We've not had any other form of progress that has moved as fast as AI, and you're impatient because in the last three months we haven't rewritten the whole world and made it a utopia of software goodness? 21:59 你基本上是在说，现在人人都该换 Linux 了。其中一个论点就是：我们可以把 Linux 上没有的软件全部在 Linux 上重写一遍。 You're basically saying everybody should be switching to Linux now. One of the arguments is like, we can rewrite all the software that's not available in Linux in Linux. 22:06 Linux 是我多年的挚爱，但我还离不开 Windows、现在又离不开 Mac 的原因之一，呃，是视频剪辑，Premiere。 Linux has been the love of my life for many years, but one of the reasons I'm still attached to Windows and now, uh, Mac is, uh, because of video editing, Premiere. 22:15 我们会解决这个问题的，不过—— We're gonna fix that, but- 22:17 所以问题是：谁来为 Linux 打造 Premiere 和 Photoshop？感觉现在一个人就够了。 So the question is, who's gonna build Premiere and Photoshop for Linux? And it feels like one person can now. 22:27 百分之百，一个人就行。 100% one person can. 22:28 那么——我可以心急。其实，我或者任何人，心急都是自己动手的第一步（笑），对吧？就像—— And so-- And I'm, I'm allowed to be impatient. In fact, me or anyone else, impatience is the first step of doing it yourself [chuckles] , right? Like- 22:40 没错。 Correct. 22:40 而且有一个几万人的社区，用 Adobe Premiere 或 DaVinci 这些非线性视频编辑器的人，都知道那些痛点。呃，大家都有共鸣。你可以去 Reddit 看，去论坛看。 And there's a community of tens of thousands of people that use Adobe Premiere or DaVinci, all these, uh, nonlinear video editors, that know the frustrations. Uh, everybody shares them. You can look at Reddit. You can look at the forums. 22:53 人人都知道。呃，我不知道原因是什么。可能是管理层，可能是会议。就是你一直在讨论的那些东西——公司内部的官僚体系，把功能开发拖得慢吞吞。 Everybody knows it. Uh, I don't know what it is. It may be the management, it may be the meetings. All the things you've always discussed is the, uh, the, the bureaucracy that's in, in- inside companies that's slowing everything down in the development of features. 23:06 但感觉只要——只要把一些开发者放出来，配上一堆智能体，就能把这些全解决。 But it feels like there-- if you just let some of the developers loose with a bunch of agents, you can fix all that. 23:13 不过，按你刚才说的一切，也许解决之道是从零开始重做，也许通过开源，或者有人创办一家新公司，因为在大公司内部，现在很难 But again, based on everything you're saying, maybe the way to do that is to start over from scratch, maybe through open source or somebody launches a new company that within large companies now it's hard to, 23:25 真正放手让开发者去构建，从而全速进入智能体时代。 to really accelerate into the agentic era by letting developers just build. 23:33 这就是经典的创新者窘境。这些公司已经把旧玩法练得炉火纯青、根深蒂固，因此它们的整个结构、管理层级、 This is the classic innovator's dilemma. These companies have gotten so good, so established at the old way, and therefore their entire structure, management layers, 23:45 流程都是为一个已不复存在的时代调优的。但你没法让它们掉头，它们是超级油轮。 processes are tuned for a time that no longer exists. But you can't pivot that. These are supertankers. 23:57 掉头根本不会发生。这就是为什么科技行业终于迎来了这场颠覆。 It just doesn't happen. This is why we're finally getting this upset in the technology industry. 24:04 有段时间，我对 Apple 和 Google 在移动端的双头垄断非常恼火，因为移动端感觉是我们拥有的最重要的计算平台， For a while, I was very upset about the duopoly between Apple and Google on mobile, because it felt like mobile was the most important computing platform that we had, 24:14 而我看不到任何路径能掀翻 Google 或 Apple——它们高高在上，控制一切，收着过路费。但游戏变了。 and I could not see a path to unseat either Google or Apple sitting on top of that, controlling everything, taking their toll booth money. But the game has changed. 24:28 那不再是最重要的平台了。手机依然重要，但很多其他形态正在到来，不管是眼镜、耳机还是别的什么。 This is no longer the most important platform. Mobile phones are important, but there's also a lot of other form factors coming, whether it's glasses or it's earpieces or whatever else have you. 24:41 现在一切都在重新洗牌，而计算平台本身大概是 40 年来第一次重新洗牌，就说桌面端吧。Linux 从 91 年就存在了。 It's all in play now, but the computing platforms themselves are in play for the first time in probably 40 years, if you look at the desktop. Linux has been around since '91. 24:54 它在桌面端一直没有起飞、没有占领市场。它占领了除此之外的一切。 It's not taken off or taken over on the desktop. It's taken over everything else. 25:00 你桌上的所有设备、你的冰箱、你的烤面包机，全都跑着 Linux——除了你的电脑。 All the devices you have on your desk, your fridge, your toaster, everything runs Linux except for your computer. 25:08 说来好笑，你的 Android 其实就是 Linux，只是包得严严实实，你都认不出来了。但现在出现了一个机会窗口，而正如你所说， Now, funnily enough, your Android is actually Linux, but wrapped so sufficiently that you can't really recognize it. But now there's an opening, and the opening is exactly as you say it. 25:21 如果有一款你离不开的软件把你绑在 Windows 上，现在完全在你个人能力范围之内，你可以开始重写它。 If there's a piece of software you depended on that bound you to Windows, it is completely within reach for you personally to start rewriting it. 25:33 也许做不到 100% 的覆盖，但这就像关于 Microsoft Office 的老笑话。“我只用 5% 的功能。”是啊，可我们每个人用的是不同的 5%。那如果我们每人都只构建自己的那 5% 呢？ And maybe you won't get to 100% coverage, but this is the old joke about Microsoft Office. "I only use 5%." Yeah, well, we all use a different 5%. Well, what if we all just build our own 5%? 25:45 嗯哼。 Mm-hmm. 25:45 如果我只取我需要的那部分功能，就把它做出来呢？ What if I just took the functionality that I need and just did that? 25:50 那是一个完全不同的挑战，而且是智能体此时此刻就极其擅长的事，并且已经干成过很多次了。 That is a completely different challenge and one agents are incredibly capable of doing right now today, and have done it a lot of times over. 25:59 我在最新的智能体时代发现的一件神奇的事是：我成了一个多语言程序员。这以前跟我完全不沾边。 So one of the amazing things I've found with the latest agentic age is I have become a polyglot programmer Something I was absolutely not before. 26:11 我以前是 Ruby 程序员—— I was a Ruby programmer- 26:13 对 Yeah 26:13 ……首先是 Ruby，然后在不得已的时候随便写点 Bash—— ... first and foremost, and then I dabbled a little bit in Bash- 26:16 对 Yeah 26:17 ……而过去两个月里，我写了 C++。我写了三个已经随 Omachi Quattro 发布的应用。我写了一个写作应用。 ... when I had to. And in the last two months, I've written C++. I've written three applications that have shipped in Omachi Quattro. I wrote a writing app. 26:30 我之前在用一个叫 Typora 的应用，是个很好的应用—— I was using this app called Typora, which is a very nice app- 26:32 很棒，很棒的应用 Great, great app 26:33 ……它本身的灵感来自另一个叫 iA Writer 的应用，那才是我在 Mac 上的真爱—— ... which in itself is based on another app called iA Writer, which was my real love on the Mac- 26:38 嗯哼 Mm-hmm 26:38 ……一个干净、简单的 Markdown 写作环境。我所有的文章都是在那儿写的。 ... for a clean, simple markdown writing environment. That's where I write all my essays. 26:44 嗯哼。 Mm-hmm. 26:45 后来我换到 Linux，用不了 iA Writer，就转到了 Typora。它是个共享软件，有一大堆我根本不需要的功能。 And then I moved to Linux, and I couldn't get iA Writer, so I moved over to this other tool called Typora. It was a piece of shareware, and it had a lot of features I just didn't need. 26:55 大概六周、七周前，我想：“你知道吗？我只需要 Typora 的 5%”——而它本来就已经是个很基础的应用了。 And about six weeks, seven weeks ago, I thought, "Do you know what? I only need 5% of Typora," which is already a basic app. 27:04 我直接让智能体开工，说要用 C++ 和 Qt 写，因为那跟我在 Quattro 上构建的美学风格很搭。 I literally told the agent to get going, that I wanted it written in C++ and Qt because that would fit well with the aesthetic of what I was building with Quattro. 27:14 大概 20 分钟，第一版就出来了。我开始用，感觉还不太对。 And in, I think, about 20 minutes, it had the first version. I started using it, and it wasn't quite right. 27:20 两天之内，我就放弃了 Typora。从那一刻起，我所有的文章都是在 OmaWrite 里写的。 Within two days, I'd given up Typora, and I wrote and have written all of my essays since that moment in OmaWrite. 27:30 我能请教一下你的建议和愿景吗？关于智能体，我发现一件事：你完全可以写一个像 Typora 替代品那样的东西，完美贴合你自己、贴合你的需求—— Can I ask for your advice and your vision about something? So what I find with, with agents, you can actually write something like a replacement for Typora that is perfectly customized to you, to your needs- 27:42 没错 Correct 27:43 ……只为你一个用户服务；然后还有另一种版本——从一个用户出发，但扩展到更大的受众。 ... with just one user in mind, and then there's a version of that that is- starts with one user in mind but expands to a larger audience. 27:53 我其实已经写了大量只给我自己用的软件，我发现这样做更容易，也更能把我的事办好。 So I've actually written a large number of software just for me, and I find it easier to do and better to get my situation done. 28:01 比如，我给自己写了一个视频编辑器，只服务我一个人，这很简单。 For example, I have a video editor I've written, which is very easy to do just for me. 28:06 对。 Yes. 28:06 跟你不一样，我从没做过被大量用户使用的东西，我觉得那有点吓人、有点让人发怵。 I've never, uh, unlike you, built something that's used by a large number of people, and I find that a bit scary and intimidating. 28:14 那么在这个智能体时代，怎么迈出那一步——先做出解决你自己问题的东西（我觉得这是很美的构建方式），再把它扩展到对别人也有用？ So, like, at this, with this agentic age, how do you take a step to actually build something that solves your problem, which I think is a beautiful way to build, but also expand it to actually useful for other people? 28:27 你有什么建议吗？另外，你是否也看到了这个问题：只给自己造工具实在太容易了？ Like, is there advice you can give, and do you also see the problem of how easy it is to just build a tool for yourself? 28:35 是的，但这比你想的简单。等你把给自己用的工具做完，直接叫你的智能体把它放上 GitHub 就行。 Yes, but it is simpler than you think. By the time you're done building the tool for yourself, you simply tell your agent to put it on GitHub. 28:43 对。 Yeah. 28:43 别的你什么都不用做。它会自己弄明白怎么把仓库放上 GitHub，会写一份漂亮的 README，会开始用 GitHub Releases 帮你追踪版本。 You don't even have to do anything else. It will figure out how to put that repo on GitHub. It'll write a nice README. It'll start using GitHub releases so that you can track things. 28:52 它实际上会成为一个比你强得多的软件维护者—— It'll actually be a better software maintainer than you could ever be- 28:56 嗯哼 Mm-hmm 28:57 ……因为它远比你有耐心，在打理开源软件那些苦差事上，它远比你勤勉。 ... because it is far more patient. It is far more diligent in doing the drudgery of managing open source software than you are. 29:05 嗯哼。 Mm-hmm. 29:05 而你只管享受使用自己构建的软件的乐趣，按自己的想法让它演进。这就引出了我们在开源界争论已久的大问题。 And you can simply get on with the joy of using the software that you built and evolving it as you see fit. Now, this gets into the big argument we've been having in open source for a while. 29:18 AI 贡献对维护者来说，到底是好事还是坏事？ Is AI contributions good or bad for the maintainer? 29:23 现在很多维护者相当恼火，因为他们突然收到海量涌入的 PR——来自一些可能不是顶尖程序员、甚至压根不是程序员的人的贡献—— There's a lot of maintainers right now who are quite upset about the fact that they're suddenly getting a huge influx of pull requests, of contributions from people who may not be the best programmers or programmers at all- 29:40 嗯哼 Mm-hmm 29:40 ……涌向他们的软件。 ... towards their software. 29:43 我看着这种争论就想：“你开玩笑吧？”眼前明明有一大条免费的贡献矿脉，你要也行不要也行， I look at that argument and go, "Are you kidding me?" Here is a vein of free contributions that you can take or don't take, 29:52 可你居然抱怨它们的存在？这——听起来就像“我的牛排太多汁了，我的龙虾黄油太多了”。 but you're complaining about the fact that they're there? It- Sounds straight out of my steak is too juicy and my lobster too buttery. 30:05 你到底在抱怨什么？ What are you complaining about? 30:07 对。 Yeah. 30:07 这简直是最美妙的事。我们念叨了那么多年，说开源的魔力就在于人人都能参与贡献，但那当然从来没成真过。 This is the most amazing thing. This is what we proclamated for so long was going to be the magic of open source, that we could all contribute to it, but it was, of course, never true. 30:17 真正给开源做贡献的一直只有一小撮人，那些技艺高超的巫师。所以这有点像一场宗教改革时刻。 There was only a small subset of people who contribute to open source, and that was the very highly skilled wizards. So this is a bit of a reformation moment. 30:26 我们和计算机之间曾经隔着一层中介，那就是一个叫“程序员”的神职阶层，而现在他们正在被绕过、被去中介化，就像—— We had this disintermediation between us and the computer, and it was this class of clerics and priests called programmers, and suddenly they're being disintermediated here by, 30:37 呃，是不是 16 世纪路德把九十五条论纲钉在教堂门上那回事？ um, was it the 95 theses by Luther in, in 1500s, nailing them to the door? 30:44 其中一条就很明确地讲中介问题：人应该可以直接接触到更高的智慧。 And one of those was very clearly about the intermediation, that you should have direct access to the higher intelligence. 30:51 嗯哼。但你不需要那些巫师来维持代码库的高水准吗？ Mm-hmm. But don't you need the wizards to keep a high bar of excellence in the code base? 30:56 百分之百需要。这一直是开源的真相，也引出了第二个让我很不爽的论点。我运营开源项目 25 年了。 100% you do, which is also something that's always been true about open source, which is the second argument that really grinds me. I've been running open source projects for 25 years. 31:06 我实打实看过几千、甚至上万个程序员的产出。 I have literally looked at the output of thousands, if not tens of thousands, of programmers. 31:12 我觉得我有足够的统计样本敢说：大多数程序员——他们很菜。 I feel like I have the statistical basis to assert that most programmers they suck. 31:21 对。 Yeah. 31:21 我不是你们听起来那个意思。 And I don't mean that in the way it sounds. 31:24 对。 Yeah. 31:24 所以我刚才特意停顿了一下。我的意思是，他们菜在：写不出我想要的代码，提交的 bug 报告不带齐相关信息， That was why I paused for a hot second here. I mean, they suck in the sense that they don't write the code I want to have written. They don't prepare their bug reports with all the relevant information. 31:36 PR 里不写清楚为什么，懒得补必要的代码注释，不检查自己的工作，不写单元测试。 They don't detail their pull requests with the why. They don't bother to fill in needed code comments. They don't double-check their work. They don't write unit tests. 31:47 创造好软件需要的这些事他们统统不做，但你知道谁会把这些全做了吗？智能体——只要你吩咐。它们非常勤恳地按你的指令办事，绝大多数时候如此，有时甚至勤恳过头。 They don't do all of these things it takes to create good software, but do you know who'll do all that stuff? Agents, if you tell them to. They're very diligent at following your instructions, most of the time, sometimes to a fault. 32:00 这实际上意味着：拿普通水平的程序员和他们投向普通开源项目的 PR 来说，他们已经被智能体比下去了。 But it actually means that if you take the median programmer and their pull requests towards an average open source project, they're already getting outclassed by agents. 32:11 比起人类写的 PR，我更愿意收到智能体写给我的项目的 PR。而且不只是因为质量更好。 I would rather get an agent-written pull request to one of my projects than I'd get one written by a human, and it's not just because the quality's better. 32:22 还因为我拒绝它的时候心理负担小得多。代码又不是你亲手写的，所以我大可以看看你的智能体替你写的东西，说：“呃，不想要。” It's also because I feel a lot less bad if I just reject it. You didn't even write it, so I can simply look at what your agent wrote on your behalf and go, "Eh, don't want it." 32:36 这在开源界一直是真理。事实上，我认为这正是开源维护者的问题之一。 That's always been true in open source, and in fact, I think this is one of the problems with open source maintainers. 32:41 他们背着太大一团神经质和焦虑，总觉得任何贡献都是自己的义务，必须想尽办法把它合进代码库。 They have way too much ball of neuroticism and anxiety that requires them to look upon any contribution as an obligation on their part to do everything to get that to land in the code base. 32:59 根本不是这样。你大可以说“不”，或者更好，说“不了，谢谢”。 That's not true at all. You can simply say, "No," or even better, you can say, "No thank you." 33:05 一旦你进入这种心态——你的项目有权按你的愿景、你的路线图存在和演进——在智能体时代拒绝你不想要的贡献就轻松多了，因为你甚至不会麻烦到或伤害任何一个真人的感情。 And if you get into the mode of realizing that your project is allowed to exist and evolve according to your vision, according to your roadmap, it becomes so much easier in the agentic age to decline the contributions you don't want because you don't even inconvenience or hurt the feelings of a human. 33:27 那只是个“铁疙瘩”（clanker），而铁疙瘩不会介意。事实上，生产铁疙瘩的那些实验室巴不得你白烧 token，反正钱照收。 It's just a clanker, and the clanker won't mind. In fact, the producers of clankers, the labs, love when you spend tokens in vain. They get paid just the same. 33:38 嗯哼。 Mm-hmm. 33:39 所以……依我看，现在是有史以来当开源软件维护者最好的时代。 So I... In my opinion, this is the absolute best time to ever have been an open source software maintainer. 33:46 没错。 Right. 33:46 我们不仅收获了一大堆智能体写的、各项清单全打勾的精彩 PR，还能接触到一群以前根本没机会给项目做贡献的人的创造力。 Not only do we get this wealth of glorious pull requests made by agents with all the boxes ticked, we also get to tap into a creativity of people who did not have access to contribute to a project before. 34:02 嗯哼。 Mm-hmm. 34:02 我运营 Omachi 项目一年多了，过去三个月做 Quattro 期间，我合并了超过 1000 个 PR。 I've been running the Omachi project now for a little over a year, and in the last three months working on Quattro, I have merged over 1,000 pull requests. 34:11 嗯哼。 Mm-hmm. 34:12 其中相当多的 PR 出自非传统程序员之手，或者是其他领域的程序员——不是做 Linux 操作系统或发行版开发的。 Quite a lot of those pull requests were written by people who were not classical programmers, or were programmers in other domains, not Linux operating system or distribution development. 34:25 因为智能体，他们得以贡献自己的好想法，而我得以从中挑最好的—— They were able to contribute their good ideas because of agents, and I got to cherry-pick the very best of that- 34:32 嗯哼 Mm-hmm 34:32 ……因为那些想法就摆在那儿。否则，这些想法只会烂在他们脑子里。再说一次，这不就是…… ... because it was there. Otherwise, those ideas would just have lived inside their heads. Again, is... 34:40 这不正是开源的意义吗？汇聚全地球的集体智慧和创造力，把它引向一片公地， That's not the purpose of open source, that we tap into the collective intelligence and creativity of the whole goddamn planet, and we channel that towards a commons, 34:49 让我们所有人都受益于所有人共同的努力。 where we all benefit from the combined efforts of everyone. 34:55 嗯哼。 Mm-hmm. 34:56 现在 Omachi 上大概有 400 个未合并的 PR，差不多是一周前的两倍。所以加速是真的，但工具也是真的强。 Now, there's currently, I think, about 400 unmerged pull requests on Omachi, about double what it was a week ago. So the acceleration is real, but so are the tools. 35:07 嗯哼。 Mm-hmm. 35:08 我不再亲自审查每一个 PR 了。已经有好一阵子不亲自审了。我让智能体—— I'm not reviewing every pull request anymore. I haven't been reviewing them for quite some time now. I have the agents- 35:14 嗯 Mm 35:15 ……替我审—— ... review them for me- 35:16 嗯哼 Mm-hmm 35:16 ……然后它们给我一份摘要，告诉我哪些已经就绪、可以交给人类做决策了。 ... and then they will give me a summary about whether something is ready for the decision, the human decision. 35:22 嗯哼。 Mm-hmm. 35:22 该合并，还是不合并？ Should we merge, or should we not merge? 35:24 嗯哼。 Mm-hmm. 35:25 那些跑偏的、重复的、糟糕的 PR，我不需要看。 I don't need to look at all the pull requests that are either wrong, duplicated, bad. 35:31 嗯哼。 Mm-hmm. 35:31 智能体可以替我把那些糟粕筛掉。我只看珍珠—— An agent can sort that chaff away from me. I just get to look at the pearls- 35:35 嗯哼 Mm-hmm 35:35 ……那些已经可以上手的好想法，那些智能体已经替我在虚拟机里验证过的 bug 修复。 ... the good ideas that are ready to go, the bug fixes that the agent has validated in a VM on my behalf. 35:44 复制和管理开源项目的所有苦差事正在以闪电般的速度蒸发，留给我们的都是黄金多汁的部分， All the drudgery of replicating and managing open source projects is evaporating at lightning speed, and we are left with the, with the golden juicy parts, 35:54 软件开发的骨髓—— the bone marrow of software development- 35:59 嗯哼 Mm-hmm 35:59 ……决定这东西该做什么、该往哪儿走。 ... deciding what should this thing do and where should it go. 36:03 你认为好想法——好想法的内核——最终源自人类的大脑吗？也就是那些个人贡献者……能不能全靠智能体， Do you think the good ideas, ultimately, the kernel of the good ideas are originated in the human mind, so the individual contributors- Like, can it be done by agents, 36:13 还是说 PR 最终必须有人类的想法？比如怎么改进 Amachi，所有那些涌进来的……还是说可以完全就是一池子智能体？ or is the PR ultimately requires a human idea? Like how to improve Amachi, all those incoming... Or can it all just be a pool of agents? 36:24 那是我在第一个智能体时代——11 月 24 日到 2 月 28 日——对智能体的思考模式和方式。 That was the mode and the way, uh, I was thinking about agents in the first agentic age from November 24 to February 28. 36:35 嗯哼。 Mm-hmm. 36:35 我以为所有的想法都会源自人类，他们告诉自己的智能体要造什么，然后智能体就开工。 I thought all the ideas would originate with the humans. They would tell their agents what to build, and off they went. 36:42 嗯哼。 Mm-hmm. 36:43 我现在完全不这么认为了。我见证了你们人类难以置信的东西。模型里蹦出来的想法好到让我自愧不如—— I don't think that's true anymore at all. I have seen things you people wouldn't believe. Ideas coming out of models so great that it makes me humble- 36:54 对 Yeah 36:54 ……要知道我可是个一向以自己点子好为傲的人。 ... as a person who otherwise prides himself on having good ideas. 36:58 嗯哼。 Mm-hmm. 36:59 智能体的创造性思维能力惊人，而那些还抱着“智能体只是鹦鹉”那套分析不放的人—— The agents are incredibly capable of creative thought, and folks who are still stuck in the analysis that agents are parrots- 37:11 嗯哼 Mm-hmm 37:11 ……认为它们只是在反刍已有想法的人，对过去六到九个月取得的进步活在妄想里。 ... just regurgitating the ideas that are already there are delusional about the progress that's been made in the last six to nine months. 37:21 现在有些听你说话的人会说：DHH 这是犯了老毛病——AI 精神病。 So some people listening to you right now will say DHH is suffering from the old case of, uh, AI psychosis. 37:28 呃，你能不能为“你确实处在妄想状态”这个观点做一个最强版本的论证，就像《飞越疯人院》那样，然后再反驳它？ Uh, can you steel man the case, uh, that you are in fact experiencing, uh, in a state of delusion, like One Flew Over the Cuckoo's Nest, and can you argue against it? 37:37 我处在狂喜状态。我就处在那种状态里。 I'm in a state of delirium. That's what the state I'm in. 37:41 因为我跟计算机打了 40 年交道，从没见过仅仅过去两个月里我看到的这些东西。 Because I have been working with computers for 40 years, and I've never seen the things that I've seen in just the last two months. 37:53 对。 Yeah. 37:54 整个职业生涯、整个人生都献给对计算机的热爱，工作时间乃至工作时间之外都以它为中心，突然之间被彻底颠覆， A whole career, a whole life dedicating to the love of computers as my main focus for the working hours and then some, suddenly completely upended, 38:04 突然之间被彻底重写。我当然狂喜。看着这个现实的人，谁不狂喜？ suddenly completely rewritten. Of course I'm delirious. Who looking at this reality is not? 38:14 如果你不狂喜，或者至少不非常兴奋……其实，我想这说法都不太对。如果你意识不到当下这一刻的分量，那才是妄想。 If you are not delirious or at the very least very excited... Actually, I guess that's not even true. If you're not recognizing the gravity of the moment, that's the delusion. 38:27 那才是精神病。真正的精神病是相信世界几乎没变，觉得我们只是有了几只背诵句子的电子鹦鹉。 That's the psychosis. The psychosis is believing that the world is barely different. We just have some electronic parrots reciting things to us. 38:37 依我看，这种论点之所以被提出来，是因为人们看不到进步的成果。这正好接你刚才说的：惊人的软件在哪儿？ Now, in my opinion, the reason this argument is actually brought up is because people are not able to see the fruits of the progress. That goes to your point earlier. Where is the amazing software? 38:49 美国 GDP 怎么没同比涨到 12%？首先，给点时间行不行？到现在连一年都不到，真是的。 Why isn't American GDP running at 12% year over year? First of all, give it a minute. It hasn't even been a year, goddammit. 39:00 但其次，就我个人而言，我觉得我已经把布丁端上桌了。 But second of all, I feel on my own account, I have actually brought the pudding. 39:07 嗯哼。 Mm-hmm. 39:08 Umachi Quattro 周五发布了，已经被几万人下载—— Umachi Quattro launched on Friday. It's been downloaded by tens of thousands of people- 39:14 嗯哼 Mm-hmm 39:14 ……而且他们喜欢它，非常喜欢。 ... and they like it. They like it a lot. 39:17 我们应该提一下，呃，根据 Perplexity 的说法：Umachi 是一个高度有主见的、基于 Arch Linux 的桌面发行版，围绕 Hyprland Wayland 平铺合成器构建。 We should mention that, uh, going to Perplexity here, Umachi is a highly opinionated Arch Linux-based desktop distribution centered on the Hyprland Wayland tiling compositor. 39:27 它由 David Heinemeier Hansson，也就是 DHH 创建，是一个打磨精致的开发者工作站，视觉风格统一，预装配置好了许多日常工具。 It was created by David Heinemeier Hansson, DHH, as a polished developer workstation set up with a cohesive visual style and many everyday tools already configured. 39:39 让我来说：Umachi 是一个漂亮、现代、有主见的 Linux 系统。 Let me put it. Umachi is a beautiful, modern, and opinionated Linux system. 39:45 嗯哼。 Mm-hmm. 39:45 它是 macOS 和 Windows 之外的另一种操作系统选择，而且它好用到爆。 It's an alternative operating system to macOS and Windows, and it's freaking amazing. 39:50 而 Quattro 这个版本，呃，是你在向智能体时代推进。 And the Quattro thing is, uh, you're pushing towards the age of agents. 39:54 没错。那是我们刚发布的最新版本。说起来也好笑，Umachi 项目才刚满一年多一点。我去年夏天在勒芒 24 小时耐力赛的比赛间隙启动的它—— That's right. That's the latest version that we just dropped out, which is also funny. The Umachi project is only Be a year old and a little bit. I started it last summer in between sessions at the 24 Hours of Le Mans- 40:07 嗯。 Yep. 40:08 ……当时我在 YouTube 上看了太多 Linux 桌面美化（ricing）视频，中了那个毒。 ... where I had watched, uh, one too many YouTube videos on Linux ricings and got bitten by that bug. 40:15 于是开始做我这第二次尝试——造一个更好的 Linux 操作系统。 Started working on this second iteration of my attempt at creating a better Linux operating system. 40:22 第一次尝试我们上次聊过，叫 Omaku，是建立在现成的 Ubuntu 系统之上的，还不错。但我能给新版本 Omachi 注入的野心完全不同， The first iteration we talked about last time, was called Omaku, was built on top of an existing system called Ubuntu, and it was fine, but the ambition I was able to pour into the new version, 40:31 因为我是从软件栈往下深七层的地方开始的。不过它依然是在前智能体时代完成的。 Omachi, because I started seven layers deeper down the stack, was completely different, but it was still done in a pre-agentic world. 40:42 最开始那些脚本全是我手写的，后来我亲眼见证了中途切换的那一刻。 I wrote all those batch scripts by hand in the beginning, and then I got to see the midway point where things changed over. 40:51 嗯哼。 Mm-hmm. 40:52 我做过几个部分由智能体加速的 Omachi 版本，然后三个月前，就全油门了：100%，一切都由智能体写，但由我来掌舵。 I did a couple of versions of Omachi that had partial agent acceleration, and then three months ago, it was full throttle, 100%, everything is written by agents, but steered by me. 41:03 嗯哼。 Mm-hmm. 41:04 结果那是一种完全不同的体验，造就了一个完全不同的操作系统，因为我的野心突然被给了无上限的天花板。 And that just ended up being a very different experience and a very different operating system because I was suddenly granted a limitless ceiling on my ambition. 41:16 嗯哼。 Mm-hmm. 41:17 我可以看着 Windows、Mac、其他 Linux 系统上的任何功能说：“我要这个，”然后智能体就会交付。 I could look at any feature in Windows, on Mac, on other Linux systems and say, "I want that," and the agents would deliver. 41:29 所以我想要的一切突然都触手可及，这意味着我有点狂喜上头了。 So everything I wanted was suddenly within reach, which meant that I got a little delirious. 41:35 嗯哼。 Mm-hmm. 41:36 如果一个精灵突然从瓶子里蹦出来说：“你想要什么都行。你在操作系统里梦寐以求的每一个功能——”，谁不会狂喜呢？ Who would not get delirious if suddenly a genie pops out of the bottle and says, "You can have whatever you want. Every feature you've ever dreamed of in an operating system- 41:46 嗯哼 Mm-hmm 41:47 ……我可以把它们交付给你，大部分五分钟搞定，有几个要二十分钟，如果我们真的放开手脚大干一场，那得花我两个小时。 ... I can deliver them to you, most of them in five minutes, a few in 20, and if we really go hog wild, it's gonna take me two hours." 41:54 嗯哼。你看过《Requiem for a Dream》那部电影吗？就是说，呃，这里面有一种类似嗑药的效果。 Mm-hmm. Have you seen, uh, Requiem for a Dream, the movie? So the, uh, there's a drug-like effect here. 42:00 是的。 Yes. 42:00 我、我个人觉得……我们会从很多层面聊这个，但、但让人难以承受的是，这么大的力量突然间就放到了我们手里，呃，很难知道该怎么办，就像《The Hobbit》里那样， I, I find it personally... We'll talk about many levels of this, but it, it is overwhelming how much power is placed in our hands im- uh, so suddenly, and it's very hard to know, it's like with The Hobbit or something, 42:12 该拿它怎么办。我发现自己，呃，真的被多任务压得喘不过气来，呃，几乎到了快燃尽的边缘。 what to do with it, and I find myself, uh, truly overwhelmed with the multitasking, uh, on the verge of almost burnt out. 42:24 呃，对很多事情都超级兴奋，但，呃，归根结底回头看，你还是想专注在一件事上，真正把它做出来发布。 Uh, super excited about so many things, but, uh, ultimately looking back, you want to focus on one thing and really ship. 42:31 我没有那种感觉。我没有被压垮的感觉，没有恐惧的感觉，没有迷失的感觉，因为我有一个使命。 I don't have that sense. I don't have the sense of overwhelm. I don't have the sense of dread. I don't have the sense of distortion because I have a mission. 42:42 嗯哼。 Mm-hmm. 42:42 我有要去的地方，因此我能把所有这些新获得的力量引导向—— I'm going somewhere, and therefore I can channel all this new power towards- 42:48 是啊，这很美妙。 Yeah, it's beautiful 42:48 ……一个单一的目标和结果：打造一台完美的电脑。因此我所有对 AI 的探索都聚焦在这个方向上。 ... a singular goal and outcome, creating the perfect computer, and therefore all my investigations with AI are focused towards that end. 43:00 当然，我还有一份正职工作。 Now, I also have a day job. 43:02 对。 Yep. 43:02 我们在那边也用智能体，那也是针对某种特定结果的。但在 Omarchy 的体验里，我等于直接接上了后脑勺，把想法涌入大脑和软件在屏幕上成形之间的带宽提升了。 And we also use agents there, and that's also targeted towards sort of specific outcomes, but with the Omachi experience, I got to tap in straight in the back of my head and increase the bandwidth between ideas arriving in my brain and software merging on the screen. 43:25 那感觉就像从、从拨号上网—— It was like going from, from dial-up- 43:28 嗯哼。 Mm-hmm 43:28 ……跨到了光纤。我记得在你的播客上，Elon 聊过真正的人类带宽—— ... to fiber. I think on your podcast, Elon talked about the actual human bandwidth- 43:35 嗯哼。 Mm-hmm 43:35 ……也就是 Neuralink 和其他项目正在努力提升的东西。 ... that Neuralink and other projects are trying to accelerate. 43:39 我们现在用来交流的带宽其实相当低，因为我们受限于自己的认知，受限于，呃，说话的速度。 Like, the bandwidth we're communicating over right now is quite low because we're limited by our cognition. We're limited by, uh, the rate of speech. 43:51 但当你和一群智能体一起工作时，就不受这些限制了。不过我得先声明一下：如果我听到自己这样说话—— You're not limited by those things when you're working with a swarm of agents. Now, let me caveat this by saying if I'd heard myself talk like this- 44:02 嗯哼。 Mm-hmm 44:03 ……在九个月前，我大概会给自己贴上“AI、AI 精神病”的标签。 ... nine months ago, I would probably have used the label AI, AI psychosis. 44:08 我觉得这个标签挺贴切的，因为九个月前这样说话的人并没有真正交付东西。我觉得这才是根本的区别：确实有人很早就看到了这一切，而我其实没那么早。 I think it would be fitting because nine months ago when people were talking like this, they weren't shipping, and I think that's the ultimate difference, that there were people who saw this early. I was not actually that early. 44:21 就像我们聊过的，我当时相当怀疑。我也在以某些方式使用 AI，但我不是第一个下载 Claude Code 的人。我记得 Boris 是在二月底发布的，我到九月份才装上 Claude Code。 As we've talked about, I was rather skeptical, and I was using AI in certain ways, but I was not the first person who downloaded Claude Code. I think Boris released that in end of February. I don't think I installed Claude Code until September. 44:35 所以中间大概有六个月，先行者们看到了未来模样的点点微光。那时它还没成真，因为智能还不够，脚手架工具也还不行，但他们能看到那些微光。我看不到。 So there was about six months there where pioneers saw these glimmers of what the future was gonna look like. It wasn't there yet because the intelligence wasn't there. The harnesses weren't there, but they could see the glimmers. I couldn't. 44:48 Tobi，Tobi Lütke，Shopify 的 CEO，现在也算是我在这个，呃，Omarchy 项目上的“同伙”，他也彻底迷上 Omarchy 了。 Tobi, Tobi Lütke, the CEO of Shopify, and now my partner in crime on this, uh, Omachi project in part, he's gotten pilled with Omachi as well. 44:58 他很早就看到了这些东西，他试图跟我说，但我没看出来。我就是看不到他所看到的。 He saw these things very early, and he tried to tell me, and I wasn't seeing it. I was not seeing what he was seeing. 45:07 你当时是个喷子。 You were a hater. 45:08 是的。我当时、我当时还在地球上呢。 Yes. I was, I was still on Earth. 45:10 好吧。 Okay. 45:11 而他……他已经登上火箭、正飞向太空了。我还记得我读了他给公司发的那份备忘录，那份 AI 备忘录，心想：“啊，是不是有点夸张了。” And he was... He had already boarded the rocket and was heading towards space. I remember actually reading his memo to the company, the AI memo, and thinking, "Ah, seems a little much." 45:24 嗯。 Mm. 45:24 就好比，你在为一件我还感觉不到是实实在在的东西大惊小怪。 Like, you're making a big deal out of something that I cannot yet feel is a tangible thing. 45:31 嗯哼。 Mm-hmm. 45:32 因为这玩意儿的产出在哪呢？我试着用过。它写的代码我不喜欢。它老想打断我。你这结论是从哪来的？呃，所以我理解，我完全理解那种怀疑。 Because where's the output of this? I try using it. It writes code I don't like. It wants to interrupt me all the time. Where are you getting this from? Uh, so I get it. I totally get it. 45:43 如果你还没见过当前这种智能水平能产出什么，你当然会看着某个人心想：“这人听着有点疯”，因为你到目前为止的所有经验都会告诉你他们就是疯了。因为如果你看看计算机编程的历史， If you've not seen what the current quality of intelligence can produce, of course you're gonna look at someone and go like, "They sound a little nutty" because all your experience up until this point would tell you that they are nutty because if you look at the history of computer programming, 46:00 过去四五十年里，人们一直在承诺普通人只要动动嘴就能写代码，说什么我们将迎来第四代语言。 for the last 40, 50 years, people have been promising That regular people could write code just by talking. We're gonna have fourth generation languages. 46:12 Lisp 曾经、Smalltalk 也曾经被包装成能让普通人自己写应用程序的环境，但这些都没真正实现——除了 Microsoft Access 数据库或 Excel 电子表格那种意义上的例外。 Lisp was once upon a time presented, Smalltalk was presented as these environments that would let regular people write their own applications, and none of it came true in the sense beyond like Microsoft Access databases or Excel spreadsheets. 46:28 那些算是终端用户编程环境，但跟我们今天能做到的完全没法比。 That, those are end user programming environments, but it's nothing like what we're able to do now. 46:33 而且，我说的还不是终端用户，我说的是我自己——一个写了 25 年代码的人——亲眼看到那种产出、那种加速、那种质量，然后把它发布出去。 And then, I mean, I'm not even talking about end user, I'm talking as me, someone who's been programming for 25 years, seeing the output and the acceleration and the quality and then shipping it. 46:45 我认为这才是会迅速改变整个讨论走向的东西。 And I think that's what's gonna change the entire conversation quite quickly. 46:49 剩下那几个落后者还在观望这东西到底好不好用，还困在 2025 年初那个“机械鹦鹉”的梗里，他们很快就会被即将铺天盖地而来的证据淹没。 The few laggards who are still on the fence about whether this is actually useful, who are still trapped in a meme from early '25 about mechanical parrots, they're simply going to be overwhelmed by the evidence that's about to flood over them. 47:05 所以我希望看到，呃，像 Omarchy 这样的项目，多来几个，真正展示智能体优先—— So I hope to see, uh, projects like Homachi, several of them, that like really show agent first- 47:15 会来的。这一切都会来的。 It's coming. It's, all of it is coming 47:17 ……出现在那些领域里，就是我已经提过的那些不起眼的领域，比如视频剪辑，所有这类应用，比如 Typora—— ... in the spaces, in the silly spaces that I already mentioned, like video editing, all of these kinds of apps that rely, like Typora- 47:24 对。 Yes 47:24 ……这类东西。 ... this kind of stuff. 47:26 它会来的，而且我已经看到了早期的苗头。 It's gonna come, and I've already seen the early inklings of this. 47:30 Omarchy 一个非常棒的特性是它有一个强大得多的插件系统，让你可以扩展你的操作系统，甚至可以说重写它， One of the really neat features about Homachi is that it has a far more robust plugin system that allows you to extend your operating system and rewrite it really, 47:41 包括它的用户界面、面板和功能——通过创建你自己的插件，而这些插件可以从我们自带的工具克隆出来。 in terms of its user interface and its panel as, and its features, by creating your own plugins that can be cloned off of tools we ship in the box. 47:54 举个例子，假设你想要一个不一样的日历。Omarchy 自带一个日历，你点那个小时钟，它会弹出日历，但它不支持 iCal，比如，它读不进你的日程。 For example, say you want a different calendar. Homachi ships with a calendar. You click the little clock, it pops up a calendar, and it doesn't have iCal support, for example. It doesn't consume your appointments. 48:07 现在已经有大约 17 个实现了。而且三天之内，Omarchy 插件市场上就有了 330 个插件。 There's about 17 implementations of that already, and in three days, we had 330 plugins on the Homachi plugin marketplace. 48:18 在我参与过的所有项目里，我从没见过这样的增长速度，也从没见过这么广泛的参与。 I have never seen growth like that with any project I've ever been involved in. I've never seen participation that broadly. 48:26 我从没见过这么多人能这么快就创造出对自己有意义、对别人也可用的软件。 I've never seen that many people be able to create software that's meaningful for them, also be usable for others so quickly. 48:36 这一切都源于 Omarchy 自带了一套技能（skills），它告诉你接入的任何智能体如何为操作系统创建扩展， All of it is driven by the fact that Homachi ships a set of skills that tells any agent you bring to it how to create extensions to the operating system, 48:45 由此让你看到真正可塑的…… and therefore affording you the vision of the true malleable... 48:53 我刚想说“agentic”。我他妈讨厌那个词。 I was about to say agentic. I hate that fucking word. 48:56 是啊。 Yeah. 48:56 我讨厌那个词的部分原因是，首先，它已经变成了营销圈里的口水话—— And the reason in part I hate that word is, first of all, it's become marketing slob speak- 49:00 对。 Yeah 49:00 ……到现在这个地步了。 ... at this point. 49:01 对。 Yeah. 49:01 它被贴到所有东西上。我真希望我们有个别的词，就单纯表示“AI 在干活”。 Like, it's just slapped onto everything. I wish it, we had a different word that just meant AI doing stuff. 49:07 “凭感觉编程”这个词感觉也不对。 Vibe coding feels wrong, too. 49:09 确实不对，因为凭感觉编程在我闻来，就跟 2000 年代初的脚本小子一个味儿。 It does, because vibe coding to me smells exactly like script kiddies did in the early 2000s. 49:15 嗯哼。 Mm-hmm. 49:15 一帮人拿着刚从网上下载的 PHP 脚本就用，压根儿不懂里面是什么。 People applying PHP scripts they just downloaded offline they don't understand anything of. 49:21 嗯哼。 Mm-hmm. 49:21 不过，我的意思是，从某种意义上说这也没错——很多凭感觉编程的东西，包括我自己凭感觉编程做的项目，我刚才说过，我有好几个。我用 C++ 写过东西，也用 Rust 写过东西。 And while, I mean, that's true in the sense that much of the vibe coding, including my own vibe coded projects, and as I just mentioned, I have several. I've written things in C++. I've written things in Rust. 49:32 嗯哼。 Mm-hmm. 49:33 Rust。我对 Rust 是深恶痛绝。要我看 Rust 代码，就像往我眼睛里泼硫酸。 Rust. I hate Rust with a passion. Rust to me is like pouring acid in my eyes when I have to look at the code. 49:42 等等，等等，为什么，为什么，你为什么这么恨 Rust？ Wait, wait, wait, why, why, why, why do you hate Rust? 49:43 在我看来，它是最丑的编程语言—— It is, in my opinion, the ugliest programming language- 49:46 哦。 Oh 49:46 ……大概是过去 40 年里发明出来的最丑的。 ... that has been invented in probably the last 40 years. 49:48 所以它是 Rails 和 Ruby 的反面。 So it's the opposite of Rails and Ruby. 49:50 它就是、就是反面。但是—— It's the, it's the opposite. But- 49:53 但你写过它。 You've written it 49:53 ……我是用 Rust 写过应用，因为 Rust 的产出确实很惊艳。内存安全、它是系统级语言、它效率极高，等等，这些都不可思议。 ... I've written applications in Rust because the output of Rust is actually amazing. The memory safe- Safety, the fact that it's the system language, the fact that it's highly efficient and, and so forth is incredible. 50:06 所以 Rust 对我来说可以同时是：为人类使用而设计过的最令人反感的编程语言，以及智能体工程的绝佳平台。 So Rust to me can both be the most repugnant programming language ever devised for human consumption and a wonderful platform for agentic engineering. 50:19 嗯哼。 Mm-hmm. 50:19 这两个事实可以毫不费力地在我脑子里共存。 These two truths can coexist easily in my head. 50:24 你觉得会不会有那么一天，我们可以直接把智能体工程就叫作编程？我们能不能直接改变“编程”的含义？因为再过几个月，还有谁会真的用老派的方式编程呢？ Do you think that there can be a point at which we can just call agentic engineering programming? Can we just change what programming means? 'Cause who's actually going to be programming the old-school way in, in a few months? 50:36 我觉得我们不应该复用这个词，因为对我来说，编程意味着理解某些基本构件：循环、条件、变量。 I don't think we should reuse the term because for me, programming implies an understanding of certain primitives, loops, conditions, variables. 50:48 编程语言的所有这些构造才构成了编程本身，而不是“做出一个程序”。因为在智能体时代之前你也可以说，嗯，你们的 CEO 也在“编程”。 All the constructs of programming languages are what constitutes programming itself, not the creation of programs. Because you could say in the pre-agentic era, well, your CEO is programming. 51:00 他雇了一帮程序员，他、他告诉他们该做什么，然后就出来一个他能卖的软件。 He's hiring a bunch of programmers, he's, he's telling them what to do, and out comes software that he can sell. 51:05 嗯哼。 Mm-hmm. 51:05 呃，我觉得大多数人不会把那位 CEO 叫作程序员，我也不觉得我们应该把凭感觉编程的人叫作程序员。而凭感觉编程，如果我们在这里下个定义，就是你让一个智能体替你构建软件。 Eh, I don't think most people would call that CEO a programmer, and I don't think we should call the, the vibe coder a programmer either. And vibe coding, if we define it here, is you tell an agent to build software for you. 51:17 你不看实现。对我来说，这就是凭感觉编程与编程——或者说智能体加速开发——之间的分界。 You do not look at the implementation. That to me is what separates vibe coding from programming or let's say agent-accelerated development. 51:26 但是，让我反驳一下，你不觉得一个程序员——老派程序员，比如你——做智能体工程，跟一个非程序员做智能体工程，是两回事吗？ But don't you think, so to push back a little bit, don't you think a programmer, old-school programmer, you, doing agentic engineering is a different kind of agentic engineering than a non-programmer doing agentic engineering? 51:39 我的意思是，如果你知道 for 循环是什么，知道函数式编程是什么，知道软件工程的一些基本好原则，那你做基于自然语言的智能体工程的方式会不一样，会更系统化， What I mean is if you know what a for loop is, if you know what functional programming is, if you know some of the basic good principles of software engineering, the way you do natural language-based agentic engineering will be different and more systematic, 51:52 而且你实际能构建的东西的规模和种类，会比凭感觉编程的人大得多。 and this, the, the, the scale and the variety of things you can actually build is, um, much bigger than a vibe coder. 52:02 那我就顺着这个观点再往前推一点，纯粹为了把话说透。 I'm just gonna extend the pro case here a little bit just for the sake of the argument. 52:05 嗯。 Right. 52:07 我实际上认为，有一段时间，懂这么多编程知识反而成了我的劣势，因为我总是指示智能体按我规定的方式去做事， I actually think for a while it was to my deficit to know as much as I know about programming because I was instructing the agents to do things as I prescribed them to do, 52:19 而它们很擅长照做。在第一个智能体阶段——咱们别叫它时代，它也就持续了三、三个月—— and they were very good at that. And in the first agentic moment, let's not call it an era, it was something that lasted three, three months- 52:31 对。 Yeah 52:31 ……阶段。在第一个智能体阶段，那样做感觉非常高效。 ... moment. In the first agentic moment, that felt very productive. 52:36 我通过告诉智能体“按我本来会用的方式去做”获得了更高的产出，我把自己作为程序员的全部经验都用在了这上面。 I could get more productivity out of telling the agents what to do in the way I would've done it, and I used all of my experience as a programmer to do just that. 52:45 嗯哼。 Mm-hmm. 52:47 然后下一个阶段我入场晚了一点。那个阶段让我、让任何人都可以只描述结果—— Then I was a little late on the next moment, and the next moment allowed me, allowed anyone, to describe outcomes- 52:57 嗯。 Mm 52:57 ……向智能体描述问题，然后得到比由程序员规定路径更好的解决方案。 ... to describe problems to the agents and get better solutions than if you had a programmer prescribe the path. 53:05 你看，好吧，咱们这争论挺有意思。呃，所以你是说，存在某些问题，在智能体工程上程序员反而比非程序员做得差？ See, okay, so we're having fun arguing here. Um, so are you suggesting that there's problems for which programmers are worse than non-programmers at agentic engineering? 53:16 百分之百。我为什么这么说，咱们在这儿说清楚。我这么说的原因是：很多程序员并不是很好的产品经理。 100%. And the reason I say that, let's define it here. The reason I say that is there's a lot of programmers who are not very good product managers. 53:27 软件就是产品管理。它该做什么？为谁做？怎么做？看起来该是什么样？我们的优先级是什么？先做什么？第一版包含什么？ Software is product management. What should it do? Who should it do it for? How should it do it? How should it look? What's our priorities? What do we start with first? What does version one include? 53:39 所有这些能力并不会轻易、也不会均匀地分布在所有程序员身上。 All of those skills are not easily or equally distributed across all programmers. 53:46 而在智能体时代，你让智能体、让 AI 去做实现，你需要的正是这些能力。而我恰好两边都占，这就是为什么我热爱这个时代。 And in the agentic era where you are letting a agent, an AI, do the implementation, these are the skills you need. And I happen to sit in both camps, and this is why I love this moment. 53:59 我仍然有一部分活在旧世界里——在那些我觉得看完整实现仍然能带来价值的领域，我会去看——同时我也活在未来。 I still get to live a little bit in the old world where I look at the full implementation because there are domains and areas where I feel that still brings value, and then I also live in the future. 54:10 就像我说的，我创建 Amara Write 的时候，一行 C++ 代码都没看过。我其实是特意跟自己约定：我要把它当作一个实验。这是个百分之百的黑盒。 And as I said, when I created Amara Write, I've not looked at a single line of that C++. I actually made it a point to myself that I was gonna treat it as an experiment. This is 100% a black box. 54:22 我会把自己当成任何一个对自己的数字打字机该怎么工作有想法的普通用户，因此我觉得这是个公平的实验——就好像任何一个对软件该怎么做、该怎么工作有想法的写作者一样。 I will treat it as though I was any other user who had opinions about how their digital typewriter should work, and therefore felt like it was a fair experiment as if a writer of any other sort who had opinions about how software should be done and how it should work. 54:39 就像你提到的那些视频剪辑师，他们清楚地知道自己不喜欢 Adobe Premiere 的哪些做法、它怎样才能更好，但没有能力去改进它—— As you mentioned, video editors who know exactly where they don't like what Adobe Premiere's doing and how it could be better, but don't have the capacities to improve it- 54:49 嗯哼。 Mm-hmm 54:49 ……现在他们能做到了。 ... how that could be. 54:50 不过反驳一下：我觉得程序员——优秀的程序员——擅长的是系统化设计，也就是用系统的视角思考—— The pushback though, I think what programmers do well, good programmers do well, is systematic design, like think in terms of systems- 54:59 没错。 Correct 54:59 ……以及严谨。 ... and rigor. 55:00 对。 Yes. 55:00 而且我认为，在未来很长一段时间里，即使是用自然语言控制智能体系统，也仍然需要某种、某种严谨性。 And I think still, and probably for a long time to come, even natural language-based control of agentic systems will require a, a, a certain kind of rigor. 55:11 不是那种过度结构化的严谨，而是要非常具体地讲清楚，呃，系统的目标、如何做验证、做什么样的，呃，安全测试。 Not overly structured rigor, but really explain very specifically w- uh, the goals of the system, how to do the verification, what kind of, uh, security testing. 55:22 你会想把所有这些事情都交代清楚。不过再反驳一下这一点：从长远看，系统应该足够聪明—— You wanna do all those kinds of things. Although then to push back on that, the system will, should be smart longer term- 55:29 六个月前我也会这么说。 I would've said the same thing six months ago. 55:30 是啊，所以这已经变了。 Yeah, so this has changed. 55:30 我觉得六个月前这话确实成立。但自那以来我的发现——尤其是最近这几周用智能体加速开发 Omarchy 的过程中——是：多数情况下， And I think the same thing would've been true six months ago. What I have found since, especially even just the last few weeks of the agentically accelerated development of Mach-E, is that more often than not, 55:42 我得谦卑地承认，智能体才是最懂的。 I have the humility to recognize that the agent knows best. 55:49 你大可以直接说：“确保它是安全的。” You can legitimately say, "Make sure it's secure." 55:52 没错。而且它对这意味着什么懂得——懂得比你多。哎，你笑了—— Correct. And it would know- ... more about what that entails. Now, you're laughing- 55:59 这世界太疯狂了。 What a crazy world 55:59 ……但这是真的。而且事实上—— ... but it's true. And in fact- 56:01 我的天。 Oh my God 56:02 ……一个很好的类比是 AGENTS.md/CLAUDE.md 文件。有段时间特别流行——大家都热衷于对它做微优化。“哦，你要告诉你的智能体这个—— ... a good parallel to this is the Agents MD/Claude MD files. There was a hot moment where- It was all the rage about micro-optimizing that. "Oh, you tell your agent this- 56:14 对。 Right 56:14 ……你要告诉你的智能体那个”，最后你整出一个巨大的、给智能体看的指令文件。 ... you tell your agent that," and you ended up with this huge file of instructions for your agent. 56:20 呃，做 Claude Code 的 Boris 最近在一次采访里谈到 Opus 5 时提到一点：他们为 Opus 5 搭载的系统提示词， One of the things that, uh, Boris working on Claude Code shared in an interview recently about Opus 5 was that the system prompt that they ship for Opus 5, 56:30 估计 Fable 也一样——缩减了 80%，因为智能体不仅需要的人类指令少得多，而且实际上还会被过度事无巨细的人类指令拖后腿。 and presumably also Fable, shrunk by 80% because the agent not only needed far less human instruction, it was actually being damaged by overly prescriptive humans. 56:45 任何一个遇见过尖头发老板的程序员都知道那是什么滋味。 And any programmer who's had a pointy-haired boss knows exactly what that is like. 56:51 老板走进房间，啥也不懂，就开始告诉你该怎么编程、怎么写代码。你怎么办？你只能生闷气。 When the boss walks into the room, doesn't know anything, starts telling you how to program, how to code. What do you do? You sulk. 57:01 如果你被强制要求做违背自己专业判断的事，你写出来的代码就更烂。 You write shittier code if you're mandated to do things that are against your better judgment. 57:05 对。 Yeah. 57:06 智能体凭什么就不一样呢？ Why would an agent not be the same? 57:08 其实，我也反驳一下自己：我确实发现我目前正在磨练的真正技能——呃，顺着这个思路，大概也是大家都在磨练的——是不要过度规定细节，是别挡系统的道。 Actually, to push back against myself, I do find the actual skill I'm currently developing, which I, b- uh, along those lines is probably what everybody's trying to develop, is not to over specify things. Is to get out of the system's way. 57:20 对。 Right. 57:20 是说清楚到刚好能提供一种高层的愿景和引导就够了。 Is to specify enough to provide a kinda high-level vision and guidance. 57:25 其实更妙的是：你连那个都不需要。现代软件开发最根本的洞见来自敏捷软件开发运动。 It's even better because you don't even need that. The fundamental insight of modern software development came from the agile software development movement. 57:34 在 90 年代末、2000 年代初，一群聪明、诚实又勇敢的人聚在一起说：“过去四五十年来我们做软件开发的方式行不通，以后也行不通。 In the late '90s, early 2000s, a group of smart, honest, and brave people came together and said, "The way we've been trying to do software development for the last 40 to 50 years have not worked, will not work. 57:48 我们一直试图在一开始就把软件该是什么样规定清楚。 We've been trying to specify upfront what software should look like. 57:53 我们以为可以问问人们想要什么，把一切都写下来，把我们的系统思维和严谨性用在一份规格说明书上，再把这份说明书交给一群程序员，让他们原封不动地实现出来， We thought that we could ask humans what they wanted, write it all down, apply our systems thinking and our rigor to a spec sheet, hand that spec sheet over to a group of programmers who would then implement just that, 58:04 然后大家就都会满意。”结果呢？没人满意，因为人们在拿到东西之前根本不知道自己想要什么。你不上手玩一玩，就不知道一个程序该做什么。 and everyone would be happy." Well, guess what? No one was happy because no one knows what they want until they receive it. You don't know what a program should do until you play with it. 58:17 所以在智能体时代，你应该抵制住一开始就说得太细的诱惑。尽量模糊一点，先把某个东西变出来，然后去跟这个东西互动。 So in the agentic age, you should resist the temptation to be overly specific upfront. Be as vague as you can to manifest something, then interact with the something. 58:28 嗯哼。 Mm-hmm. 58:29 做出好软件的方法是：先写一点点软件，然后试着去用它。正是在使用软件的过程中，你才会发现自己真正想要什么。 The way you arrive at good software is you write a little bit of software, and then you try to use it. It is in the process of using software that you discover what you really want. 58:38 嗯哼。 Mm-hmm. 58:38 也是你发现什么重要、什么不重要的过程。而智能体的神奇之处在于，它们让你能跌跌撞撞地撞进“意识到自己想要什么”这种体验——因为你并不知道自己想要什么。 It's where you discover what's important and what's not important, and agents are incredible at allowing you to bumble into this experience of realizing what you want because you don't know what you want. 58:51 嗯哼。 Mm-hmm. 58:52 我、我其实做过和你说的类似的事：实现，呃，不同的设计、完全不同的实现，然后给自己做 HTML 页面，呃，PHP 页面，在上面给我更喜欢的方案投票之类的。 And I- I've actually done s- so something similar to what you suggested, which is implement, uh, different designs and totally different implementations, and then I create HTML pages for myself, uh, PHP pages, where I vote on what I like more and so on. 59:06 对。 Yes. 59:06 它会这样不断迭代，感觉很好也很有意思。跟智能体系统交互——如果你给自己做一个好用的界面——可以超级好玩，而且你专注的是设计部分，是设计里最有趣的部分。 And it does this iterative thing, and it's very nice and fun. The interaction with the, the agentic system, if you create a nice interface for yourself, it can be super fun, and you're focusing on the design part and the most fun parts of the design part. 59:17 没错，而这正是人类真正擅长的。 Correct, and this is what humans are really good at. 59:20 对，品味。 Yeah, taste. 59:21 差异化评估。你给我三个选项，我挑一个。不过如果你给人类 22 个选项，他们就直接掉下悬崖了。这就是选择的悖论。 Differential evaluation. You give me three options, I pick one of them. Now, humans actually fall off a cliff if you give them 22 options. That's the paradox of choice. 59:31 但你给他们三个选项，他们一瞬间就能告诉你他们喜欢哪个。 But you give them three options, they very quickly within a split second will tell you what they like. 59:36 有很多方法都利用了这一点：人类会做出这种瞬间判断，而这种判断其实是先于理智的，对吧？ There are all these ways that exploit the fact that humans will make these snap judgments that are actually pre-intellectual, right? 59:44 它们从直觉里冒出来，然后才传到、传到大脑，大脑再试图为直觉说过的话找理由。 Like, they're coming from the gut, and then they arrive up in the, in the brain, and the brain tries to rationalize why the gut said what it said. 59:53 如果你愿意放下一些那种理智上的预先合理化，干脆让直觉来掌舵，智能体时代就是一种启示。 And if you're willing to let go of some of that intellectual pre- processing rationalization and simply let your gut drive, the agentic age is a revelation. 60:06 所以问你这个问题特别有意思，因为你过去很长时间以来一直很有名地谈论，比如，对优美的 Rails、Ruby 代码精雕细琢， So this is a fascinating thing to ask you because you have sort of famously for a long time talked about, like, detailed chiseling of beautiful Rails Ruby code, 60:16 而现在你在几个月之内就转向了不再那样做。那你，呃，现在的工作方式是什么样的？ and now you have switched in a matter of months to not doing that. So, like, what is your, uh, current setup look like? 60:27 因为我猜你仍然在追求美、追求雕琢、追求手艺，只是你在换一种媒介做手艺。 'Cause you're, I assume, are still looking for the beauty, for the chiseling, for the crafting, but you're just crafting in a different medium. 60:35 而且在不同的抽象层次上。当我在处理 Ruby 代码时——我们有很多 Ruby 代码，因为它贯穿我们整个业务——当我让智能体去改那些代码时，我仍然会抠细节。 And at different levels of abstraction. When I'm working in Ruby code, and we have a lot of Ruby code because it's in our entire business, and I'm asking agents to make changes to that code, I still sweat the details. 60:46 嗯哼。 Mm-hmm. 60:48 我逐渐意识到的是，这种抠细节带来的经济回报正在迅速缩水。 What I'm coming to realize is that the economic payoff of that sweat is diminishing rapidly. 60:58 我之所以 25 年来如此审慎地抠每一行代码，是因为我知道，保持架构一致且可塑的回报是：软件可以快速变化和演进—— The reason why I, for 25 years, was sweating every line of code so judiciously was because I knew the payoff of keeping an architecture coherent and malleable was software that could change and evolve- 61:16 嗯。 Mm 61:16 ……靠一个小团队、不用高昂的成本，也不会因为改了一处就在另一处引入一堆 bug。 ... quickly with a small team, and not exorbitant cost, and not introducing a bunch of bugs when you change one thing over the other. 61:24 这就是“为什么要写优美的代码”的核心经济论据：优美的代码更容易理解、更简单、更可塑。 That was the driving economic argument for why you should write beautiful code, because beautiful code is easier to understand. It is simpler. It is more malleable. 61:36 但那个前提是“由人来修改代码”。这件事在今天还在多大程度上成立，我觉得是个开放的问题。 That was premised on humans doing the modifications. I think it is an open question to which degree this still matters. 61:44 不过，它确实还重要。我这么说的原因——至少眼下——是 token 仍然是稀缺的。在当前这个时间点，我们所有人都受 token 限制。 Now, it does matter, and the reason I say that, at least for the moment, is that tokens are still scarce. At this moment in time, we are all token limited. 61:58 好吧，不是所有人，但任何没有无限预算的人都受 token 限制。 Well, not all, but anyone who, who doesn't have endless budgets are token limited. 62:03 所以，写出让智能体更容易处理和演进的系统——不用每次都重新学习整个上下文——回报是很大的。 So therefore, there is great payoff to writing systems that agents have an easier time dealing with and evolving without having to relearn the entire context. 62:13 就像人一样，如果它们能在不破坏架构的情况下迭代和修改代码库，那么下一次修改的成本就能和上一次一样低。 Just like humans, if they can make iterations and changes to the code base without wrecking the architecture, they can make the next change just as cheaply as the last one. 62:24 这就是经典的“大泥球”问题：你最后得到一个烂成一团的系统—— And this is the classic ball of mud where you end up with a system that's a ball of mud- 62:31 嗯哼。 Mm-hmm 62:31 ……因为它的拼凑方式让各部分之间毫无关联，整个就是、就是一团乱麻，对吧？ ... because it's just put together in a way where nothing is connected, and it's just a, a real mess, right? 62:37 我在智能体身上见过这种情况，也在我们自己的代码库里见过：第一个 PR 质量就马马虎虎，然后你在上面再叠一个 PR，接着再叠五个，结果就不怎么样了，对吧？ I've seen that with agents, and I've seen it in our own code bases where, again, the first PR is, like, mediocre of quality, and then if you add another PR on top of that, and then five more down the line, it's not very good. Right? 62:50 所以这方面仍然有回报，但那个回报是建立在我们当前这个时点上的。而我现在已经能够……顺便说一句，“AI 精神病”真正发作的地方就在这里：当你试图推演九个月后会是什么样子的时候。 So there's still a payoff to that, but that payoff is premised on our current moment. I am now able to... And this is, by the way, where the AI psychosis really comes in, when you try to extrapolate what nine months from now it's gonna look like. 63:02 是啊。 Yeah. 63:02 推演两年后会是什么样。不过作为一个思想实验：我用过的第一台电脑是 Commodore 64。 What two years from now it's gonna look like. But as an intellectual experiment, I think the first computer I used was a Commodore 64. 63:10 我也是。很巧。 Me too. Cool. 63:11 它有一颗 1 兆赫兹的 CPU 和 64K 内存。 It had one megahertz CPU and 64K of memory. 63:15 嗯哼。 Mm-hmm. 63:16 每个为那台机器写软件的人都把某些约束刻进了骨子里。好吧，是所有约束。那台机器就是纯粹由约束构成的。 Everyone who wrote software for that machine internalized certain constraints. Well, all the constraints. That machine was nothing but constraints. 63:25 如果你把当年的程序员放进时光机，把他或她传送到 2026 年，给他们一台今天的现代电脑，然后说：“来，给我写个软件”， If you put that programmer in a time machine and teleported them or her to 2026 and gave them a modern computer today and said, "Now write me a piece of software", 63:35 他们会迷茫一阵子，因为他们所有的技术和经验法则全都不适用了。或者说不是“不对”，因为高效的软件仍然是美的。 they'd be lost for a little bit because all their techniques and all their heuristics would simply be wrong. Or not wrong, because efficient software is still beautiful. 63:48 只是他们能创造的价值已经过时了。比如，要让一台 1 兆赫兹的电脑跑出一个电子游戏，你需要做的优化量级，跟今天需要做的完全不同。 They would be out of date with the value that they could create. Like, the amount of optimization you have to apply for a one megahertz computer to produce video game is just very different from what you have to apply today. 63:59 想象一下当年那个在 Commodore 上编程的人。回想一下，呃，编程当时是什么感觉、意味着什么：资源高度受限—— Just imagine that guy that was programming back then on a Commodore. Just rem- remember all those, like, what programming felt like and what it meant, highly constrained resources- 64:10 对。 Yes 64:10 ……一切都那么慢。 ... how slow everything is. 64:11 在很多方面那也是美的。很美。 Beautiful in many ways. Beautiful. 64:12 那就是精雕细琢。但是—— It's chiseling. But- 64:14 它、它—— It, it- 64:15 ……想象一下那个人来到现在。 ... imagine that person coming now. 64:17 不过这也挺浪漫的。咱们也把这一点想进去。有这么个现象：比如每当我们哀叹现代汽车文化的时候，对吧？ It's romantic too, though. Let's also imagine that. This is one of the things. Whenever we bemoan modern car culture, for example, right? 64:26 然后我们会想：“哦，以前大家都坐马车的日子是不是更好？”不，并不好。你知道当年用马匹当交通工具的纽约是什么味儿吗？ And then we think like, "Oh, wasn't it better when we all had the horse-drawn carriages and so on?" No, it wasn't. Do you know what New York smelled like when we used horses for transportation? 64:35 嗯。 Yeah. 64:35 那就是个露天下水道。马粪拉得到处都是。 It was an open sewer. They just shat everywhere. 64:38 是，但你玩过 Red Dead Redemption 吗？你知道那些牛仔有多酷—— Yeah, but have you played Red Dead Redemption? You know how cool those cowboys were- 64:42 这正是那种—— That is the way- 64:42 ……骑在马上的样子？ ... on the horse? 64:43 ……体验怀旧过去的方式：你不会接收到那些，呃，毛茸茸的，呃，感官刺激。 ... to experience the nostalgic past, when you don't get the, uh, all the furry, uh, impulses and senses. 64:50 所以基本上，老派的手写编程有点像我们浪漫化、拍成电影的那种牛仔文化。 So basically, programming, old school programming by hand is kinda like the cowboy culture that we romanticize and make movies about. 64:57 对手写代码的浪漫化已经开始了，我自己也有一点，因为那确实是个非常浪漫的时代。 It's already happening, the romanticization of handwritten code, and I have some of it because it was a very romantic era. 65:05 我很庆幸自己亲历了 20 年“手写代码具有经济价值”的岁月。那是、那是一段好时光。 I'm grateful to have had been alive for 20 years of economically valuable handwritten code. That was a, that was a good time. 65:16 一个手写优美代码也有价值的时代。 Where handwritten beautiful code was also valuable. 65:20 没错。因为手写优美代码今天依然存在，和昨天一样多。 Correct. Because handwritten beautiful code exists as much today as it did yesterday. 65:26 现在仍有人主动为 Commodore 64、为 Sega Mega Drive 或其他复古主机写新游戏。 There are people who willfully write new video games for the Commodore 64 or for the Sega Mega Drive or for other vintage consoles. 65:35 事实上，我最近最喜欢的游戏之一是，呃，Mod Retro 的。你知道，呃，Palmer Luckey 的副业吗？就是复刻老式主机的那个。 In fact, one of my favorite video games of late is, uh, Mod Retros. Do you know, uh, Palmer Luckey's side geschift, which is recreating old consoles? 65:47 他复刻了 Game Boy。 So he's recreated the Game Boy. 65:49 不错。 Nice. 65:51 特别棒。你去查一下。真的、真的很酷。 It's amazing. Look it up. It's, it's really cool. 65:54 但比“他复刻了这个硬件、保留了原版的厚重感，又用，呃，现代屏幕之类的东西把它带进当代”更酷的是：他们现在还出了 Nintendo 64 的复刻。太不可思议了。 But what's even cooler than the fact that he recreated this hardware and the original chunkiness, but brought it up to date with, um, sort of modern screens and so on, now they just put out the Nintendo 64 too. Incredible. 66:06 超级酷。对细节的疯狂关注、输入延迟，所有这些都考虑到了。不过还有一个也很酷的——那个，我把这个挥一下。就那边那个，你把镜头凑近点，这样你就能看到我在说什么了。 Super cool. Wild attention to detail, input lag, and all the other things. But what's also very cool, that one, I have the wave. That one right there, and hover over it, because then you get to see what I'm talking about. 66:19 那是 Chromatic—— That is the Chromatic- 66:21 啊。 Ah 66:21 ……的俄罗斯方块重新实现版。 ... Tetris re-implementation. 66:23 哇。 Wow. 66:23 我在原版 Game Boy 上玩过超多俄罗斯方块。 I played a lot of Tetris on the original Game Boy. 66:26 嗯。 Yeah. 66:26 它大概是我有史以来最喜欢的前三款游戏之一。 Probably one of my top three favorite games of all time. 66:29 嗯。 Yeah. 66:30 他们重写了它，只改了一处：如果你按“上”，方块会直接砸到底。 And they rewrote it with one change. If you hit up, the brick slams. 66:38 嗯哼。 Mm-hmm. 66:39 这一改让游戏节奏快了大约 400%。这是个不可思议的游戏，而且它是一个新的 Game Boy 游戏。 And that speeds up the game by about 400%. It is an incredible game, and it's a new Game Boy game. 66:47 嗯哼。 Mm-hmm. 66:47 我记得 Game Boy 是 89 年还是 88 年出的。那是非常老的机器了。现在还有人在写“新的老游戏”，这太棒了。 I think the Game Boy came out in '89 or '88. It's a very old machine. The fact that there's still people writing new old games is awesome. 66:57 他们这么做的部分原因——在这里他们做是因为想要俄罗斯方块——但也有人纯粹是出于对约束的热爱，出于对“写那种东西”这种浪漫情怀的热爱。 And the reason they do it in part, now here they're doing it because they want Tetris, but there are people who do it just for the love of the constraints, for the love of the romantic notion of, of writing that. 67:07 就像现在还有人骑马出行，不是为了快点到达目的地，而是因为他们喜欢这种交通方式：胯下是一个实实在在的生物，推着你向前走。 Just like there are people still riding around on horses, not to get places fast, but because they like the mode of transportation where it's a literal biological being below you propelling you forward. 67:21 另外我们还应该说：那些优美的代码是，呃，绝佳的训练数据。所以从某种意义上说，我们几十年间构建的美—— And also, we should say that that beautiful code is, uh, great training data. So in some sense, the beauty that we've built over a few decades- 67:30 对。 Yes 67:30 ……延续了下来。 ... continues. 67:31 我们孕育了——……孕育了这一刻。 We gave birth- ... to this moment. 67:35 是啊。 Yeah. 67:35 这是何等的幸运。我、我个人对此感受很深，因为我写过的几乎所有代码都是公开的代码。我职业生涯的绝大部分都投入在了写开源代码上。 What a privilege. And I, I feel that very personally because I, almost all of the code I've ever written is public code. The vast majority of my career has been dedicated to writing open source code. 67:47 所以，其中一些代码、一些优美的代码行就在训练集里。事实上，我听说有人写 Ruby 代码时会这么做： And therefore, some of that code, some of those beautiful lines are in the training set. And in fact, I've heard people do this when they write Ruby code. 67:57 他们跟智能体说：“按 DHH 的风格来写。” They ask the agent, "Write it like DHH would." 68:00 嗯哼。 Mm-hmm. 68:01 而他们对结果很满意，这确实让我心里暖了一下：我以自己微不足道的一份力，参与孕育了这一刻。还有所有贡献过开源代码、或者任何代码的人——只要智能体在过去这些年里能接触到， And they're pleased with the result, and that does warm my heart a little bit, that I helped give birth to this moment in my tiny little small part, and anyone else who contributed open source code or any code at all that the agents have had access to over the last, 68:12 好吧，一直以来—— well, entire time- 68:16 嗯。 Mm 68:16 ……编程作为一门学科存在以来的所有代码，如今都成了这些新的智能体主宰的一部分。 ... programming has been around as a discipline, is now part of these new agent overlords. 68:22 那么你、你会不会有一点伤感：手写优美代码正变得越来越没用了？ So you, do you don't have a little bit of sadness that writing beautiful code by hand is now less and less useful? 68:29 如果我明天得出去找工作，很可能没法再靠“手写一行行代码”这门手艺挣钱了——我为此伤感吗？不。 Am I sad that if I had to get a job tomorrow, it's unlikely that I could apply my skills in such a way that I would be paid to write these manual lines of code? No. 68:42 我干了 25 年了。我想看点新东西——尤其是如果我们都要长生不老的话（我希望不会，也许以后再说吧）。我已经做过了，做够了，挺好的。 I did it for 25 years. Like, I wanna see something new here, especially if we're all gonna live forever, which I hope we don't. Maybe we get to that. But like, I've done that. That, that's enough. That's fine. 68:52 我对此的怀念，并不比我对“不用扛着锄头下地、不用站在流水线上”的怀念更多。 I don't feel any more nostalgic about that than I do about the fact that I don't have to spend my time in the field with a hoe or at the assembly line. 69:01 等等，等一下。你可是世界上最擅长手写精美代码的程序员之一，是极有品味的美学鉴赏家。 Well, hold on a second. You are one of the best programmers in the world at writing handcrafted beautiful code, appreciators of beauty with great taste. 69:11 没错。 Correct. 69:12 而这已经被取代了。 That has been replaced. 69:13 我谦虚地说一句。 I say humbly. 69:14 对，那已经被取代了。而你现在，呃，就是在快速进化，你能够——你非常了不起的品质之一就是：你能迅速改变想法、快速进化。 Yeah. That has been replaced. Like you had now, you're just fast evolving and you're able to, one of your, uh, really great qualities is you're able to change your mind and evolve very quickly. 69:27 但某种意义上，你得亲手杀死过去的那个自己。 But like, you had to kill that other person. 69:30 我感激把我带到此刻的每一个瞬间。因此我也一直在努力——也许这确实需要练习，我承认，也许我并不一直都能做到这样。 I am grateful for every moment that brought me to where I am right now. And therefore, I have also tried, and maybe this did take some practice. I will concede that. Maybe I wasn't always like this. 69:42 但斯多葛哲学、amor fati——爱你的命运——是一种极其解脱的活法。 But stoic philosophy, amor fati, loving your fate, is an incredibly liberating way to live. 69:52 你可以攻击这一点说：“好吧，那是因为你站着说话不腰疼。”没错，但我在没有这些所谓的“特权”时，秉持的也是同样的方向。 And you could attack that and say, "Well, that's a point of privilege." True, but also I had the same Directional point when I didn't have all this, quote-unquote, "privilege." 70:05 所以我认为有一种活法、一种与世界相处的方式：接受你能改变的和不能改变的事情，然后爱上这一切。 So I think there's a way to live and interact with the world where you accept the things you can change and the things you cannot change, and fall in love with all of it. 70:14 嗯哼。 Mm-hmm. 70:15 爱上这个事实：今天的世界相比去年，已经在科技行业里向前跳了二十年。 Fall in love with the fact that the world today has jumped two decades forward into technology industry from where it was last year. 70:24 那么我们能聊聊程序员们普遍感受到的焦虑吗？呃，他们也在经历同样的转型。 So can we talk about the general anxiety that programmers feel, uh, going through the same transformation? 70:30 他们可能上了大学，呃，主修计算机科学，梦想着当程序员、做东西、拿高薪，而现在一切都在变， They've maybe gone to university, uh, majored in computer science, dreamed of being programmers and building stuff, high salary, and now everything's changing, 70:41 他们对于“我的人生该怎么办”有很深的焦虑。你能共情这种焦虑吗？他们到底该怎么办？ and there's deep anxiety about what do I do when... with my life. Can you empathize with that, and what the hell are they supposed to do? 70:51 非常能。因为我自己虽然有这种天性，但我清楚地知道它并不是平均分配的，不是每个人都处在我的位置上、都对未来抱有这样的乐观。 Hugely. Because just because I have this disposition, I'm keenly aware that that's not evenly distributed, that not everyone sits with this position and this optimism for the future. 71:03 不过我确实想把几件事分开来说。 But I do think I wanna separate things a little bit. 71:08 如果你对编程的热爱只在于“把正确的逻辑构件拼在一起、产出别人让你产出的东西”这种机械的部分，那你确实会很难适应新现实。 I think you're gonna have a hard time coping with the new reality if the only thing you loved about programming was the mechanical bits of putting the right logical constructs together to produce something other people told you to produce. 71:28 因为那个机械过程正受到威胁。但如果你像你刚才说的那样，对“创造东西”感到兴奋，我觉得你根本没有受到威胁。 Because that mechanical process is under threat. If you are, as you just mentioned, excited about building things, I don't think you're under threat at all. 71:40 事实上，有个很有力的论点是：我们需要比现在多得多的创造者。而如果你看就业数据，情况很模糊。 In fact, I think there's a great argument for us needing far more builders than what we have now. And if you look at the employment stats, it's fuzzy. 71:53 会发生什么完全说不准。 It's not clear what's gonna happen at all. 71:55 有些数据实际上显示岗位空缺在增加，因为 AI 的到来大幅拉低了软件的价格，人们想要多得多的软件。 Some stats actually show an increase in openings because the advent of AI is so dramatically lowering the price of programs that people want a lot more programs. 72:07 这就是经典的杰文斯悖论：某样东西的价格下降，对它的需求就会增加。ATM 也是个例子。 This is the classic, the Jevons paradox, that says when the price of something goes down, there's gonna be more demand for it. And this is the example of the ATMs too. 72:17 ATM 刚出现的时候，很多银行柜员非常担心自己的工作，因为突然有了一台能从人们账户里吐钱的机器。 When the ATM originally came, a lot of bank tellers were very afraid for their job because suddenly there was a machine that could dispense money from people's account. 72:27 但 ATM 实际做的是降低了网点的成本，银行一下子开得起多得多的网点，结果我们的银行柜员反而比以前更多了。当然，这些都不是必然保证的。 Well, what the ATMs did was lower the price of a branch, and suddenly banks could afford to open a lot more branches, and we ended up with more bank tellers than we did before. Now, none of this is guaranteed. 72:39 也确实有事情真正改变的时刻。 There are also moments where things do change. 72:41 直到 19 世纪末，绝大多数人都还在田里劳作，然后突然有了机械化收割工具，我们就不再需要扛着锄头、赶着牛的人力了。 The vast majority of people worked in the fields up until the late 1800s, and then suddenly we got mechanized harvesting tools, and we did not need human labor with a hoe and an ox out there. 72:56 我们有了，呃，机械工具替我们做这些事。在那个时刻，有没有人感觉工作和生计受到了威胁？有。这种受威胁的感觉合理吗？ We had, uh, mechanical tools to do that for us. In that moment, are there people who were threatened that their job and their livelihood was at stake? Yes. Was that a reasonable thing to feel threatened about? 73:09 合理。当年砸织布机的卢德分子—— Yes. The Luddites smashing the weaving machines- 73:13 嗯哼。 Mm-hmm 73:13 ……在英格兰也是同样的事。 ... in, in England had the same thing. 73:16 而且我觉得这个类比可能比种田的人更贴切：那些人其实是高技能的专业人士，做着自己喜欢的工作，条件还相当不错，不是整天在太阳底下卖苦力。 And I think maybe that's the better parallel than people working the fields because these were actually highly skilled professionals doing a job that they liked on rather favorable conditions, not slaving outside in the sun all day. 73:28 但生活在今天的我们，还愿意让衣服是一种极度稀缺的资源吗？比如，如果这件 T 恤要 400 美元呢？ But where we live now, would we still like for clothing to be this heavily constrained resource? Like, what if this T-shirt was $400? 73:40 我是说，好吧，那我就只有两件了。不过我的衣柜看起来大概也就只有两件的样子。 I mean, okay, then I'd have two. Well, my wardrobe probably looks like I have two. 73:45 但我认为人类的整体进步取决于生产率的提升。而什么是生产率的提升？ But I think the general progress of mankind depends on productivity incr- improvements. And what are productivity improvements? 73:56 这一点非常重要，因为抽象地讲，我觉得每个人都会说：“太好了，生产率提高，好事。”你以为这一切都凭感觉吗？不是的。 This is very important because in the abstract, I think everyone says, "Yay, productivity. That's good." What did you think, this was all just vibes? No. 74:04 生产率意味着用更少的人完成同样多的工作。当然，你想完成的总工作量可能会增加，于是人会变多，但也可能不会。 Productivity means fewer people to do the same number or the same job. Now, the amount of job you want done may increase, and therefore you get more people, but it also may not. 74:16 可能存在一些局部情况：一家公司只需要完成一组固定的任务，突然之间他们用十分之一的人就能做完。顺便说一句，这可能…… There may be pockets where a company just needs a certain set of fixed tasks done, and suddenly they can do them with a 10th the number of people. This, by the way, could... 74:27 对被裁员的个人来说，当下是悲惨而艰难的。但对整个经济体来说，这也是件好事：你突然释放出了这些资源，他们可以去做更有生产率的事情。 is tragic and difficult in the moment for the individual being laid off. It's also amazing for the economy at large. Suddenly you've freed up these resources who can now go do more productive things. 74:38 整个经济就是这样演进和改善的，增长就是这样来的，而增长是好事。 This is how the whole economy evolves and improves and how we get growth, and growth is good. 74:46 我觉得这是另一件值得强力辩护的事：增长总体上是好事。我们不会想把时钟拨回 1920 年或 1950 年。 I think this is the other thing to make a strong defense for, that growth in general is a good thing. We do not want to wind the clock back to 1920 or 1950. 74:59 我们应该为“我们发现了能让我们做成远超从前的事情的新技术”而兴奋。虽然有些可能被 AI 接手的任务确实是人类乐于去做的， We should be excited about the fact that we've discovered new technology that allows us to do vastly more, and while some of the tasks that may be taken over by AI were tasks that humans were excited to do, 75:09 它们只是在经济上不再成立了，但也有很多任务纯粹是苦差事。拿我自己的编程来说，这一点肯定成立。 and they're, they're just no longer economically viable, a lot of the tasks were just drudgery. This is certainly true of my programming. 75:21 编程中有些时刻我觉得妙不可言。那些时刻会产生我们上次聊过的那种心流状态，令人狂喜。 There were moments of programming I thought were amazing. These were the moments that produced that state of flow we talked about last time, and they were ecstatic. 75:33 但这种时刻一年里能有几次？随便找一个普通程序员，一年 2000 个小时的工作里，他有多少时间处在心流状态？ How often did they happen out of the course of a year? If you took your average programmer, how much time out of 2,000 hours at the job did that person s- spend in a flow state? 75:45 嗯哼。 Mm-hmm. 75:47 100 小时？200 小时？ 100? 200? 75:50 还是 25 小时？我觉得这些答案对某个人来说都有可能成立。而把那份苦差事交给机器，正是文明的历史。 25? I think all those answers could be plausible for someone. And taking that drudgery, handing it over to machines, that is the history of civilization. 76:03 是啊，我是说，花在调试上的那些时间。编程里有太多痛苦的……痛苦的时光。 Yeah, I mean, the hours spent debugging. So the pr- programming has so many painful ... painful hours in it. 76:10 对。 Yes 76:10 而现在它们没了。我是说，在很大程度上，编程现在只剩下乐趣了。 And now they're gone. For the, I mean, for the most part it's just fun. 76:15 在很多领域里，大部分痛苦都消失了，剩下的可能很快也会消失。所以有些浪漫化情绪也是从这里来的—— A lot of it is gone in a lot of domains, and the rest of it may be gone very soon. So that's where some of the romanticism also comes in- 76:24 哦。 Okay 76:24 ……因为身处其中的痛苦是难受的、煎熬的、烦人的。 ... because pain in the moment is difficult, it's hard, it's annoying. 76:30 而后视镜里的痛苦是成就、是自豪、是学习、是前进的脚步，是这一切，对吧？ Pain in the rear view mirror is accomplishment, proud, learning, steps forward, all the things, right? 76:38 我们把这一切都当作人类体验的一部分：身处其中时我们憎恨痛苦，但过去五分钟之后我们又珍惜它——我们不记得那份艰辛，只记得那份进步。 So we encompass all of this as part of the human experience, that we hate pain in the moment, but we cherish it when it's five minutes passed and we don't remember the, the hardship, we just remember the progress. 76:51 所以在社会层面、在，呃，宏观经济层面，增长当然令人兴奋。但在个体层面，会有大量焦虑，呃，可能还有大量痛苦。 So certainly at the societal level, at the, um, macroeconomic level, growth is exciting. But the individual level, there's going to be a lot of anxiety, uh, potentially a lot of suffering. 77:02 那么给点建议的话：如果面对的是一个年轻的 DHH，或者一个年轻的开发者、程序员，你会、你会建议他们做什么？ So by way of advice, if you're like a young DHH or young developer, programmer now, what would you, what would you advise they do? 77:11 不要试图预判任何事。你真的会疯掉的。 Don't try to anticipate anything. You will literally go crazy. 77:15 嗯。 Mm. 77:15 因为即使是这行里最聪明的大脑，也无法预判再迭代两代模型之后世界会是什么样。 Because even the smartest brains in the business cannot anticipate what two model hops from here is going to look like. 77:24 试图推演两年后的样子纯粹是浪费时间，而且你会因此得上“AI 精神病”。 It's an absolute waste of time, and you will develop an AI psychosis trying to deduce what two years from now is gonna look like. 77:31 专注于当下。而当下是投身计算机领域最不可思议的时代。 Focus on right now, and right now is the most incredible time to be into computers. 77:38 嗯哼。 Mm-hmm. 77:38 只要你全身心投入，就能让电脑做出最惊人的事情。 You can make them do the most amazing things if you lean in. 77:42 只要你强迫自己花点时间去了解最前沿进展到哪里了，我打赌你不可能不为“什么是可能的”感到兴奋。 If you maybe, for a hot moment, force yourself to learn where the state of the art is, I double dog dare you not to get excited about what's possible. 77:57 那，呃，你会建议公开地做项目（build in public）吗？ And, uh, would you recommend building publicly? 77:59 我觉得公开做其实是可选项。这么做挺好的，因为你能成为一个社区的一份子，开源里有一种同袍情谊。 I think it's optional, actually, to build public- publicly. I think it's, it's nice to do because then you get to be part of a community, and there's some camaraderie in open source. 78:09 不是“一些”同袍情谊，是海量的同袍情谊。 Not some camaraderie, there's an enormous amount of camaraderie. 78:13 开源是社区和情谊的绝佳源泉，也是抵御个人生存焦虑的好方法。 Open source is an incredible source of community and camaraderie, and it's also a great way to stem some of that personal existential dread. 78:24 如果你身处人群之中，你会感染上好的“思想病毒”、令人兴奋的“思想病毒”——那种告诉你“一切都会好起来、未来一片光明”的病毒。 If you're amongst other humans, you can get infected with good forms of mind viruses, exciting forms of mind viruses, the ones that tell you that it's all gonna be okay, and the future looks bright. 78:37 如果我们一起做出一堆东西，我们就能做到我们任何人做梦都想不到的事。 And if we build a bunch of things together, we can do things none of us ever dreamed we could do. 78:43 我刚和 Ryan Hughes 聊过，他几乎从第一天起就是我在 Omarchy 上的主要搭档。我们昨天还在聊：当初一起踏上这段旅程的时候，我们的野心非常朴素。 I was just talking to Ryan Hughes, who's been my main partner on Umachi almost since day one, and we were literally talking about yesterday how when we started on this journey together, our ambitions were very modest. 78:57 我们当时就想：“能不能让这个发行版别因为一个不受我们控制的 Arch 软件包刚推出来，就在凌晨三点挂掉？能做到那就太好了。” We're just like, "Can we get this distro to not crash at 3:00 AM because of an arch package that just got pushed out outside of our control? That would be amazing, then it'd be great." 79:08 现在我们靠在椅背上想的是：“我们怎样才能征服世界？怎样才能让 Omarchy 上火星？ Now we're leaning back and thinking, "How can we take over the world? How can we make Umachi go to Mars? 79:16 怎样才能实现我们有过的所有梦想？”这种野心膨胀一部分是因为智能体让事情变好了，但很大程度上也是因为我们身边聚集了一群对同一段旅程充满热情的人。 How can we do everything we've ever dreamed of?" And that level of ambition inflation is partly because things have gotten better with agents, but hugely also because we've surrounded ourselves with others excited about the same journey. 79:32 嗯哼。 Mm-hmm. 79:33 而如果你只是一个人坐在与世隔绝的小洞穴里为未来担忧，那你确实会、会有点发疯。 And if you are just sitting in your little isolated cave worrying about the future, yeah, you're gonna go, you're gonna go a little nuts. 79:41 嗯哼。 Mm-hmm. 79:42 如果说新冠教会了我们什么，那就是：人要是独自憋在自己的小洞穴里担心将要发生的事，是会疯掉的。 I think if we learned anything during COVID is that people will go nuts sitting in their little cave inside by themselves worrying about what's gonna happen. 79:52 那跟抑郁没什么区别。无休止地反刍那些——无法控制的事情。 That is indistinguishable from depression. Ruminating endlessly about things- Can't control. 80:01 嗯哼。 Mm-hmm. 80:02 恕我在这里引用一下 Jensen Huang：那只是失败者的论调。你不必当失败者。你可以选择全身心投入，然后—— I'm sorry to channel some Jensen Long here, but that's just loser talk. You don't have to be a loser. You can choose to lean in and- 80:12 嗯哼。 Mm-hmm 80:12 ……赢。而“赢”的定义很宽：学得更多、做得更多、贡献得更多、参与得更多。 ... win. And win is very broadly defined as learning more, making more, contributing more, being part of more. 80:23 说到底，你有什么选择？你没得选，朋友。 And at the end of the day, what are your choices? You don't have a choice, mate. 80:28 嗯哼。 Mm-hmm. 80:29 不管你喜不喜欢，未来都会来，所以你不如选择为它兴奋。 The future's coming whether you like it or not, so you might as well choose to be excited about it. 80:34 但是，我、我是说，我们应该强调一个事实：有了 AI，做计划几乎是不可能的。我们大多数人多少都会规划一下人生。年轻的时候，你有希望和梦想，你去上大学是有原因的。 But, I, I mean, we should emphasize the fact that planning is nearly impossible with AI. I mean, most of us plan our life a little bit. When we're younger, you have, like, hopes and dreams. There's a reason you go to college. 80:46 那背后是有一个计划的。 There's a plan underlying that. 80:47 我同意。 I agree with that. 80:48 但—— But there- 80:48 这一点我承认。 I'll concede that 80:49 ……这确实有点—— ... it is kinda- 80:51 这次不一样。 This is different 80:51 ……疯狂：真的没法做计划了。 ... nuts that it's impossible to really plan. 80:54 是的。 Yes. 80:54 因为，你真的不知道，呃，六个月后还会不会有 Claude Code、Codex、Cursor。也许到时候你是对着 WhatsApp 说话。 Because, like, you really don't know, uh, if there will even be a cloud code, a Codex and Cursor six months from now. Maybe it'll be you'll be talking to WhatsApp. 81:06 也许全是开放调用的形态，也许全都变成语音了。也许 Omarchy 会征服世界，然后全是语音交互。也许操作系统都不复存在了。 Maybe it'll be all open call type. Maybe it'll just all be voice. Maybe Omachi will take over the world and they'll be voice. Maybe there won't be operating systems anymore. 81:15 到时候可能就有一个发光的球体，你只要跟它说话就行。因为如果自然语言就足以完成人生所有任务，那人生是干嘛的？人生中需要完成的主要任务又是什么？ You'll just, there'll be this glowing orb that just, you just talk to because if natural language is enough to accomplish all tasks in life, like, what is life about? Like, what are the major tasks you need to accomplish in life? 81:28 如果你用语音就能搞定一切，程序都由它来写，一切都系统化了，那、那还需要什么工作？也许剩下的更多是服务业。 If you can just do that with voice, and it writes all the programs for you, and all is systematized, and what, what are the jobs needed then? Maybe it's more about the service industry. 81:38 也许以后根本不会有程序员了，呃，什么智能体工程师都不会有。 Maybe there won't be programmers, like, uh, agentic engineers at all. 81:43 是的，你建议专注当下很明智，但这并不意味着这就是一场疯狂的过山车，因为我们不知道今年十二月或明年一月会是什么样子。 And so yes, your advice is wise to focus on the moment, but that doesn't mean that it's just a crazy rollercoaster ride, 'cause we don't know what this December or what this January will look like. 81:55 我们不知道，所以，我们应该选择相信事情会有好结果。 We don't know, and therefore, we should choose to have faith that it's gonna pan out. 82:01 我对抗焦虑最好的建议就是，在很大程度上，你可以选择往哪边倾斜。 That is my best advice towards fighting anxiety, is that you have a choice, to a large degree, of how to lean. 82:13 嗯哼。 Mm-hmm. 82:13 你想倾向末日论那边吗？ Do you wanna lean into P doom? 82:17 末日论不管你倾不倾向它，它该发生还是会发生。 P doom is gonna happen whether you lean into it or not. 82:21 你控制不了。 You don't have control. 82:22 但也有实际的日常安排：你每天学多少东西、做多少东西、呃，做多少多元化的事情。 But there's also actual daily schedule of how much learning are you doing, how much building are you doing, how much, uh, uh, diversification you're doing. 82:31 比如，我最近去了中国农村旅行，而且我是特意这么做的。 And for example, I traveled across rural China recently, and I did it quite deliberately. 82:37 因为你在和人们交谈，在中国，在任何一个乡村，那里时间流逝得慢得多，没有人在讨论智能体工程。 Because you're talking to people, uh, in China, in, in rural anywhere, where time moves much slower, where there's no discussions about agentic engineering. 82:49 而，而我一开始是去了中国的长城。那是永恒的，跨越了几个世纪。你知道，当时有一种 FOMO 的紧张感，好像我可能错过了什么。 And the, the ti- and I, I started by, uh, going on the, on the Great Wall of China. That's timeless. That spans centuries. And, you know, there was, like, this kind of tension of FOMO, like maybe I'm missing out on something. 83:02 但后来我非常刻意地想去寻找人类存在的永恒感，真正去感受这样一个事实：好吧，纵观人类历史，似乎一切都在变。 But then I very deliberately wanted to find the timelessness of human existence, really plug into the fact that, okay, it, it always feels like everything's changing, uh, uh, throughout human history. 83:15 但你知道，那些普适的东西依然是真的，关于生命中什么才是最重要的，我只是想确保自己始终与之保持连接。 Uh, but, you know, the universals are still true, uh, of what matters in life, and I just wanted to make su- keep connected to that. 83:22 你不会错过任何东西的。这一点在这场转型的早期阶段真的让我很不爽。每个人都特别激动，觉得你要是不…… You're not gonna miss anything. This is the part that actually grinded my gears in the early phases of this transition. Everyone was so up in arms that if you didn't... 83:35 现在流行 loop 了。哦不不，loop 过时了，现在流行 graph 了。哦不不，那个也过时了，现在是 harness 之类的，对吧？他们不断地在 frontier 上翻烧饼，这在某种意义上其实非常令人兴奋。 It's loops now. Oh, no, no, we're done with loops. It's graphs now. Oh, no, no, we're done with that. It's, it's harnesses this, right? They're, they're constantly churning through the frontier, which in one way is actually very exciting. 83:46 新领域打开的时候就是这样，就像一扇新的传送门，东西全都涌出来。这很令人兴奋。 This is what happens when a n- new field opens up. It's like a new portal, and it's all spilling out. That's exciting. 83:53 但这也意味着，如果我过去一年一直在背包旅行，没碰过电脑，没有见证这个智能体时刻，昨天才出现，你知道吗？ But it also means that if I had just been backpacking for the last year, hadn't touched a computer, hadn't witnessed this agentic moment, and I just showed up yesterday, do you know what? 84:06 我两周就能追上。这里没有任何积累，这在某种程度上是巨大的解脱。如果你错过了过去一年，你可以在两周内赶上 frontier。 I would've been caught up in two weeks. There's not any accumulation, which is in some ways a great relief. If you missed the past year, you can catch up to the frontier in two weeks. 84:19 如果你是一个程序员，在喜马拉雅山徒步了一年然后回来，你两周就能追上。这实际上恰恰证明了我们正在经历的进步有多快。 If you're a programmer who was out hiking the Himalayas for a year and you come back, you can catch up in two weeks. And this is actually the great credit to the progress we're experiencing. 84:30 我们现在有无数实验在同时进行，不断而无情地筛选什么有效、什么无效。你不需要记住甚至不需要参与整个过程。 We have so many experiments running simultaneously right now that are constantly and ruthlessly sorting what works, what doesn't work. You don't have to remember or even be part of that entire journey. 84:43 你只需要直接来拿结果。 You can just show up for the results. 84:45 但关键的一步是，当你背包旅行回来时，要呃，愿意成为一个完全不同的人，因为事情变化得太快了。我是说—— But the key step there, when you come back from the backpacking journey, is to, uh, be willing to become a totally new, different human because things are changing so fast. I mean- 84:57 没错。 That's right 84:57 ……真的，如果你去年十月去背包旅行……今年四月或五月回来—— ... literally, if you left for the backpacking journey- ... in October last year and came back in April or May- 85:04 那你跟被冻在低温舱里一百年也差不多了。 You might as well have been in the cryo chamber for 100 years 85:06 ……就像，你什么意思？我们不再编程了？这，这是一场巨大的变革。 ... it's like what, what do you mean? We're not programming anymore? This, it's a v- it's a huge transformation. 85:12 这很惊人，而且我其实……我是说，我有一部分是故意夸张了一点为了效果。我也承认哀悼一阵子是可以的。那不是我的性格。 It's startling, and I actually... I mean, I'm putting some of this on slightly for effect. I also recognize that it's okay to grief for a hot moment. That's not my disposition. 85:23 我不是那样处理事情的。 It's not how I do it. 85:24 但我接受那是一个经受时间考验的应对失落感的机制，有失落感是可以的——你所熟知、也许还很热爱的那个世界已经不一样了。 But I accept that that's a time-tested mechanism for coping with a sense of loss, and it's okay to have a sense of loss that the world that you knew and maybe you were very fond of is no longer the same. 85:43 我也许应该——我应该大声明确说出来，我确实有那么一种感觉，就像那个眼泪顺着脸颊流下来的印第安人……有一种悲伤。这是一次告别，是对旧编程世界的告别。 I maybe sh- I should say out loud and explicitly that I do have that, the Native American with a tear rolling down my... Like, there's a sadness. It's a goodbye. It's a goodbye to the old world of programming. 85:55 这，这让我非常着迷。 It, it's so fascinating to me. 85:56 我花了那么多年—— I've spent so many years- 85:58 但这一切都汇聚到了这一刻。对我来说，这正是整个旅程如此有意义的原因。AI 不是从天而降的。 But it all amounted to this moment. That is to me what makes the entire journey so meaningful AI did not arrive from the sky. 86:07 它不是什么穿越宇宙、突然出现在我们岸上的外星技术。AI 存在的时间和计算机科学一样长。 It was not alien technology that had blasted through the universe and suddenly showed up on our shores. AI has been around for as long as computer science has been around. 86:17 在五十年代，当时最聪明的人们认为这个问题大约十年内就能被攻克。 In the '50s, they thought the problem was gonna be cracked in about a decade, the smartest people at the time. 86:23 结果花的时间更长了一点，但花这么久的部分原因是我们在通往 AI 和神经网络的路上拐了几个弯，以为还有其他符号表示方法会胜出。 It took a little longer, but the reason it took a little longer was in part because we took some blind turns on the road to AI and neural networks, and thought that there were other symbolic res- presentations that were gonna carry the day, 86:35 结果并没有。那大概让我们倒退了十五年。还有一部分原因是，我们必须经历游戏革命，3D 游戏。 and they didn't. And that probably set us back about 15 years. So some of it was also, we had to, um, go through the gaming revolution. 3D games. 86:47 如果没有 Quake，如果没有 Duke Nukem，如果没有 Unreal Tournament，我们就永远不会有 AI，因为我们就永远不会有 GPU，因此也就永远不会有今天这种形态的 AI。 If we had not had Quake, if we had not had Duke Nukem, if we had not had Unreal Tournament, we'd never have gotten AI because we would never have gotten the GPUs, and therefore we would never have been able to get AI in the shape it is now. 86:59 所以我们到此刻为止所做的一切，都是抵达这里所必需的。 So everything we did up until this moment was required to get there. 87:05 对。 Yeah. 87:05 你也参与其中。天哪……这是多大的福气。你以自己的小小方式做出了贡献，无论是靠训练还是靠打电子游戏，这都是多大的福气。真的，你完全可以选择这样去看待它。 You were part of that. Holy... What a blessing. What a blessing to have contributed in your small part, either by training or by playing video games. Literally, you could just choose to think at it that way. 87:18 拿我来说，九十年代那些打电子游戏的时光，它们是—— All those times you spent in, for my case, in the '90s just playing video games, they were- 87:23 你在做贡献。 You were contributing. 87:24 对，我当时就在为 AI 革命做贡献，就在那时那刻。 Yeah, I was contributing to the AI revolution right there and then moment. 87:27 为人类文明的进步做贡献。不，确实。文明伴随着旧事物的消亡和新事物的诞生而进步，它就这样延续下去，但我认为，哀悼的过程仍然是其中的一部分。 To the progress of human civilization. No, for sure. Civilization progresses with the death of the old and the birth of the new, and it continues in this way, but it's still the grieving process is a part of it, I think. 87:39 缓一缓。没关系的。缓一缓……而且，你甚至不需要接受一切都死了。 Take a moment. It's okay. Take a... And also, you don't even have to accept that it's all dead. 87:45 比如，如果你还非常呃，眷恋手工雕琢代码这件事——如果你一年前问我，我会预测我会更加眷恋。我很惊讶—— Like, if you are still very, uh, attached to the hand chiseling, which if you'd asked me a year ago, I would've predicted that I would've been more attached. I'm surprised- 87:58 我有点惊讶。 I'm kinda surprised 87:59 ……有点惊讶我竟然没有更眷恋。但我想，我没有更眷恋的原因是我找到了更好玩的东西。 ... a little that I've not been more attached. But the reason I've not been more attached, I think, is that I found something more fun. 88:07 如果不是这样，如果 AI 只是取代了我热爱的东西，然后给了我一个我讨厌的东西，我可能会有点怨气。 And if it hadn't been like that, if AI had just replaced a thing I loved and gave me something I hated, I'd probably be a little bitter. 88:16 我是说，这真的感觉……看着这一切发生，既好笑又 awesome，而且鼓舞人心。 I mean, this really feels... It's, it's hilarious and awesome to watch and inspiring to watch. 88:19 就像毕加索突然接触到了图像生成，超级兴奋，几个月内就转型了。这就是真实发生的事情。 It's like Picasso all of a sudden f- like, getting access to, uh, to image generation and getting super excited and switching in a matter of months. That's literally what happened. 88:32 就像一个欣赏优美代码的艺术家，突然之间你就在谈论玩得很开心—— So like, it's an artist that appreciated beautiful code, and all of a sudden you're talking about having fun- 88:37 我觉得毕加索其实是个很好的类比，因为如果你看他早期练手艺的作品，他画的是写实画，对吧？ I think Picasso is actually a good parallel because if you look at his early work where he learned his craft, he was painting realistic paintings, right? 88:48 他当时是在训练自己成为用旧有方式描绘现实的大师，那是我们所知道的最好的方式。然后各种其他表达形式出现了，立体主义出现了。 Like, he was training to be a master of the old ways of depicting reality the best way we knew how. And then along comes all these other forms of expression. Cubism comes along. 89:01 嗯哼。 Mm-hmm. 89:01 还有所有这些抽象表达的形式。他没有哀叹自己不再画文艺复兴式的画了，不再只是画一个完美的苹果了。 And all these o- forms of abstract expression. He's not decrying that he's not doing Renaissance paintings anymore, that he's not just painting the perfect depiction of an apple. 89:14 他把苹果重新想象成一个该死的方块，而且他对此兴奋不已。 He's reimagining the apple to be a freaking square, and he's getting excited about that. 89:20 我感受到了这种转变，我们从一种表达模式，突然打开了一扇大门，那里的……就像，色彩光谱宽了太多。 And I feel that transition, that we've gone from this one mode of expression, and we've suddenly opened the gates where the... Like, the color spectrum is so much wider. 89:33 我看到了更多东西。这也许是因为我成为程序员的方式。我成为程序员不是因为深爱 if 语句。 I'm seeing more. And it's perhaps because the way I became a c- programmer. I did not become a programmer because a deep love of if statements. 89:46 我成为程序员是因为我想要程序。 I became a programmer because I wanted programs. 89:49 嗯哼。 Mm-hmm. 89:49 我想要让不存在的东西存在。一开始，我必须学会编程才能做到这一点。 I wanted things to exist that did not exist. And at first, I had to learn how to program to make that happen. 89:56 嗯哼。 Mm-hmm. 89:56 后来我恰好爱上了编程这门手艺，作为——花了二十年真正深入钻研它，但现在我回来了。 Then I happened to fall in love with programming as a craft and as- Spent two decades really diving deep on that, but now I'm back. 90:05 我回到了最初的状态：有一个想法，然后迫不及待地想看到它存在于这个世界上，而 AI 把这个过程缩短到了几乎为零。 I'm back to having an idea and being impatient beyond belief to see it exist in the world, and AI has just shrunk that down to almost nothing. 90:18 所以这是一次重新发现的加速。当然，我明白并不是每个人都是这样。 So it's an acceleration of a, of a rediscovery. Now, I understand that that's not true for everyone. 90:26 有很多程序员从一开始就纯粹热爱编程，他们被逻辑结构之类的东西所吸引。我在那个状态待了二十年，所以我懂。 There are plenty of programmers who just loved programming right from the beginning, and they were attracted to the logical constructs and so forth. I mean, I've been there for two decades, so I get it. 90:36 但我也认为，你能够发现自己新的一面—— But I also think that you're able to discover new sides of yourself- 90:41 嗯哼。 Mm-hmm 90:41 ……只要你让它进来。再说一次，我一直回到这个几乎是遗憾最小化的框架，借用 Jeff Bezos 的说法。 ... if you let it in. Again, I keep coming back to this almost regret minimization framework, to borrow Jeff Bezos's terms here. 90:52 两年后的你回头看自己，会觉得“哦，我把时间花在对现状、对我的行业变了这件事发脾气上，花得值”，还是说你会—— Are you gonna look at yourself two years from now and think, "Oh, I spent my time well being pissy about the present, about the fact that my industry changed," or would I, 91:02 两年后回头看这一刻，为自己迎头而上、为也许哀悼了一阵子而感到骄傲？再说一次，所有这些都有空间。 two years from now, look back upon this moment and be proud of myself for leaning in, for maybe having a moment of grief. Again, there's room for all of it. 91:14 然后想：“我要搞清楚这个世界现在是怎么运转的，我不只要充分利用它，我还要创造出更多。” And then going, "I'm gonna learn how the world spins now, and I'm gonna make, not just the best of it, I'm gonna make more of it." 91:22 我是说，这确实是真的，很多程序员，我的很多朋友，我，显然还有你，都玩得很开心。 I mean, it is true, a lot of programmers, a lot of friends of mine, me, obviously you, are having fun. 91:30 过去三个月我用电脑获得的快乐，比之前任何时候都多。 I have had more fun with computers in the last three months than at any time previously. 91:38 先快速去趟洗手间，然后我得问问你，呃，你现在到底是怎么和这些系统交互的。 Quick bathroom break, and then I gotta ask you about, uh, how you actually interact with the systems now. 91:45 好。 Yes. 91:46 我得问问你，呃，你的编程装备是怎么变的。键盘、语音、用的什么 IDE？ I gotta ask you about, uh, how has your setup, uh, programming setup changed. So keyboard, voice, what's the IDE? 91:57 现在想起来很疯狂，但我用了 TextMate 差不多二十年。 It's crazy to think about now, but yeah, I used TextMate for almost 20 years. 92:02 嗯哼。 Uh-huh. 92:02 我想我是从 2005 年开始用 TextMate 的。我帮忙把第一个版本弄出来，之后我就没兴趣换。 I used TextMate starting in 2005, I think. I helped get the first version out, and then I just wasn't interested. 92:10 我不需要替代品，直到换到 Linux，我才被迫离开了自己的舒适区。 I wasn't in the market for an alternative, and it wasn't until the switch to Linux that I was forced out of my habitat. 92:20 而现在换到了，叫什么来着？智能体工程？哦，我真他妈讨厌这个词。 And now with the switch to, what are we calling it? Agentic engineering? Oh, I fucking hate that term. 92:28 我们得想个听起来和“编程”一样朴素、但又能体现有智能体参与的词。但—— We gotta come up with something that sounds as plain as programming, but encapsulates the fact that it's with agents. But- 92:35 我还是觉得现阶段就应该叫编程。 I still think it should be called programming at this point. 92:36 行，那就叫编程吧。 All right, let's just call it programming. 92:37 好。 Yeah. 92:38 和智能体一起编程需要一套不同的工具。 The programming with agents requires a different tool set. 92:41 确实需要，主要的变化是你从脑子里的单线程编程变成了并行处理。 It really does, and the main change here is that you're going from single thread programming in your head to parallel processing. 92:53 以前我在 TextMate、甚至不久前的 Neovim 里手工雕琢代码的时候，我一次只专注一个问题，有条不紊地把它做完，而这正是进入心流的入口。 When I was writing code, chiseling it by hand in TextMate or even Neovim, uh, not that long ago, I would just focus on one problem at the time, and I would methodically work my way through it, and that was actually the portal to flow. 93:07 心流的入口就是深深沉浸在一个问题里，从头做到尾。 The portal to flow was deep immersion into a single problem, see it through to the end. 93:13 嗯哼。 Mm-hmm. 93:15 用智能体就不是这么工作的。部分原因是智能体既太快又太慢。 That's not how it works with agents. In part because the agents are at once both too fast and too slow. 93:23 你让它们做一件事，它们不会像敲键盘那样立刻给你回复。所以你得让智能体先煮一会儿—— They don't give you an immediate reply on something that you ask them to do that's the same as typing on a keyboard. So you have to let the agent cook- 93:34 嗯哼。 Mm-hmm 93:34 ……煮一阵子，然后你就会意识到，嗯，如果我就干坐着等它们，首先，那感觉不像在干活。哪怕一个智能体效率很高，感觉也不像。 ... for a bit, and therefore you realize, well, if I just sit around waiting for them, first of all, that doesn't feel productive. Even if the agent, just one of them, can be highly productive, it does not feel productive. 93:47 感觉不好。实际上会让你觉得自己有点没用。 It doesn't feel good. It feels actually like you're a little bit useless. 93:49 嗯哼。 Mm-hmm. 93:50 也许第一个智能体时刻到来的时候，我有过那么一阵子——当时我一次基本只跑一个智能体——我心想：“我对这事儿拿不准。” And maybe I had a moment when the first agentic moment was there, and we, I was running mostly one agent at a time where I felt like, "I don't know about this." 94:00 但很多难题都可以通过简单地堆更多资源来解决。这就是 AI 本身的整套 scaling law，对吧？ But you can solve a lot of hard problems by simply throwing more resources at it. This is the, uh, whole scaling law of AI of itself, right? 94:09 如果你把这些东西并行化，你不是只跑一个智能体，而是跑一把，你就能呃，感觉自己处于心流状态，因为你在不断地做编程意义上的工作：你在做决策， That if you paralyze these things and you're not running one agent, but you're running a handful, you can, um, feel like you're in a flow state because you're constantly doing programming work in the sense that you're making decisions, 94:22 你要么在帮某个智能体解锁——因为它纠结该往哪个方向走——要么准备好接新任务。而要做到这一点，你需要一套不同的装备。 and you're helping either unblock an agent because it has a question about which direction to take, or you're ready for a new task. And to do that, you need a different setup. 94:35 我一开始是在 Tmux 里做，开不同的面板、不同的分屏。基本上可以把它想成一个带标签页的终端，你打开一堆标签页。 I started first doing it in Tmux and just having separate panes and having separate splits. Basically a terminal with tabs is a good way to think about it. You open a bunch of tabs. 94:45 我觉得大多数人都很清楚那是怎么回事，没人只在一个标签页里干活。 I think most humans know exactly how that works. They don't work in just one tab. 94:50 所以呃，还是坚持用终端、CLI。 So, uh, still sticking to the terminal, CLI. 94:53 当然。 Absolutely. 94:53 所以你不用 Claude Code app 或者—— So you're not using Cloud Code app or the- 94:56 不用。 No 94:56 ……Codex app 或者—— ... Code Ex app or the- 94:57 我，我很喜欢这场智能体革命是从终端里开始的，因为我本来就超爱 TUI——终端用户界面——以及整个终端世界。 I, I love the fact that this agent revolution was kicked off in the terminal because I was already a huge fan of TUI's terminal user interfaces, and the terminal in general. 95:07 那感觉是个特别好的地方，是个很美的地方。现代终端就是一个赏心悦目的工作环境。所以我喜欢这一点。 That feels like a, a, a really nice place to be. It's a beautiful place to be. The modern terminal is just a good-looking place to work. So I like that. 95:19 对。 Yeah. 95:19 然后，呃，当你有多个智能体的时候，尤其是不只是在你自己机器上跑多个智能体，而是开始在多台机器上跑的时候—— And, um, and then this fact of having multiple agents, especially once it's not just multiple agents running on your own machine, but you start running multiple machines. 95:31 光靠 Tmux 就不够用了，所以最近我换到了一个叫 Herder 的东西。 Now Tmux alone is not enough to keep track of it, and that's why as of late I've switched to this thing called Herder. 95:39 嗯哼。 Mm-hmm. 95:40 Herder 本质上就是 Tmux 加智能体通知。每当你的智能体干完了、需要你做点什么，它就“叮”一声。 And Herder is essentially Tmux plus agent notifications. So whenever your agent is done and needs something for you, it goes ding. 95:50 一个小呃，一个小铃铛告诉你它准备好等它的人类了，呃……到底是主人还是仆人？我一直没太搞明白。 A little, uh, a little bell telling you it's ready for its human, uh... Is it, is it master or servant? I'm not quite sure always. 95:58 但它是准备好让你做决策了。它还会追踪这些智能体是不是在工作。所以我搭了这套 Herder，实际上我在好几台机器上各跑了一套 Herder。 But it is ready for a decision And it also keeps track of these, is it is working or not. So I have this Herder set up. I have multiple Herder setups actually running on individual machines. 96:10 大约一个月前我进入了一个疯狂阶段，意识到在一台机器上干这些活不够快。 I went on this crazy phase just about a month ago, realizing that doing this work on a single machine is not fast enough. 96:19 就像我发现了多核编程，但我只有两个核。 It's like I've discovered multi-core programming, but I only have two cores. 96:23 嗯哼。 Mm-hmm. 96:23 我就想，如果我有 16 核呢？32 核呢？64 核呢？于是我立刻出去买了这些超棒的 KVM，叫 gli.net Comet。 I'm like, what if I had 16 cores? What if I had 32 cores? What if I had 64 cores? So I instantly went out and I bought these amazing KVMs called gli.net Comets. 96:35 嗯哼。 Mm-hmm. 96:35 它们就是一个小盒子，你插上 HDMI，插上呃，USB，连到电脑上。就是个 KVM。 And what they do is it's this little box. You plug in HDMI, you plug in, um, USB and connect it to the computer. It's like a KVM. 96:44 KVM 就是一种远程控制电脑的方式，但这个的特别之处在于它简单得离谱。你把这东西接上，打开一个网页，登录一次，设一个密码。 So a KVM is a, a remote way of controlling a computer, but what's special about this is just how easy it was. You connect this thing in, you go to a webpage, log in once, set one password. 96:56 然后这东西就能加入你的 tailnet。 Now this thing can hop on your tailnet. 96:59 嗯哼。 Mm-hmm. 96:59 这对我来说是过去一年的另一场革命：发现了这些 WireGuard 网络。 This has been the other revolution of the last year for me, is discovering these WireGuard networks. 97:07 Tailscale 本质上就是把你所有的电脑变成一个局域网，不管你在哪。 E- tail scale is essentially turning all the computers you have into a local network wherever you are. 97:15 嗯哼。 Mm-hmm. 97:16 比如现在，我在手机上就能直接访问我 Malibu 办公室里的所有电脑，也能访问我 Copenhagen 办公室里的所有电脑。 Like right now on my phone, I have direct access to all the computers in my Malibu office. I also have access to all my computers in my Copenhagen office. 97:25 嗯哼。 Mm-hmm. 97:26 我可以把它们当成就在我旁边一样用，不用在防火墙上打洞，也不用搭复杂的 VPN。 And I can treat them as though I sat right next to them, without having to punch holes in a firewall or set up complicated VPNs. 97:34 它的作用就是降低了让新算力上线的摩擦。 And what this does is it just decreases the friction it takes to get new compute online. 97:40 嗯哼。 Mm-hmm. 97:41 所以我一发现这个，就翻了翻我的储藏间，发现里面有一堆以前实验用的迷你 PC。我就想：“要是把它们全连上呢？” So as soon as I discovered this, I looked in my closet and I realized I had a bunch of mini PCs from prior experiments. I just said, "What if I just connected all of them?" 97:50 哦，哇。 Oh, wow. 97:51 然后我就把其中四台电脑连上了—— And I just connected four of the computers- 97:53 不错。 Nice 97:53 ……放在储藏间里。每台都插着自己的小 Comet，突然间我就能同时在更多电脑上跑智能体，而且全部用 Herder 控制。 ... in a closet. They all had their little Comet, and suddenly I could run agents on more computers at the same time, and I could control them all with Herder. 98:04 后来确实到了我自己的处理能力跟不上的地步。 And it did get to a point where I maxed out my own processing power. 98:08 对。 Right. 98:08 我觉得大概在…… That I think at about... 98:12 怎么说呢？大概四五台机器，每台跑，我不知道，三个智能体。我大概能跟 16 个线程。 What do we want to put it? About four to five machines running, I don't know, three agents. I have about 16 threads. 98:20 对。 Yeah. 98:20 这是我能驾驭的上限。当然，智能体跑得越快，我能跟的线程就越少，但按现在的速度，我大概能跟 16 个线程—— That's what I can run. And the faster the agents run, of course, the fewer threads I can run, but at the current pace, I can run about 16 threads- 98:26 嗯哼。 Mm-hmm 98:26 ……全速运转。 ... at, at full acceleration. 98:29 这也是为什么我从只是对智能体时刻感到兴奋，变成了有点狂喜，因为我觉得，就像你说的，其中一些确实有点冲击。 And that's part of why I've gone from just being excited about the agent moment to being delirious because I think, as you say, some of it is a little assaulting. 98:42 再——再说一次，我们和电脑之间的带宽还是拨号级别的。这真的挺好笑的。你想想，一年前我坐在这儿，对吧？我在手工雕琢代码。 A- again, we're on, we're on a dial-up bandwidth-wise with our computer. I mean, it's actually hilarious. So you think of I'm sitting here a year ago, right? I'm chiseling my code. 98:53 对。 Yeah. 98:53 我在写我的一行行代码。 I'm writing my lines. 98:55 对。 Yeah. 98:55 大概一个小时，我会写出一个漂亮的 controller 或者一个漂亮的 model。就一个文件，而且我是认真打磨过的，对吧？最后可能剩下 60 行。 And over the course of, like, an hour, I will have written one beautiful controller or one beautiful model. Like, that's one file, and I've really worked on it, right? So maybe there's 60 lines left. 99:07 所以那时候我的带宽大概是每小时 30 行。我觉得这都算高的，可能只有每小时 20 行。现在我跑 16 个线程。 So maybe my bandwidth at this moment is, like, 30 lines an hour. I think that's even high. Maybe it's 20 lines an hour. Now I'm running 16 threads. 99:17 有时候我每小时能产出几百行，呃，几百行代码。等一下，我先停一下。我讨厌这个指标，对吧？ I'm producing at some times hundreds of lines of hour or, or, uh, hundreds of lines of code per hour. Now, let me pause myself there for a hot moment. I hate that metric, right? 99:28 代码行数总体来说是个愚蠢的衡量标准，而且我觉得编程社区里有些人嘲笑那种 AI 狂热是对的——人人都在谈论自己写了多少行代码， Like, lines of code is a stupid metric in general to measure these things, and I think it's right from some in the programming community to ridicule the AI psychosis when everyone talks about the number of lines of code they're writing, 99:39 然后你问他们：“你做出了什么？”他们就说：“呃，我……”然后就答不上来了，对吧？ and then you ask them, "What did you build?" And then like, "Well, I..." And then no good answer comes out, right? 99:47 这只是个方便的简称，用来描述这件事。 It's just a nice shorthand to frame it. 99:49 它确实是个方便的简称，是对产出量的一种表示或概括。至于产出是好是坏—— It is a nice shorthand. It is a representation or encapsulation of how much output that's coming out. Now, whether the output is good or bad- 99:57 嗯，对。 Mm. Yeah 99:57 ……那另当别论，但我现在的产出确实多得多，对吧？所以我才能让所有这些线程同时跑。这就是这套装备。 ... is not a referendum on that, but I was producing And producing so much more now, right? And therefore I'm able to keep all these threads going. So that's the setup. 100:06 还是用 Neovim，但到这个阶段我已经不怎么写代码了。 It's still Neovim, but at this point I'm not writing a lot of codes. 100:11 我把 Neovim 当项目浏览器用，然后用它启动 Lazy Git 来看改动日志。 I'm using Neovim as a project browser and then as a way to kick off, uh, Lazy Git to see the change log for what's there. 100:20 不过说实话，如果 GitHub 显示 pull request 能再快一点，网页端可能其实是更好的选择。不好说。 And even that I'd say if GitHub was a little faster at showing you your pull request, the web is probably actually a nicer place to do that. I don't know. 100:29 还有另一个我玩过一阵子的工具叫 Hunk，它能以非常漂亮的方式展示 diff。你也可以看那个。但我发现看 Hunk 的时候我只能看到变更集。 There's also this other tool I've been playing with a bit called Hunk, which just produces, uh, diffs in a really nice way. So you could look at that, too. But I find that when I look at Hunk, I only see the change set. 100:41 而我审查智能体产出的时候，经常想看周围的上下文。哦对，它改了这个文件，但另一个没被碰过的文件里有什么？是不是其实也该改？ And when I'm reviewing output from an agent, I often wanna see the surrounding context. Oh yeah, so it changed this file, but actually what do we have in this other file that wasn't touched but maybe should've been touched? 100:51 所以这就是为什么我还是喜欢用 Neovim 来做这件事。呃，不过当然，这一切都发生在 Omarchy 里，也就是都发生在 Linux 上。而这正是智能体带来的另一个重大突破。 So that's why I still like Neovim as a way of doing it. Um, but it's all happening by the way, of course, in Omachi. So it's all happening on Linux. And this was the other major breakthrough with agents. 101:03 智能体热爱 Unix 哲学。 Agents love the Unix philosophy. 101:07 它喜欢可以通过命令行调用的独立工具，而在主流操作系统里——我这里数三个—— It loves individual tools that it can invoke through the command line, and there is no operating system on earth of the majors, I'm counting three here, 101:17 Mac、Windows、Linux，没有一个能像 Linux 那样完美契合这种机制。Linux 里的一切要么是配置文件，要么是 CLI 工具。 Mac, Windows, Linux, that works as well with that mechanism as Linux. Everything is in Linux is either a config file or a CLI tool. 101:30 而这在五分钟前还是它最大的缺点。这正是人们不喜欢 Linux 的原因：全是配置文件和 CLI 工具。 Now, that was its main drawback five minutes ago. This was the reason people didn't like Linux. It's like it's all config files and CLI tools. 101:38 嗯哼。 Mm-hmm. 101:39 宇宙跟我们开了个多大的玩笑：Linux 五分钟前的缺点，现在成了它最大的卖点。 What great irony that the universe has played it upon us that now the drawbacks of Linux five minutes ago are now its major selling points. 101:50 这就是我意识到的其中一件事，因为四个月前我被迫用 Mac 过了一个周末。 This was one of them 'cause I thought I was stuck for a weekend with a Mac, uh, four months ago. 101:58 我本以为我要去的地方有一台 Linux 机器，所以没带笔记本，结果到了才发现只有一台 Mac Mini。 I thought I had a computer at the place I was going that was a Linux machine, so I didn't bring my laptop and I I found out when I arrived I only had a Mac Mini. 102:08 所以我打算将错就错，呃，用我一直在 Linux 上琢磨的那些思路来配置这台 Mac。其实现在也能做不少事。 So I was gonna make the best of a bad situation here and, um, set up my Mac in some of the ways I've been thinking about with, um, with Linux. And you can actually do a lot now. 102:20 Homebrew 已经变得相当好了。Homebrew 是 Mac 缺失的包管理器，在简化装机这件事上，没人比它更推动 Mac 前进。 Homebrew has gotten really quite good. Homebrew is the missing package manager for the Mac, and no one has done more to move the Mac forward in terms of ease of setup. 102:31 但哪怕是像 Raycast 这么简单的东西——我不知道你用没用过。 But still something as simple as Raycast, I don't know if you've used that. 102:34 嗯哼。 Mm-hmm. 102:34 就是那个—— That's the- 102:35 用过。 Yep. 102:37 你没有一个可以直接访问的配置文件。你得进 GUI，导出一个文件，然后把这个文件——不知道——装进你那该死的背包里的 U 盘上—— There's no config file that you can just access. You have to go into the GUI, export a file, then take that file, I don't know, in your freaking backpack on a USB key- 102:47 嗯哼。 Mm-hmm 102:48 ……然后才能在别的地方导入。你没法自动化整台机器的配置。你完全没法自动化 Mac 默认快捷键的配置。 ... and then you can import it somewhere else. You cannot automate the entire setup of your machine. You can't automate at all the configuration of Mac's default key bindings. 102:59 嗯哼。 Mm-hmm. 102:59 那只能手动操作，像山顶洞人一样用鼠标点来点去来配置你的机器。 That has to be a manual process where you're clicking with a mouse like a caveman to set up your machine. 103:05 我是说，总有变通的办法。 I mean, there's ways around it. 103:07 没有好办法。我找过，我很努力地试过。 Not good ones. I looked. I tried hard. 103:10 那说说我的经历吧。显然我是个 Linux 用户，但因为 Adobe Premiere 这些 Adobe 产品，我也是个 Windows 用户，所以我经常会用 WSL，也就是 Windows Subsystem for Linux。 So here's, here's my, been my journey. Obviously I'm a Linux person, but because of Adobe Premiere, Adobe products, I'm also a Windows person, so oftentimes I would use WSL, Windows Subsystems for Linux. 103:23 对。 Yes. 103:23 也就是 Windows 里的 Linux，但在智能体时代这不是一种好的 Linux，因为它只是—— So Linux inside Windows which is, uh, in the agentic era is not a good kind of Linux 'cause it's just- 103:28 它是个沙箱。 It's a sandbox 103:29 ……它是个沙箱，而你希望 Linux 是被彻底放开的，什么都能做。 ... it's a sandbox and you want the Linux to be unleashed to be able to do everything. 103:33 没错。 Correct. 103:34 所以我现在开始醒悟，比如 Ghostty，我——你知道，我是个键盘党。配置文件在哪儿？ And so I have now woken up to things like Great Cat and have w- w- you know, I'm a keyboard person. Where's the config file? 103:41 我能不能，呃，我能不能基本上让一切都自动化：一个智能体帮我把所有东西都配置好、保存配置，让我可以跨系统复制，一切自动化？ Can I, can I, um, can I basically do everything where an agent can, can set everything up for me and save the configuration so I can replicate it across systems and everything's automated? 103:53 我不得不问很多这类问题，而很多都还没有答案。 And so I had to ask a lot of those questions and a lot of them are missing. 103:56 它们没有好答案。 They don't have good answers. 103:57 但很多时候你其实可以直接自己写这个应用。 But oftentimes you could actually just build the app yourself. 104:01 不过是有一些 hack 的。 But there are hacks. 104:02 对，是有 hack。 Yeah, there are hacks. 104:03 是有 hack。 There are hacks. 104:03 是有 hack。 There are hacks. 104:03 你不必这样过日子，Lex。有更好的……其实，这正好是个好时机。 You don't have to live like this, Lex. There's a better... Actually, this is a great moment. 104:08 在你有视频剪辑需求之前。 Until you have a video editor. 104:11 所以这是个好时机。 So this is a good moment. 104:12 对。 Yep. 104:13 呃，我听说那天是你生日。 Um, I heard it was your birthday. 104:16 对。 Yeah. 104:16 所以，呃，我带了个小礼物。 So, uh, I brought a little gift. 104:18 不会吧。 No. 104:19 不管你愿不愿意，我们都要把你拉上智能体操作系统这艘船。 We are going to get you onto the agentic operating system whether you like it or not. 104:23 太棒了。 This is awesome. 104:24 我，呃，我问了我的朋友—— And I, uh, I asked my friends- 104:26 嗯。 Yeah 104:26 ……Dell 的朋友，看他们是不是有一台机器，呃—— ... at Dell whether, uh, maybe they had a machine that um- 104:30 好。 Okay 104:30 ……我可以，呃，可以送你当生日礼物。 ... I could, uh, I could get you for your birthday. 104:33 哦，不会吧。 Oh, no. 104:33 这就是我自己用的那台机器—— So here's the machine I use- 104:35 嗯。 Yeah 104:35 ……也就是—— ... which is- 104:36 嗯。 Yeah 104:36 ……呃，和我用的一样，但现在它是你的了。这是一台—— ... well, the same one I use, but it's now yours. It's a- 104:40 不会吧。 No 104:40 ……Dell XPS 14……已经装好了 Omarchy 4—— ... Dell XPS 14- ... already set up with Omachi 4- 104:45 哦，哇。 Oh, wow 104:45 ……随时可以让你配置。而且，呃，等你意识到你的所有……对了，我们几乎应该计个时。你马上就能跑起来。 ... ready to be configured for you. And, um, once you realize that all your... By the way, we should almost time this. You can be up and running. 104:55 你只需要回答大概五个问题，一分钟搞定。你想的话甚至可以现场直播做。 You have to a- answer, like, five questions in about one minute. You could literally do it live if you wanted to. 105:00 哦，Omarchy，DHH 做的漂亮、现代、有态度的 Linux。好，开搞。 Oh, Omachi, beautiful, modern opinion Linux by DHH. Yeah, let's do it. 105:05 你边填我边说。那么—— Fill it out while I talk. So- 105:07 好。 Cool 105:07 ……正如你刚才意识到的，Mac 上的那种痛苦，正是我那个周末的切身体会。 ... exactly what you realized, that pain on the Mac was the exposure I got for about a weekend. 105:14 我很努力地试过，也找到了一些不错的变通方案，最主要的是 Homebrew 现在支持安装 GUI 应用了，因为回想不久之前还觉得好笑：人们需要 Zoom 的时候， And I tried really hard and I found some decent workarounds, the main one being that Homebrew now has tabs for GUI apps because it's hilarious to think upon the recent past where people would go to websites like I need Zoom. 105:30 他们会打开 zoom.com，找到下载链接，下载一个 DMG…… They would enter zoom.com, find the download link, download a DMG- ... 105:36 双击它，看到一个奇怪的手动交互，让他们把一个图标从屏幕的一个地方拖到另一个地方，而那就是安装应用的方式。 double-click on that, see a weird manual interaction telling them to drag a icon from one place of the screen to the other, and that was how you installed apps. 105:48 然后你一遍又一遍地重复这个过程，直到装齐所有想要的应用。如果你要配置一台新机器、有十二个应用要装，你就得像个彻头彻尾的山顶洞人一样去十二个不同的网站下载。 And you just do that over and over again until you had all the apps that you wanted. And if you were setting up a new machine and you had like 12 apps, you'd go to 12 different websites to download them like a total caveman. 106:01 Homebrew 解决了其中很大一部分，但没有全部解决，而且在——这方面什么也没解决—— N- homebrew has solved a lot of that, but it has not solved all of it, and has solved nothing in terms of the- 106:08 抱歉打断一下，但你看这速度多快。 Sorry to interrupt, but look how fast this is going 106:10 ……哦，大概再……四十秒就好了。 ... Oh, it's gonna be done in- ... in about 40 seconds. 106:15 然后你就进入 Omarchy 了。 And then you're gonna be inside of Omachi. 106:17 Super 就是你键盘上的 Windows 键或 Command 键。哪一个是 Super 键？ Super is the Windows or Command key on your keyboard. So which one is the Super key? 106:21 对，就是那个——Windows 键。哦，不对不对，你这是特别版。 Yep, that's the- the Windows key. Oh, actually, no, no, you have a special edition. 106:25 对。 Yeah. 106:25 全世界大概只有六台这样的电脑。 There are only about six of these computers in the world. 106:27 哦，有一个 Super 键。 Oh, there's a Super key. 106:29 上面写着 Super，再看 Copilot 键原本的位置，在键盘另一端，那是 Omarchy 的 logo。 It says Super, and look at where the Co-pilot key is, on the other end of the keyboard. That's the Omachi logo. 106:35 漂亮。好，酷。 Nice. Okay, cool. 106:36 这是，呃，Dell 的 Spencer Bull，我和他合作了很多这方面的工作。他们非常棒。 This was, um, Spencer Bull at Dell, who I've been working with on a lot of this stuff. They've been incredible. 106:42 他们一度放弃了 XPS 这个品牌。 They, for a moment, let go of the XPS brand. 106:46 是的，去年。 Yes, they did. Last year. 106:48 我当时就想：“这太蠢了。XPS 多好啊。” And I was- I was like, "This is dumb. This is... XPS is awesome." 106:51 蠢的不只是去掉了 XPS 品牌，他们还去掉了 escape 键，在顶部换成了电容触控键。 It was not just dumb that it got rid of the XPS branding. They also got rid of the escape key and put in capacitive touch buttons at the top. 106:59 就是 Apple 大概三年前意识到很糟糕的那个错误，而且所有人都因此骂他们。 The mistake that Apple realized, like, three years ago was terrible, and everyone hated them for them. 107:04 对。 Yeah. 107:04 但现在他们有一支非常棒的团队，而这其实是我第一台自用的 Dell 笔记本。 But now they have an incredible team, and this was actually my first Dell laptop for myself. 107:12 嗯哼。 Mm-hmm. 107:12 因为我们的服务器一直在用 Dell。我们的 Cloud Exit 就是靠 Dell 支撑的。我想我们从两千年代中期就开始用 Dell 了。 Because I've been using Dell on our servers. Dell was what powered our Cloud Exit. We've been running Dell since the mid 2000s, I think. 107:20 但我从不用他们的笔记本，因为和 MacBook 比，它们对我来说就是不够格。 But I never ran their laptops because compared to a MacBook, they just weren't up to the task for me. 107:27 2016 年——不对，2026 年的 XPS，搭载 Intel Panther Lake 芯片，是第一批我个人真正愿意用的 Dell。 The XPSs from 2016, or- or from 2026, with those Intel Panther Lake chips, are the first Dells that I would personally actually use. 107:37 有意思的是，我最初对 Apple 生气的时候就想：“你知道吗？我干脆买台 Dell 算了。”反正我们已经在用它们了。然后我买了台 Dell，结果：“兄弟，我不行，我真的不行。 And it's funny because when I got mad at Apple, originally, I thought, like, "Do you know what? I should just get a Dell," because we already used them and whatever, and I got a Dell, and I was like, "Man, I can't do it. I can't do it. 107:49 就是——就是不够好。”结果你猜怎么着，今年他们把一切都修好了。 It's not- it's not good enough." And then lo and behold, this year, they just fixed everything. 107:56 嗯。 Mm. 107:56 他们修好了续航，修好了性能—— They fixed the battery life, they fixed the performance- 107:59 嗯哼。 Mm-hmm 107:59 ……修好了兼容性，修好了重量，修好了屏幕。这是一块 tandem OLED 屏—— ... they fixed the compatibility, they fixed the weight, they fixed the screen. That's a tandem OLED screen- 108:05 对，太不可思议了。 Yeah, that's incredible 108:05 ……这意味着它的技术实际上比你在那该死的 MacBook 上能拿到的还要好。它比 MacBook 更轻。 ... which means it's actually better tech than what you'll get in a freaking MacBook. It's lighter than a MacBook. 108:13 而且芯片终于有竞争力了，因为 Intel 花了足足五年开发 18A 制程节点。花了很长时间，但现在它来了，而且强得离谱。 And the chip is finally competitive because Intel spent a good five years developing the 18A process node. It took a long time, but now it's here, and it's freaking incredible. 108:25 嗯哼。 Mm-hmm. 108:25 我们终于有了能和 Apple M 芯片竞争的东西—— We finally have competition to Apple M chips- 108:29 嗯哼。 Mm-hmm 108:29 ……续航一样长，性能也有竞争力。 ... that can go as long as on battery and can have competitive performance. 108:34 而且这机器有一个很漂亮的默认主题，然后，呃，这是什么？Super 加空格打开—— And there's a beautiful default theme to this thing, and then, uh, what is this? Super Space opens up- 108:39 其实你可以直接按那个 Omarchy 键，它会打开菜单。然后你从那里控制一切。 Actually, you can just push the, um, Omachi key. It opens up the menu. So now you control everything from there. 108:43 哦，和 Super 一样。好。 Oh, same thing as the Super. Okay. 108:44 对。 Yeah. 108:45 酷。 Cool. 108:45 和 Super 加空格一样。这对我来说是个巨大的变化：我们终于有了不需要为它找借口的 Linux 机器。 Same thing as Super Space. So this has been a huge change for me, that we finally have Linux machines that you don't have to make excuses for. 108:57 嗯哼。 Mm-hmm. 108:57 因为我不介意找借口，借口其实不过是取舍的另一种说法。 Because I don't mind making excuses because excuses are really just another word for trade-offs. 109:02 嗯。 Mm. 109:02 我很喜欢那台 Framework 13，我想上次聊过。我现在依然喜欢他们，他们依然在做好的电脑。但我觉得对很多人来说，他们不想要 DIY 体验。 I love that Framework 13 I think I talked about last time. I still like them. They still make good computers. But I think for a lot of people, they didn't want a DUI experience. 109:14 他们想要的是开箱即用、在做工各方面都感觉和 Mac 一个水准的东西——而现在我们终于有了。生日快乐。 They wanted something that just out of the box felt like it was on par in build quality and whatever with, uh, with Macs, and- and now we finally have it. So happy birthday. 109:27 太感谢了。DHH 送的 Omarchy Dell 笔记本是最好的生日礼物。这太棒了。 Thank you so much. An Omachi Dell laptop from DH is the best birthday present. This is awesome. 109:34 顺便看看屏保。 Check the screen saver, by the way. 109:36 嗯，有点惊人。我看到了。 Yeah, it's kind of amazing. I'm seeing it. 109:38 在那块 tandem OLED 屏上—— On that tandem OLED screen- 109:41 太美了。 It's gorgeous 109:41 ……简直 awesome 到让人不敢相信。 ... it is absolutely mind-bogglingly awesome. 109:46 你看。 Look at that. 109:49 整个东西有一种复古的感觉。 There's like a retro look to the whole thing. 109:50 完全正确。 Totally. 109:50 像一台打字机。 It's like a typewriter. 109:51 就像在九十年代运行一个 BBS 系统……那正是我十四岁时干的事。 It's like running a bulletin board system in the '90s- ... which is exactly what I did when I was 14 years old. 109:58 所以对我来说，这是一次彻底的——复古，也是一种很棒的美学，把 Unix 里这种现代崛起的文化和——结合在一起。 So for me, it's a complete- throwback and an awesome aesthetic, combining this sense of modern rising culture in Unix with this- 110:09 嗯哼。 Mm-hmm 110:09 ……和这种九十年代的复古结合在一起。 ... throwback of, uh, of the '90s. 110:10 对，一个智能体优先的操作系统配上复古外观。 Yeah, like a retro look for an agent first OS. 110:12 完全正确。 Exactly. 110:13 这太——太疯狂了。 It's cr- it's crazy. 110:13 而这又是为什么我当时那么兴奋—— Which again, pairs... This was why I was so excited that- 110:16 嗯。 Yeah 110:17 ……所有这些智能体 harness 都是以 TUI、以 CLI 的形式出现的—— ... all these agent harnesses arrived as TUI's, arrived as CLIs- 110:21 嗯。 Yeah 110:21 ……以终端的形式出现，因为突然间，新一代的程序员和普通用户都发现了终端的荣光。 ... has arrived as, as a terminal because suddenly, a whole new generation of both programmers and regular users discovered the glory of the terminal. 110:31 终端其实是我们能拥有的最令人印象深刻、最高效的用户界面之一。 The terminal is actually one of the most, uh, impressive and productive user interfaces we can have. 110:38 嗯哼。 Mm-hmm. 110:38 当然，在那之后我们造出了更容易上手的用户界面。 Now, we built user interfaces since then that were easier to learn. 110:43 鼠标是一项伟大的发明，它让不太懂电脑的人也能点点点、自己摸索，但终端是给想了解自己系统的人准备的， The mouse is a wonderful invention for letting someone who doesn't know a lot about computers just click around and find things out, but the terminal is for people who want to learn their systems, 110:53 给那些想学键盘命令的人。 who wanna learn the keyboard commands. 110:57 所以你觉得，呃，你觉得 TUI 会一直存在下去？ So you think, um, you think that TUI is gonna stick around? 110:59 百分之百。当然，它不会是唯一的东西。当然不是，因为这些实验室现在估值上万亿美元，他们需要触达几十亿人。 100%. Now, it's not gonna be the only thing. Of course it's not, because these labs now have trillion-dollar valuations, so they need to be able to reach billions of people. 111:09 嗯哼。 Mm-hmm. 111:09 而我不觉得我们能教会几十亿人 TUI 的精妙之处。 And I don't think we're gonna teach billions of people the intricacies of the TUI. 111:13 所以我们会有其他工具，而且已经有了其他工具。顺便说，这正是 Linux 上的一件美事——那个 ChatGPT Codex 工具。 So we're gonna have other tools, and we already have other tools, which is, by the way, one of the sweet things about, uh, Linux, the, um, ChatGPT Codex tool. 111:22 对，我看到了。 Yeah, I saw that. 111:23 那个发布刚出来两小时之内，我就把它打包好放进了 Omarchy。现在你只要去……其实你可以按那个 Omarchy 键。 Literally within two hours of that announcement dropping, I had it wrapped up and ready to go in Omachi. Already now, if you go to, uh... You can actually hit the, uh, Omachi key. 111:33 嗯哼。 Mm-hmm. 111:33 然后直接输入 ChatGPT，它应该会弹出安装选项。 And then just write ChatGPT. It should pop up as, like, install. 111:38 漂亮。 Nice. 111:38 然后直接按回车。 And then just hit return. 111:41 输入 Lex 的 sudo 密码。 Pseudo password for Lex. 111:42 对，你刚才设的那个。然后它就会安装，装完弹出来，你就有 ChatGPT 了。 Yeah, whatever you put in. And then it's just gonna install it, and it's gonna pop it up, and you have ChatGPT. 111:47 漂亮。 Nice. 111:48 很多应用都是这样，所有智能体 harness 都是按这种方式内置的。呃，我之前说的 Tailscale，你要想用的话也是这样，还有 Dropbox。这些东西都在里面的原因是我自己在用。 And that's how a lot of the apps are, and all the agentic harnesses are built in that way. Um, Tailscale, as I talked about, if you wanna hop on that, is that way, uh, Dropbox. And the reason all these things are in there is because I use them. 112:00 嗯哼。 Mm-hmm. 112:00 因为这套系统我是为自己打造的，我首先想的是：怎么为自己打造梦想中的电脑？如果我拥有了我的梦想电脑，我确信还有别人和我做着同样的梦。 It's because I built this system for me, and I thought, first and foremost, how do I build my dream computer just for me? And if I get my dream computer, I am assured there are other people who are gonna have the same dreams as I do. 112:13 对了，你对把安装时间压到一分钟以内这件事很执着，网上很多人问你为什么。所以为什么？ Now, you're obsessed about getting the installation to be under one minute, and a bunch of people asked you online why. So why? 112:21 为什么，你为什么这么执着于把安装时间压到 60 秒以内？ Why, why were you so obsessed to getting the installation to be less than 60 seconds? 112:27 首先，让我引用一下 Mitchell Hashimoto，他做了很棒的 Ghostty 终端——在 Omarchy 上也能用——他还是 HashiCorp 的创始人。 First, let me quote, uh, Michel Hashimoto, who created Ghostty terminal that's awesome and also available on Omachi and was also the founder of HashiCorp, 112:36 不久前他在 X 上发了个很棒的帖子，里面有句杀手级的话：“对卓越的追求不需要解释。” who had this great post on X not too long ago that had the killer line, "The pursuit of excellence deserves no explanation." 112:46 （笑） [chuckles] 112:47 想把一个东西做到尽可能好、尽可能快、尽可能美，这件事不需要任何理由。 Wanting to make something as good and as fast and as beautiful as possible has no need for justification. 112:58 嗯哼。 Mm-hmm. 112:58 我们就应该想要漂亮、快速、用起来愉悦的系统。关于安装时间，有意思的是它一开始并不是以一分钟为目标的。刚开始的时候我还挺谦虚的。 We should want beautiful, fast systems that are a delight to use. Now, the funny thing with the install time is it didn't actually start with a one-minute mission. I was being modest when it got going. 113:11 我当时想，如果我能在十五分钟内配置好整个系统，那相比以前已经是巨大的进步了，我就会非常满足。 I thought if I could set up my entire system in 15 minutes, it would already be such a dramatic improvement over what came before it that I would be perfectly happy. 113:20 而这源于一个事实：benchmark 已经偏得太离谱了。你说过 Commodore 64 是你的第一台电脑，对吧？ And it came off the fact that the benchmarking had been skewed so out of whack. You talked about the Commodore 64 being your first computer, right? 113:27 嗯哼。 Mm-hmm. 113:28 当你拆开 Commodore 64 的包装，按下开机键，我记得不到一秒钟，BASIC 解释器就准备好接受你的命令了。真的是几乎瞬间。 When you unwrap that Commodore 64 and you hit the on button, in I think about less than one second, the basic interpreter was ready to accept your command. It was literally almost instant. 113:40 几乎没有启动时间。你知道我配置一台新买的 Mac 花了多久吗——买它其实只为了一个我还没在 Linux 上凭感觉编程搞定的应用，Adobe Lightroom？ There was virtually no boot time. Do you know how long it took me to set up a new Mac that I had bought for actually the one app that I still haven't vibe coded my way into on Linux, which is Adobe Lightroom? 113:52 嗯哼。 Mm-hmm. 113:53 全新电脑，开箱。我开始配置，有个软件更新。等我把系统更新到最新、准备好安装 Adobe Lightroom 的时候，过去了 42 分钟。 Brand-new computer, out of the box. I set it up. There's a software update. Before I was ready to install Adobe Lightroom that I had it fully up to date, it took 42 minutes. 114:08 这太离谱了，太荒谬了。这对所有喜欢电脑的人都是一种侮辱：你拿到一台新电脑，在能用之前得先花 42 分钟做更新—— That's grotesque. That's absurd. That's an insult to everyone who likes computers, that you get a new computer and before you can use it, you have to spend 42 minutes doing updates- ... 114:20 ……更新的还是出厂时就预装好的东西。这简直荒唐，而且更疯狂的是这还不是最糟的。 on something that was already installed on it out of the box from the factory. That is just preposterous, and it's crazy that that's not even the worst thing. 114:29 我还买了一台 PC，呃，因为所有这些 Linux 机器都是以 PC 的形态来的，对吧？ I also got a PC, um, because all these Linux machines, they come as PCs, right? 114:34 对。 Right. 114:34 而且都预装了 Windows。所以我买了一台，就在三周前。全新机器，Intel Panther Lake，全套配置。 And they have Windows pre-installed. So I got one, this was just three weeks ago. Brand-new machine, Intel Panther Lake, the whole shebang. 114:44 从拆包装到能用，一小时三十五分钟。你——你觉得自己进 Omarchy 花了多久？ An hour and 35 minutes from unwrapping the thing until it's ready to use. H- how long did you take before you think you were in Omachi? 114:53 我是说，肯定不到一分钟。 I mean, it's less than a minute for sure. 114:55 一分钟。 A minute. 114:56 对。 Yeah. 114:57 我们可以这样造电脑。技术已经有了。从1981年 Commodore 64 刚出来那会儿，我们就已经有这个技术了。 We can build computers like this. We have the technology. We've had the technology since like 1981 when the Commodore 64 first came out. 115:05 如今我们的电脑快了一万亿倍，可你还得花42分钟、甚至一个半小时去设置它才能用，这件事让我恼火到极点，以至于它成了一种激情。 The fact that we now have a computer that's a trillion times faster and you have to spend 42 minutes or an hour and a half setting it up before it's ready to use is so infuriating to me that it became a passion. 115:17 说来也好笑，刚才我们休息的时候，我正跟 Ryan Hughes 聊天，就是我这边的同谋。 And it's funny, just while we were having a little break here, I was talking to, to Ryan Hughes, my co-conspirator here. 115:23 我们现在觉得，已经摸到了一条路，可以做出针对特定电脑的预制 turbo 镜像，因为这件事需要一些硬件层面的知识，比如 Dell XPS。 We now have, we think, a line where we can create these special turbo images that are pre-made for specific computers because it relies on some knowledge of the hardware, like the Dell XPS. 115:34 我们要做一个 Dell XPS 的 turbo 镜像，大概12秒就能装完。一整套 Linux 系统—— We're gonna make a Dell XPS turbo image that can install in about 12 seconds. A full Linux system- 115:43 嗯。 Yeah 115:43 ……大约12秒就装完。 ... installed in about 12 seconds. 115:44 太不可思议了。 That's incredible. 115:45 正是这种执念，让我用电脑时获得了极大的满足感——我们不接受事物本来的样子。 That obsession is what gives me this great satisfaction with using computers, that we do not accept things the way they are. 115:54 这也是为什么，我如今身处的这个被智能体加速的现实，让我如此振奋。 And this is why this agent-accelerated reality that I now inhabit is so invigorating. 116:01 嗯哼。 Mm-hmm. 116:01 因为我可以看着这些约束，然后说："你知道吗？我——"我本来可以在五分钟那儿就停下的，那其实已经够快了。这就是 X 上那场争论，对吧？ Because I can look at these kinds of constraints and go, "Do you know what? Um, I-" I could have stopped when we were at five minutes. That would already have been fast enough. This was the argument on X, right? 116:12 嗯哼。 Mm-hmm. 116:12 你干嘛还要继续？五分钟已经够快了。好吧，也许对你够快，但对我来说，12秒有意思多了。而且我一直在想，要试着从第一性原理出发去想问题。就是这样。 Why do you keep going? Five minutes is fast enough. Okay, maybe for you, but for me, 12 seconds is a lot more fun, and I always think, try to think these first principles. Okay. 116:25 在那台机器上，其实再按一下 Omarchy 的那个键，然后选 speed。你会看到一个列表弹出来，其中有一项就是 disk speed 之类的。 In that machine, actually hit the Omachi, uh, key again, and then do speed. You'll see a list come up, and one of them is, like, disk speed. 116:35 Speed test，network speed，disk speed。 Speed test, network speed, disk speed. 116:37 Disk speed。 Disk speed. 116:38 对。 Yeah. 116:38 选那个，然后运行。这是在测试内置 NVMe 硬盘的性能，你得到的结果是多少？ Do, do that one, and then run it. So this is testing the performance of the built-in NVMe drive, and what are you getting? 116:45 每秒 7 GB。 Seven GB a second. 116:47 也就是每秒七 GB。这块盘就是这么快。那我们为什么不能装一个 Linux 发行版？一秒钟，Omarchy 发行版是 5.8 GB。 So seven gigabytes a second. That's how fast that drive is. Why can we not install a Linux distribution? In one second, the Omachi distribution is 5.8 gigabytes. 116:57 嗯哼。 Mm-hmm. 116:58 如果我们能跑到每秒七 GB…… If we can run at seven gig... 117:00 当然，实际上你做不到，因为 U 盘比那慢得多。但你应该去逼近我们面对的底层物理和协议的极限，那才是你的目标。 Now, of course, you can't actually do that because the USB key is a lot slower than that, but you should be able to maximize the underlying physics and protocols that we're dealing with, and that's your goal. 117:13 我们得一路压回到那个极限上去。 We gotta get all the way back down to that. 117:15 对。我很喜欢你在做一个 Linux 发行版—— Yeah. I love that you're doing a Linux distribution- ... 117:18 因为，呃，我是说，你本来可以做任何东西，但你在做的是一个 Linux 发行版，这意味着你在重新思考电脑到底是什么，因为操作系统是最根本的，是—— 'cause, uh, I mean, you could be do- you, you could be building anything, but the fact that you're building a Linux distribution means you're rethinking about what a computer is 'cause a operating system is the fundamental, like, the foundation of- 117:32 对。 Yes 117:32 ……是人与电脑交互的地基。 ... how you interact with a computer. 117:33 对，而且它已经停滞太久了。 Yes, and it's been so stale. 117:34 对，确实是。 Yeah, it has. 117:35 今天的 macOS 和十年前的 macOS 几乎没什么区别。 The macOS we have today is scarcely different from the macOS we had 10 years ago. 117:41 事实上，在很多方面它更糟了，因为 Apple 在不断收紧对用户的控制。他们不想让——他们安装未经 Apple 批准的软件—— In fact, it's worse in many ways because Apple continues to chokehold the control on users. They don't want to- them to install software that they haven't already said good- 117:53 嗯。 Mm 117:53 ……就不行。他们不想让用户改快捷键，除非手动去弄。他们也不想去改那个切换工作区时500毫秒的动画。 ... to go for. They don't want users to reconfigure their hotkeys without doing it manually. They didn't want to con- uh, change the 500-millisecond animation switching workspaces. 118:05 这是一台封闭得令人恼火的电脑。当然，平心而论，作为一台封闭的电脑，Apple 做得还算不错。但我不想要一台被锁死的电脑。我想拥有我自己的电脑。 It is an infuriatingly locked down computer. Now, to Apple's credit, it's a pretty good computer for being locked down, but I don't want a locked down computer. I wanna own my computer. 118:17 更进一步，我想改造我的电脑。而这正是为什么智能体时代需要一个新的操作系统。 Better yet, I wanna mutate my computer, and this is where the agentic age needs a new operating system. 118:24 嗯。 Mm. 118:24 当你能凭感觉编程做出脑子里想到的任何应用时，你也应该能凭感觉编程改造你的操作系统。 When you can vibe code whatever app comes to your mind, you should be able to vibe code your operating system. 118:30 你应该能改变任何东西：它的外观、它的工作方式、有哪些面板、没有哪些面板。而这需要 Linux，就这么简单。另外两家操作系统没有一个能做到这一点。 You should be able to change anything, how it looks, how it works, what panels are there, what are not there, and that requires Linux, as simple as that. There are no... None of the other two operating systems can deliver this. 118:41 但你能不能谈谈——我记得有人问过你，比如"你算老几，凭什么觉得自己能比 Ubuntu 做得更好？"或者"Ubuntu 已经够好了，你干嘛还要做？"你的回答是："我觉得我能做得更好。" But can you speak to the, I think somebody asked you, like, "Who the hell are you to think you could do better than Ubuntu?" for example. Or, or, like, "Why are you doing it when Ubuntu is already pretty good?" And your answer was, "I thought I could do better." 118:55 那你能不能谈谈——这是在 X 上的——你这种"狂妄"是从哪儿来的？你一个人，或者加上 Ryan，就觉得自己能做出比 Ubuntu、比 Arch、比这些都更好的东西？ So can you speak to, this was on X, can you speak to, like, where the craziness comes from, thinking that you as one person or maybe with Ryan, you could start building something that's going to be better than Ubuntu or Arch or any of this? 119:12 嗯，我得说，一开始并没有那么宏大的想法。我开始折腾 Linux 的时候，其实就是从 Ubuntu 起步的，只是在它上面做加法。 Well, I'll say that it didn't start quite as grandiose. When I got going with my Linux journey, I actually just started with Ubuntu, and I just built on top of it. 119:22 这是一个已经能做很多我想做的事情的系统，至少它是 Linux，那我能不能就在它上面搭，让它更现代、真正好看，并且装上我想用的程序？ Here's a system that already does a lot of the things I wanted to do, and at least it's Linux, so can I just build on top of it and make it more modern, make it actually look good, and have the programs I wanna use? 119:33 我大概做到了那个程度的某个版本，然后我发现，原来还可以再往下走七层。你越贴近硬件、越贴近内核、越贴近一个个具体的软件包，你发现的自由就越多。 I got, like, some version of that, and then I discovered that there's seven layers deeper you could go, and as this- the closer you get to the metal, the closer you get to the kernel, the closer you get to the individual packages and so forth, the more freedom you discover. 119:47 所以随着我发现的自由越来越多，我的野心也越来越大，然后我意识到，哦，意外吧意外吧，我对电脑应该怎么工作，其实有非常强的主见， So as I discovered more and more freedom, I got more and more ambitious, and I realized, oh, surprise, surprise, I have strong opinions about how computers should work, 119:56 它们应该长什么样、该装哪些程序、安装应该花42分钟还是45秒——顺便说一句，45秒是目前的世界纪录。 how they should look, what programs should be installed, and whether they should take 42 minutes to install or 45 seconds. That's the current world record, by the way. 120:09 所以，如果有人在听完这期之后装了 Omarchy Quattro，而且打破了45秒，截个图发给我，你就会成为新的世界纪录保持者。 So if anyone installs Amachi Quattro after this and can beat 45 seconds, take a screenshot, send it to me, and you'll, you'll be the new holder of the world record. 120:16 那，你是什么时候突破60秒大关的？ When, when, uh, how long ago did you cross the 60-second mark? 120:19 这事挺有意思的，因为它就像，呃，以前那个什么极限来着？就是四分钟一英里之类的—— It's funny because it's like, um, that, uh, the old limit on, what was it? It was a four-minute mile or something- 120:27 对，嗯哼。 Yeah. Uh-huh 120:27 ……1952年那会儿—— ... like that in 1952- 120:29 嗯哼。 Mm-hmm 120:30 ……人们都觉得人类不可能做到。然后有一个人做到了，结果接下来的几个月里，真的又有三个人做到了，对吧？ ... that people thought was impossible for humans to beat. Then one guy beats it, and literally within the next few months, three more guys do, right? 120:36 对，嗯哼。 Yeah. Mm-hmm. 120:37 我们当时就有这么个目标。我最初定的目标是两分钟。 So we had this goal. I originally had the goal it was two minutes. 120:40 嗯哼。 Mm-hmm. 120:40 然后我们突破了两分钟，我就想："既然两分钟能做到，为什么一分钟不行？"我记得我当时还想："那就太疯狂了。一分钟装完一个现代操作系统，不可能，毕竟别的系统都…… And then we beat two minutes, and I'm like, "Well, if we beat two minutes, why can't we beat one minute?" And I remember thinking like, "That'd be crazy. You can't install a modern operating system in one minute when they're all... 120:52 如果别的都要花42分钟，对吧？"比如 Apple，有那么多聪明人，也要花42分钟，那这应该就是天花板了吧。不不，这不是天花板，因为根本就没有限速。 If the other ones are taking 42 minutes, right?" Like, uh, Apple, full of a lot of smart people, like, it took 42 minutes, so that's gotta be the bar. No, no, it's not the bar because there's no speed limit. 121:03 你习以为常的那些东西，几乎在任何情况下，都不是现实里固定不变的支柱。 Like, none of the things you take for granted are fixed pillars of reality, in almost any case. 121:09 在你压到那块硬盘每秒七 GB 的传输极限之前，你就还没碰到极限，所以你就该继续往下推。当然，到这地步它本身就变成了一场游戏，本身也是一种满足。 Until you get down to the seven gigabytes per second that drive is able to transfer, you've not reached the limit yet, so you should just keep pushing. And of course, then it becomes a game in itself and a satisfaction itself. 121:22 每抠掉一秒钟都让我兴奋。当我们突破一分钟大关的时候——其实也就是，我想，三周前左右，可能也就两周前——我简直疯了。 Every second I could shave off was an excitement. And when we broke the one-minute barrier, which was really only broken, I think, three weeks ago or something like that, maybe even two weeks ago, I just got crazy. 121:33 然后我让一群智能体去跑各种自动研究循环，试验各种不同的理论，比如：哦，改这个东西会怎样，改那个东西会怎样？把这个砍掉会怎样？并行做这些事会怎样？ And then I got a swarm of agents to run all these auto research loops of trying all these different theories on like, oh, if you change this thing, if you change that thing, what if you cut this out? What if you do things in parallel? 121:44 如果趁用户输入用户名的时候，我们就开始把软件包预加载进内存，是不是就能更快地把它们装进去？这变成了一场好玩的游戏。我觉得产品开发就该这样想：它应该好玩。你应该…… What if while you were typing in your username, we start preloading packages into memory system, we can fire them in faster? And it becomes a fun game, and I think that's how you should think about product development. It should be fun. You should... 121:56 它应该带点"胡闹"的劲儿。应该去超越目标，把它做到比"所谓够用"好得多的程度。 It should be frivolous. It should be trying to overshoot the target and making it so much better than what it, quote-unquote, needs to be. 122:04 尤其是现在，因为你可以动得非常快，可以尝试非常多不同的方法，因为有了智能体。 Especially now, 'cause you can move so quickly and can try so many different approaches because of the a- the agents. 122:09 对。这让事情更令人满足了。不过我觉得这道理一直如此。你看史上最伟大的 Mercedes-Benz，我记得叫 W126，我看过一部关于它的很棒的纪录片。 Yes. That just makes it more satisfying, but I think this was always true. If you look at the greatest Mercedes-Benz of all time, I think it was called the W126. I saw this wonderful documentary on it. 122:20 呃，主导这个项目的人，此前多年一直在 Mercedes 的安全部门工作，突然之间他被任命负责整个 S 级项目，然后他就把所有东西都塞了进去。 Um, it was steered by this guy who for years had worked in Mercedes' department for safety, and suddenly he gets put in charge for the whole S Class project, and he just puts in all of it. 122:32 自动收紧安全带、大灯清洗装置等等，全都是那时候来的。就市场而言，这辆车造得实在太过剩了。 This is where you got the auto-retending seat belts and, uh, washers on the headlights and everything. This car was so overbuilt for the market. 122:42 嗯哼。 Mm-hmm. 122:42 简直不可思议。但你看着它就会想："我就想要这个。"就像你看一块机械表会想："它能下到100米深？"或者，呃，我记得， It was unbelievable. But you look at that and go like, "Well, I want that," in just the same way that you look at a mechanical watch and go, "It can go down to 100 meters?" Or, or I think, 122:52 呃，Rolex 有一款能下到4000米的——那个，Deep Sea，深潜的那款—— uh, Rolex has one that can go down to 4,000 meters- ... the, the Deep Ocean, the Deep Diver Ocean- 123:00 嗯哼。 Mm-hmm 123:00 ……不管它叫什么吧。 ... or whatever it's called. 123:01 全世界买那块表的人里，顶多三个人会真的把它用到极限。但你难道不想成为那种人吗——去支持那些想把人类往前推那么远的人？ No one outside of three people in the world who ever buy that watch will ever put it to the test, but don't you wanna be the kind of person who is a patron of people who wanna push the human race that much forward? 123:16 我想要那辆打破速度极限的最快的车。我想要那块能下到马里亚纳海沟的潜水表。我想要那个能在60秒内装完的操作系统。 I want the fastest car breaking the speed limits. I want the diver watch that can go down the Mariana Trench. I want the operating system that can install in less than 60 seconds. 123:26 所以从个人层面讲，我——我敬佩那些人。 So from the individual level, w- I, I admire those people. 123:30 我觉得每个人都应该为此努力。但我们也该提一下它带来的结果：对极致的追求，哪怕那些指标看起来很无聊，它可能带来的意想不到的结果，是其他各种卓越， I think everybody should strive for that, but also the consequences of that we should mention, like the pursuit of excellence, even when the metrics seem to be frivolous, the maybe unexpected consequences of that is other kinds of excellences, 123:41 是各种发现和创新，诸如此类，跟速度本身毫无关系。 discoveries, innovations, and all that kind of stuff that has nothing to do with the speed. 123:48 百分之百。 100%. 123:49 这就是发现。那个……有一个你追逐的指标作为催化剂、作为驱动创新的引擎，是很好的，因为你会不断发现新东西。这跟登月、去火星是同一个道理。 It's discovery. It's the, the... It's nice to have the metric that you're chasing be the catalyst, the engine that drives innovation because you'll just discover... It's the same reason you go to the moon, you go to Mars. 124:01 为什么要去？谁在乎呢？其实真正有价值的是，你会在化学推进方面发现很多东西。 It's like, why? Who cares? Well, actually, the real thing is you'll, you'll, you'll discover a lot about chemical propulsion. 124:09 你会发现很多关于人类如何在太空中长期生存的东西。也许你会发现全新的在太空做太阳能的方式，或者在太空建数据中心，或者随便什么。 You'll discover a lot about, uh, how humans can live for a long time in space. Maybe you'll discover whole new ways of doing solar energy in space or data centers in space or whatever the hell. 124:21 没错。 That's right. 124:22 各种各样的……也许你会遇到—— All kinds of... Maybe you'll meet a- 124:23 没错。 That's right 124:23 ……外星人。你永远不知道。不过，定一个目标然后去追，尤其当它是数字化的目标时，真的很棒。 ... aliens, but you, you, you never know, but, uh, setting a goal and chasing it, especially when it's numerical, is really nice. 124:32 而定一个宏大的目标就更好了。当你拒绝只做"务实"的事，而去追求略微超出" plausible（貌似可行）"的东西时，魔法就会发生。 And setting a grand goal is even better. When you refuse to just do what's practical and you strive for what's slightly beyond plausible, that's when the magic happens, 124:44 我们就该全力投入这种事。这也是原因之一……其实某种程度上，这个具体目标也是受了别人的启发。 and we should be going all in on that. This is one of the reasons... And in fact, to some degree, this specific goal was motivated by seeing others do this. 124:56 我是说，在这个话题里，Elon 是独一无二的那个人。那家伙的目标简直太离谱了，乍一看完全荒诞不经。 I mean, Elon is the singular individual in this conversation. The guy is so out there in his goals. They seem so preposterous at first glance. 125:09 我确实有好多眼都觉得它们荒诞。然后你发现，它们有时候真的会实现。 I've certainly considered them preposterous on many glances. And then you realize that they come true sometimes. 125:19 也许不总是按时实现，那又怎样？那又怎样？我们是没按时等到自动驾驶汽车，2017年那会儿我还狠狠嘲讽过 Elon 一通，但现在它就在这儿了。 Maybe not always on schedule, and so what? So what? So we didn't get the self-driving car on schedule. I gave Elon a lot of shit for that back in '17, but now it's here. 125:31 除了那个在2017年就相信它可能的人，还有谁能把我们带到今天？他对现实如此没有耐心，以至于那股纯粹的蛮力，就要把现实硬推到他想要的位置上。 And who would've gotten us here except for the person who thought it was possible in '17, who is so impatient with reality that the sheer force is just gonna propel that reality to the place he wants it to be? 125:48 我觉得这特别鼓舞人心。我们需要有人不断移动这样的目标柱。我们需要有人去跑那个四分钟一英里，好让我们其他人意识到："好吧，原来我也能跑得更快。" I think that's deeply inspiring, and we need these kinds of, uh, goalposts to be moved. We need someone to run the four-minute mile so that the rest of us realize, "Well, okay, but I guess I can run it quicker." 126:00 这就是我努力的方向，我在 Omarchy 上也是真的这么努力的，因为——那样活着就是更有意思，有更高的追求和野心。不过我得说，在这一点上我其实改变了一些看法。 That's what I try to strive for, and I really strive for it with Amachi because- It's just more fun to live that way, to have high aspirations and ambitions. And I will say I've actually changed my mind on that a little bit. 126:12 我在另一本书里花了挺多篇幅讲，很多人其实并不需要那样的野心。而且我也……这话也没错。 I spent quite a lot of time speaking the other book that there are a lot of people who don't actually need those ambitions. And I also... That's also true. 126:24 我们可不想要80亿个 Elon 满世界跑。我觉得那会相当疯狂。 We don't want, uh, 8 billion Elons running around the world. Would be rather crazy, I think. 126:29 但我渐渐意识到，对创业者来说，有一个足够大的、有分量的目标，是健康的。 But I've come to realize that for entrepreneurs, it's healthy to have suitable chunky goals. 126:38 开意大利餐厅这件事，不能就只是"开个餐厅"。它可以是我只开一家意大利餐厅，但我要做出全世界最棒的披萨。 The idea of just running the Italian restaurant can't just be that. It should be maybe I just run the one Italian restaurant, but I make the best damn pizza possible. 126:50 我要做出最好吃的帕玛森料理。我们必须追求卓越，才能攒出继续走下去的动力。 I make the most delicious Parmesan. We have to strive for excellence to muster the motivation to keep going. 127:00 而且，呃，对，要敢于往大了想，比如你敢去造一个 Linux—— And, uh, yeah, to dare to, to think big like you could build a Linux- 127:05 对。 Yes 127:05 ……发行版。你还记得吗，为了达到这么快的速度，都做了哪些事？你提到了一些，都是智能体发现的。 ... distribution. Is there, um, some stuff you remember about what it took to get us to be so fast? You mentioned a few things that were... The agents were discovering. 127:15 其中一件事，就是把人类输入的延迟当作预加载的机会来利用。所以—— One of the things was to treat the lag of human input as an opportunity to preload. So- 127:23 嗯。 Hmm 127:23 ……你在设置一台 Omarchy 机器的时候，要回答五个问题—— ... there's five questions you answer when- 127:25 对。 Yeah 127:25 ……就这么个过程。 ... you set up an Omachi machine. 127:26 所以它是在后台干活。 So it's doing stuff in the background. 127:27 我们在后台干活。这一点我之前甚至都没考虑过，可它其实是教科书里最老的把戏了。 We're doing stuff in the background, and that was one of the things I didn't even consider, which is, I mean, the oldest trick in the book. 127:34 呃，长期以来，各种电子游戏都是这么做的：让玩家跟某个东西交互，然后你在后台做预加载之类的事。这是个老掉牙的技术，但我就是没想过把它用在安装器上。 Um, all sorts of video games for a long time have done this, where you let the player interact with something, and then you do preloading and, and so forth in the background. So it's an age-old technique, but I just hadn't thought about it for an installer. 127:45 我原以为，你得先问用户一串问题：密码、用户名、时区之类的，然后才开始干活。可实际上你完全可以预加载。 I thought first you ask the user some questions about their password, and their username, and their time zone, and so forth, and then you do the work, when in fact you could preload that. 127:55 然后剩下的就是一堆对安装器本身的各种调整，比如它做事的顺序。还有一些就是纯粹地瘦身。 Now, the other things were just a bunch of other tweaks to, um, how the installer itself was doing things in a certain order. Some of them was also just shrinking it. 128:06 上一个版本的 Omarchy 是 7.5 GB，而我们如今省下来的时间里，有很大一块其实就是解压缩那些压缩包的时间。 The last version of Omachi was 7.5 gigabytes, and a lot of the time we're down to now is literally decompressing compressed package files. 128:16 ISO 里的大头就是这个，安装时间的大头也是这个。所以把它瘦身下来——我们现在大概是 5.85 GB——相当可观的一块削减。 That's the bulk of what is on the ISO, and that's the bulk of the install time. So by shrinking it down to, I think we're at 5.85 gigabytes now, a pretty chunky reduction. 128:26 这几乎就是一比一地转化成了时间。而且我在那儿还做了些挺有意思的事。Omarchy 用的是 JetBrains 字体，这是一款—— It almost literally translated one-to-one, and I did some really interesting things there. So Omachi uses the JetBrains font, which is a- 128:34 嗯。 Hmm 128:34 ……很棒的字体，很漂亮，其实我第一眼并不喜欢它。但后来它成了唯一一款在 Mitchell Hashimoto 的 Ghostty 里能完美显示的字体。 ... awesome font, beautiful font, and actually a font I didn't like at first glance. But then it was the only font I could get to look perfect in Mitchell Azimoth's Ghosty. 128:45 Ghostty 不知道出于什么原因，就是不愿意按我习惯的方式渲染我心爱的 Bitstream Vera Sans。 Ghosty, for whatever reason, did not want to render my beloved Bitstream Vera Sans just the way I was used to it. 128:53 嗯哼。 Mm-hmm. 128:53 但它渲染 JetBrains 这款字体就很完美，所以我就换了。不过…… But it rendered, um, this JetBrains font perfectly, so I switched over. But it... 128:59 Arch 上 JetBrains 字体的标准软件包有 200 MB，因为它包含了所有这些变体、所有这些子字形什么的。然后我意识到：这些我们根本没用。 The standard package for JetBrains font on Arch is 200 megabytes because it includes all these variations and all these subtypes and whatever. And I realized, like, we're not using any of that. 129:12 我们只用等宽版，而且是那个 Nerd——呃—— We're using the monospace version in this one NERT, uh- 129:16 对。 Yep 129:16 ……打过补丁的 Nerd Font 版本。你知道它多大吗？我记得才 16 MB。 ... font patched edition. Do you know how mu- big that is? I think it's 16 megabytes. 129:21 嗯。 Hmm. 129:21 所以我就干脆新做了一个包，一个瘦身版—— So I simply just came up with a new package, the slim version- 129:25 嗯哼。 Mm-hmm 129:25 ……的 JetBrains 软件包。 ... of the JetBrains package. 129:26 漂亮。 Nice. 129:27 就这么一下，我省了 180 MB。我就这样一个个地过，做了好多次。NVIDIA 驱动我们也这么干了。 And there, right there, I saved 180 megabytes. I just went through and did that a bunch of times. We did that with the NVIDIA drivers too. 129:35 NVIDIA 驱动直接从——我记得是 Arch 仓库——拿来的时候，并没有用最极致的压缩格式，我记得叫 ZSTD，那是一种压缩起来非常慢的格式。 So the NVIDIA drivers, when they came straight off the, I think the Arch repo, were not compressed with the most extreme form of compression, which I think called ZSTD, which is a really slow form of compression. 129:47 生成压缩包很慢。但如果你是在构建一个软件包—— It's slow to create the archive. But if you're building a package- 129:50 嗯哼。 Mm-hmm 129:50 ……你就有的是时间，只要这能让 ISO 变小。光是那两个 NVIDIA 包，我记得我们就省了 200 MB。 ... you have all the time in the world if that means you can then shrink the ISO. So on just the two NVIDIA packages, I think we saved 200 megabytes. 129:58 嗯哼。 Mm-hmm. 130:00 这件事让我特别享受的地方在于，我把自己想象成了 McLaren 的汽车设计师。 What I loved about this was I imagined myself being a car designer at McLaren. 130:06 McLaren 现在造的是世界上最轻的一些超级跑车，他们对抠掉车上每一克重量这件事，执念到了极点。 So McLaren makes some of the lightest super sports cars in the world right now, and they are absolutely obsessed with shaving every damn last gram off their cars. 130:18 所以哪怕是一台新的 750S McLaren，也比竞争对手轻得多。 That's why even a new 750S McLaren is just way lighter than the competition. 130:25 他们用碳纤维单体壳，这就让他们胜过了，呃，比如 Ferrari——Ferrari 用的是铝。 They use a carbon fiber monocoque, and that helps them over, uh, like Ferrari, for example, does it- uses aluminum. 130:32 但他们就是对重量有这种执念。我看过一个采访，McLaren 的人跟一个记者在那儿一条条过："对，这一块我们省了，呃……370克——" But they also just have this obsession with weight, and I would see this interview with the McLaren folks talking to some journalist just going over, "Yeah, we saved, uh- ... 370 grams- 130:42 对。 Yeah 130:42 ……"就这一个部件。"我就想："等等，什么？这辆车总共也就1040公斤或者1380公斤什么的， ... off that one." And I just went, "Wait, what? There's someone on a car that weighs maybe 1,040 kilos or 1,380 or something. 130:53 他们居然在为370克较真？" They're s- they're worrying about 370 grams?" 130:56 对。 Yeah. 130:57 真是一群执念狂魔。 What obsessed maniacs. 130:59 嗯哼。 Mm-hmm. 130:59 我爱死他们了。 I love them. 131:00 对。 Yeah. 131:00 我就想要这种东西出现在我的生活里。就冲这一点，我都想要一台 McLaren。 I want that in my life. I want a McLaren- ... just for that reason. 131:05 所以当我在那儿一个 MB 一个 MB 地给软件包瘦身的时候，我就想：天呐，我这会儿还真有点像 McLaren 的汽车工程师，虽然—— So I thought in that moment as I s- I was shaving individual megabytes off the packages, man, I- I'm kinda like a McLaren car developer here, even though- 131:14 对。 Yeah 131:14 ……我是说，我们这活儿甚至还没干完呢，这些包还能再瘦。但这其实就牵扯到 Omarchy 跟传统 Linux 社区之间的另一个冲突：臃肿。 ... I mean, we haven't even completed that work. The package could shrink even more. But this is actually this other conflict that we've had with Omachi, and the traditional Linux community is bloat. 131:25 嗯哼。 Mm-hmm. 131:25 在某些圈子里，出厂预装软件是不合规矩的，因为"你算老几，凭什么替大家决定该用什么程序？" This idea that it's actually not kosher in certain circles to ship with pre-installed software because who are you to make decisions about what programs people should use? 131:37 而我就想说："哥们儿，名字里就写着呢。它叫 Omarchy，因为 Oma 这部分就是 omakase（お任せ）的缩写，字面意思就是'主厨之选'。我就是那个主厨。 And I'm like, "Dude, it's in the fucking name. It's called Omachi because the Oma part is short for omakase, which literally means chef's choice. I'm the chef. 131:47 选择权在我手上，而这一套就是我觉得很棒的应用合集。"顺便说一句，这里面还有视频编辑器。它出厂自带 OBS—— I'm making the choices, and this is the collection of applications I think is awesome," which by the way has a video editor. It ships with OBS- 131:56 嗯哼。 Mm-hmm 131:56 ……用来做录屏。它就在里面。 ... to do your recor- It's just in there. 131:58 对。 Yeah. 131:58 OBS 是预装好的。 OBS is pre-installed. 131:59 漂亮。 Nice. 131:59 Kdenlive 是一个视频编辑器，时间线编辑器，很棒。我所有视频都是用它剪的。然后我自己还做了一个剪片段的工具。你之前说到的那个同步的事。 Kdenlive is a video editor, timeline editor, which is great. I edit all my videos with that, and then I made my own clip editor. You talked about the thing with syncing up. 132:09 对。 Yeah. 132:09 我有个剪片段的需求，比如我录完播客什么的，需要一个发到 X 上的片段。 I had this thing with clips, so I would do a podcast or whatever. I need a clip for X. 132:13 嗯哼。 Mm-hmm. 132:13 而用 Kdenlive 或者别的什么—— And it's kind of a hassle to use Kdenlive or other- 132:15 嗯哼。 Mm-hmm 132:15 ……时间线编辑器干这个挺麻烦的，所以我就做了 Omacut。 ... timeline editors for that, so I made Omacut. 132:18 嗯哼。 Mm-hmm. 132:18 Omacut 也包含在里面。 Omacut is included, too. 132:19 Omacut，酷。 Omacut, cool. 132:19 它就是专门用来剪片段的，而且整个流程都可以用键盘驱动。你可以精确跳到想要的时间点。 It's just for making clips, and you can drive the whole thing with keyboard. So you can go to just the right time mark. 132:25 嗯哼。 Mm-hmm. 132:25 然后按 Control + Space，片段的起始线就移到那个位置。然后你跳到—— And then you can hit Control + Space. It moves the clip line to that line. Then you go to- 132:29 漂亮。 Nice 132:29 ……片段的结尾，按 Alt + Space，结束线就移过去了。 ... the end of the clip, and you hit Alt + Space, and it moves the end line. 132:33 漂亮。 Nice. 132:33 然后按 Control + S，就保存了。 And then you hit Control + S, and it saves. 132:35 漂亮。 Nice. 132:35 我现在用 Omacut 剪片段快得飞起，太爽了。而且系统里还带着全套软件：Neovim、Hyprland 那套、Tmux、终端。 I can make clips so damn fast now with Omacut. It's amazing. But it has all the software on it. It has Neovim on it. It has Herder on it. It has Tmux. It has a terminal. 132:46 有一堆主题，有一堆背景图，都是我能挑出来的最好的。这些都占点空间。 It has a bunch of themes. It has a bunch of, uh, background images, the best ones I could pick out. That takes up some space. 132:54 它们就应该占空间。因为这不该是一片荒芜之地，什么都得你自己从零重建。 It should because this is not supposed to be this just barren landscape where you have to reconstruct everything. 133:00 嗯哼。 Mm-hmm. 133:00 它应该是你一开箱就能立刻干活的系统。 This is supposed to be a productive system the minute you unpack it. 133:04 它是用什么写的？ What is it built in? 133:05 Linux 配置那部分其实主要是用 Bash 写的。 So it's actually mostly built in Bash for the Linux configuration part. 133:11 对。 Yeah. 133:11 Bash 干这个很合适。它，呃，就是为这个而生的，真的就是为系统管理而生的。当然它也有局限，你不该什么都用 Bash 写。 Bash is a great language for that. It is, um, made for that, really for system administration. It also has its limitations. You should not build everything in Bash. 133:21 智能体生成 Bash 的水平怎么样？挺好的？ How are agents with generating Bash? Pretty good? 133:23 哦，厉害极了。 Oh, amazing. 133:24 这样啊。 Okay. 133:24 除了一件事：每次抓到它犯这毛病，我都得敲它后脑勺，提醒它去看 agents.md 文件，因为我在里面写了一条指令：不要用提前退出。 Except for one thing, and I have to smack the agents over the back of the head every single time I catch it and remind it to look at the agents MD file because I will have an instruction in there not to make early exits. 133:37 嗯哼。 Mm-hmm. 133:37 智能体特别喜欢用前置条件判断，而不是完整展开的条件分支。 The agents love to have preconditions instead of fully expanded conditionals. 133:43 嗯哼。 Mm-hmm. 133:43 我喜欢的是那种"如果是这样，就做那个"的写法。 I like something that says, "If this, then do that." 133:47 嗯哼。 Mm-hmm. 133:47 智能体喜欢写成：前置条件，否则退出；再来一个前置条件，否则退出。 Agents like to have precondition or exit, another precondition or exit. 133:52 对。 Yeah. 133:53 然后一路掉到它真正要做的那件事上。我特别讨厌这种风格。 And then it falls down to the final thing it actually wants to do. I, I hate that style. 133:57 除了这一点，它们已经好到惊人、好到离谱了。我们上次聊的时候还说，我会让智能体给我一段 Bash，然后我自己敲一遍，但我其实记不住。 So except for that, i- it's gotten shockingly, dramatically, awesomely good, to the point where when we talked last time, we talked about how I would ask the agent for a Bash thing, and then I would type it in myself, and I kinda wouldn't remember. 134:11 后来我特意下了功夫，确保这些本事不会从我指头上流失掉，所以我真的去学了怎么自己写 Bash。结果现在我自己一行 Bash 都不写了。 And then I took the effort to make sure that all that competence wasn't draining out of my fingers, so I actually learned how to write Bash myself, and now I'm not writing any Bash myself. 134:19 我大概已经好几个月没亲手写过 Bash 了，因为智能体已经强到这种程度：你只要给它们一点风格上的指引，偶尔在复杂度上顶回去一下…… I have not written any Bash myself for probably a couple of months because the agents have just gotten so good, that if you give them the few pointers on style and occasionally put pushback on complexity... 134:29 这事我——我前几天还发了条推。让我震惊的是，经常发生这样的事：智能体做完一件事，说做完了，另一个智能体评审一遍， This is w- I, I tweeted about this a couple of days ago. It blows my mind how often an agent can do something, say it's done, have the work reviewed by another agent. 134:39 嗯哼。 Mm-hmm. 134:39 也说做完了。然后我看一眼："呃，我觉得这有点太复杂了。"它就会说："哦对，你说得对，我确实过度复杂化了。"然后咔咔砍掉一半。 They say it's done, and then I go like, "Eh, looks a little too complicated for me." And then it'll go like, "Oh, yeah, you're right. You're right. I totally overcomplicated this," and it'll cut it in half. 134:50 我就想："它一开始为什么不直接这么写？"然后有人提醒我：人类也一样。 I'm like, "Couldn't it just do this, like, from the outset?" And then someone else reminded me, humans are the same way. 134:57 你做了个东西，别人评审的时候说一句："呃，这看着有点复杂。"你往往也会说："呃，你说得对。" If you work on something and someone else reviews it and they just are like, "Eh, it looks a little complicated," quite often you're able to go like, "Eh, you're right." 135:05 你看，正是这些小迹象，让我觉得程序员仍然是必需的，那种直觉。因为我觉得，你那句话是怎么说的来着？"我对代码的全貌有一种感觉。" See, that, that just, th- those little indications t- to me show that the programmer is still needed, that intuition, 'cause I think, what, what did you say? I have a sense of the landscape of the code. 135:17 当我做 Omarchy 这样的东西时，我——我确实会看代码。我看它的形态。 When I build something like Omachi, y- I, I do look at the code. I look at the shape. 135:22 对。 Yeah. 135:23 我看它的比例。 And I look at the proportions. 135:24 对。 Yeah. 135:24 我已经不必事事亲自动手凿了。 Like, I don't have to chisel everything myself anymore. 135:27 我又开始自我陶醉了：前一秒我还是 McLaren 的——呃——汽车工程师，下一秒我就成了 Da Vinci，带着他整个画室的学生们—— This is, again, when I'm flattering myself for, not, uh, one moment I'm a McLaren, uh, audio, oh, uh, auto engineer, and the next moment I'm a Da Vinci working with his whole studio of students- 135:39 对。 Yeah 135:39 ……那些学生在细细雕琢大理石，对吧？然后 Da Vinci 走过来说一句："啊，不对，比例不太对。"这就是现在跟智能体协作的感觉：我可以像个编辑一样给反馈。 ... who are doing all the chiseling on the, the fine marble, right? And he goes like, "Ah, no, the proportions are not quite right." That's how it feels working with agents now, that I can give feedback like an editor. 135:50 嗯哼。 Mm-hmm. 135:50 就是——你知道吗？我不必每个字都自己写，但我看得出这段论证有没有立住。 That, do you know what? I'm not gonna write everything myself, but I can tell you when the argument doesn't land. 135:56 我看得出这个形态跟问题是否相称，或者它是不是就是太复杂了。而且令人震惊的是，这么干经常就是有效。 I can tell you when the shape isn't proportional to the problem or when it just feels like it's too complicated, and shockingly often it just works. 136:05 早年间那个玩笑是"别犯错"。我觉得现在不用再对智能体说这个了，因为现在的框架（harness）太好了，它们会跑自动化测试、拿到反馈。但你确实得告诉它们：做得简单点。 It's like the early days where the jokes was Don't make mistakes. I don't think you have to tell the agents that anymore because now the harnesses are so good, they run automated testing, they get that feedback. But you actually do have to tell them, make it simpler. 136:19 你跟智能体交互时用语音吗？还是说……你打字的占比多少？你—— Do you use voice at all for the interaction? Or, uh, are you... How much are you typing? How much are you- 136:23 我全程打字。其实 Omarchy 内置了 VoxType—— I type the whole thing. I actually, in Omachi's built-in VoxType- 136:27 对。 Yeah 136:27 ……呃，语音转写，其实做得相当好。只是我发现我的思维方式不是那样的。我自己都挺意外的：我不喜欢跟电脑说话。 ... um, audio transcription, and it's actually really good. I just find that I don't think that way. And it surprised my- self that I don't like to talk to the computer. 136:39 呃，在车里的时候我喜欢用语音，但坐在键盘前，我就想打字。而且我本来就喜欢打字。你在用语音吗？ Uh, I like when I'm in the car, but when I'm in front of the keyboard, I just wanna type. I also just love typing. Are you using the voice? 136:46 我是个内向的人，不喜欢社交—— I'm an introvert who doesn't like social interaction- 136:48 对。 Right 136:48 ……也不爱说话。而且—— ... or speaking. And- 136:50 我觉得我一部分也是这样。 I think that's part of mine too 136:53 但对我来说，不知道为什么，语音特别强大。我想给你看看这个……我随身带着这个小东西，叫 Plaud。我就把它夹在，你看，夹在我的 T 恤上。 But, like, for me, for some reason, voice is really powerful. I want to give this... So I, I run with this thing, it's called Plaud. So I, I just hook it up on my, you know, um, on my T-shirt here. 137:05 对。 Yeah. 137:05 按一下按钮，它就开始录音。就心流体验而言，这其实比坐在电脑前好得多。 You press a button, it starts recording. A- and this is actually more, in terms of flow, much better experience than behind a computer. 137:13 然后我会存下来——通常是一些智能体生成的问题，我想把它们彻底想透。 And I, uh, save, well, usually it's agent gen- generated problems that I wanna really talk through. 137:22 于是我就口述一个很长的提示词。里面很多东西都是意识流式的推理。 And so I prog- I write, I speak a long prompt. Like, and a lot of it, it's like stream of consciousness reasoning. 137:31 所以我有时候说着说着会改变对一个设计的想法，但这些全都在录音里。然后我让一个 LLM 来处理它。Plaud 那边，呃，显然是有 API 的。 So I- I'll sometimes just change my mind about a design, but all of that is in there. And then I have an LLM that processes that. W- with Plaud, y- there's, uh, obviously an API. 137:42 你可以把录音取出来、转写成文字、清理转写稿。因为你需要它……因为我会提到文件，提到各种技术概念，比如文件、函数—— You can grab it, transcribe it, clean up the transcript. 'Cause you, you want it to be... 'Cause I'm referring to files. I'm, I'm referring to technical concepts, like files and functions- 137:53 对。 Yes 137:53 ……之类的，你得确保转写、确保—— ... and so on, that you wanna make sure the transcription, the- 137:56 对。 Right 137:56 ……语音转文字能转对。所以你得有一个词典，处理文件时还得让它感知你的代码库。这样你就有了一个巨大的东西，常常是我用10、15、20分钟把整个设计讨论一遍。 ... the speech-to-text gets correctly. So you wanna have a dictionary, you want it to be aware of the code base when you process the files. And then you have this gigantic thing that's often 10, 15, 20 minute me discussing the full design. 138:09 我发现这个方式特别、特别强大，因为它不存在过度规约（overspecification）的问题，因为里面有大量意识流的思考， And I find that to be really, really powerful because it doesn't have the problem over s- of overspecification because of how much stream of consciousness thinking the, 138:21 系统对你想象的东西有非常丰富的上下文。这对设计的早期阶段特别有用。 the system has about what you're imagining. And this is really good for the early steps of a design. 138:29 比如说，如果我要从零开始构建 Omarchy，最初阶段我会口述，可能长达一个小时——我确实有过一小时的提示词——我一边跑步一边—— Like, if I wanted to build Omachi from scratch, the first stages, I t- I talk through, it could be an hour, there's been prompts that I had for an hour, and I'm just running and- 138:40 太厉害了。 That's amazing 138:40 ……一边说。我强烈推荐大家试试这种基于语音的口述—— ... and speaking. I highly recommend people try, uh, voice-based speaking- 138:48 这样啊—— So- 138:48 ……长时间的口述。 ... for prolonged periods of time. 138:51 我其实还没—— I haven't actually- 138:51 而不是非常具体的那种。 As opposed to very specific 138:52 ……我还没这么用过，对，我还没试过。也许这正是我用错了的地方。我一直只是打两句话、或者说两句话，然后就有点心烦，觉得我打字可能也一样快。 ... I haven't, th- yeah, I haven't used it that, and maybe that's how I'm using it wrong. I try to just type or speak two sentences, and then I get a little frustrated that I could have typed that just as fast maybe. 139:04 但我没试过这种连续说上20分钟的做法。 But I have not tried this whole thing of just talking continuously for 20 minutes. 139:09 对，尤其是你，你说话特别快。 Yeah, and especially you because you speak very fast. 139:12 对。嗯，我能想象那个效果。 Yeah. No, I could see it. 139:14 因为我之前一直拿它做回邮件之类的事，用到了我觉得它能帮上忙的程度——确实有点用——但它会犯个小错，我就得去纠正。 I think because the reason I've been, I've been trying to do it for things like replying to emails, and it got to a point where I thought it could help, and it sort of did, but then it would make a small mistake and I'd had to correct the mistake. 139:27 但如果你是对着一个 LLM 做意识流口述，它本来就不在乎那些细枝末节—— But if you're doing stream of conscious to an LLM who doesn't care about all this other stuff anyway- 139:32 对。 Yeah 139:32 ……那效果可能好得多。 ... that probably works way better. 139:34 有两件事你必须做。第一，拿到原始音频，转写。我用 ElevenLabs 做转写，他们大概是最好的语音转文字模型。但仍然会有错。 There's two things you have to do. One, okay, you have the raw audio, transcribes it. I use ElevenLab for transcription. They're probably the best speech-to-text model. There's still mistakes there. 139:44 所以你得有一个提示词，能访问一个词典，还能访问大量—— So you need to have a prompt that has a- access to a dictionary and to a lot of- 139:50 这个词典怎么构建？ How do you build the dictionary? 139:52 首先，有一个跟你相关的术语词典。然后让它感知你的代码——比如你有一个智能体，当你谈论代码时，它会扫你的整个代码库，找你常用的关键术语。 First of all, there's a dictionary of terms that are relevant to you. Then have it aware of a c- so you have an agent that looks through when you're talking about code Looks through your entire code base for key terms that you use often. 140:03 另外，当我们谈到消息类的东西，Gmail、WhatsApp 之类的，它得知道那些人名、那些你常被转错的常见词。 Also, when we're talking about messaging, Gmail, and WhatsApp, and so on, it's aware of all the people and common terms that you get wrong 140:11 你在这上面真是钻得深。 You've really gone deep in this. 140:13 对我来说，这是一个必须解决的问题—— That to me is a problem that needs to be solved- 140:14 对。 Yes 140:14 ……因为你用语音转文字时的问题之一，就是它有时候会出错。 ... because, uh, one of the issues you probably have speech-to-text is, uh, that it makes mistakes sometimes. 140:22 没错。那我就没法拿它口述邮件了。 Correct. Then I can't use it for dictating an email. 140:25 对。如果它从不出错，你—— Right. If it never makes mistakes, you- 140:28 那吸引力就大多了。 Then it would be more appealing. 140:29 对。而且其实你并不介意它不是完全实时的。你可以接受等五秒—— Yeah. And you are okay being not perfectly real time, actually. You, you, you're okay waiting five seconds- 140:35 对。 Right 140:35 ……对吧？我不——我不知道你的使用场景是什么。我可以接受等五秒—— ... potentially. I, I don't, I don't know what your use case is. I'm okay waiting five seconds- 140:39 嗯哼。 Mm-hmm 140:39 ……甚至10秒，等它把所有东西处理完。 ... 10 seconds for it to process everything. 140:43 你得凭感觉编程把这一整套做成一个应用，让这一切—— You gotta vibe code this into an app that does all of this such that- 140:45 对。 Right 140:46 ……就是能用。 ... it just works. 140:47 我记得我私下跟你说过，让我很意外的是，翻译这件事至今没有被做到完美。 I think I mentioned to you offline that it's very surprising to me, uh, that, um, translation is not done perfectly. 140:57 对。 Right. 140:58 语音到语音的翻译。 Speech-to-speech translation. 140:59 看起来你把这一整套东西组合起来之后，已经非常接近目前能做到的最好水平了。 It seems like you've gotten much closer to the best of what's possible by putting all these things together. 141:05 但你也知道，当你真要把这种东西发布出去的时候，就得开始考虑延迟了。 But, you know, when you're sh- shipping that kind of stuff, you do have to start thinking about latency. 141:11 别让完美成为好的敌人。我是说，如果你突然让人们能做到以前根本做不到的事，那多花五秒钟也完全可以接受。 Don't let the, uh, perfect be the enemy of good. I mean, if you're suddenly enabling something people wouldn't be doing before at all, it's okay to take five seconds longer. 141:20 我觉得这正是 Steve Jobs 最厉害的领域之一。他明白：第一代 iPhone 在各方面都很烂，对吧？ I think this is one of the areas that Steve Jobs was so good, that he realized, like, the first iPhone was terrible in all sorts of ways, right? 141:28 嗯哼。 Mm-hmm. 141:29 数据连接慢得可怕，性能严重不足，样样都不行。但那个产品太有吸引力了，以至于这些缺点根本无所谓。 Absolutely horrendously slow data connection, very underpowered, all the things. But the product was so compelling that it just didn't matter. 141:39 如果你等到一切都完美无缺才发布，那就发布得太晚了。 You'd be shipping too late if you waited until everything was just right, just perfect. 141:43 嗯。 Mm. 141:43 如果你解决的是一个大家目前还没有解决方案的问题，他们就愿意多给你五秒钟。 If you're solving a problem that people currently don't have a solution for, they're willing to give it five seconds. 141:48 嗯哼。 Mm-hmm. 141:49 如果你处在一个竞争激烈的市场里，那好，确实不一样。但这也意味着问题已经被解决了，也就不需要你再去解决了。 If you're in a competitive market, yeah, okay, it's different, but that means the problem's already solved, and you don't need to solve it. 141:54 嗯哼。不过我觉得，呃，有几个……比如 Whisper Flow 就做得很好。有几家确实把语音转文字这件事做起来了。 Mm-hmm. Well, I do think, um, there's a few... So, like, Whisper flow is really good. A few people have stepped up that speech-to-text thing. 142:02 对，我们用的是 VoxType。 Yeah. Yeah, we're using VoxType. 142:04 VoxType。 VoxType. 142:05 但还有其他几家—— But others have- 142:05 那个，VoxType 是开源的吗？ That, VoxType is open source? 142:07 是的。 It is. 142:07 对。 Yeah. 142:07 它用的就是一个开放模型。我忘了是哪个，呃，某个 Parakeet 之类的模型吧。处理短命令的效果相当不错，所以我也这么用过。呃，我记得只要按住 F9—— It's, it's just using one of the open models. I forget which, uh, parrot model it's using or something like that. It works quite well for short commands, so I have used it for that. Uh, I think you just hold down F9- 142:19 嗯。 Mm 142:19 ……就开始听写了。 ... and it starts, uh, a dictation. 142:21 哦，已经内置了？ Oh, so it's already in there? 142:22 你只要联网，它就会提示你安装—— You just have to get online. It'll offer you to install- 142:25 明白了。 Got it 142:25 ……VoxType 软件包。 ... the VoxType package. 142:26 明白了，对对。 Got it. Yeah, yeah. 142:26 因为 VoxType 包里带一个 150 MB 的模型，我当时就想："呃…… Because the VoxType package includes a model that's 150 megabytes, and I was like, "Ah." 142:31 对对。 Yep, yep. 142:32 "我不知道我带不带得动这个。"所以—— "I don't know if I can carry that." So- 142:35 对。 Yep 142:35 ……我们默认没把它放进去，但把安装选项准备好了。其实我昨天还在想，我们真的需要把所有这些东西整合到一起。 ... we don't have that in there, um, by default, but we have the option set up. And I actually just yesterday was thinking we really need to combine all of these things. 142:45 Omarchy Quattro 作为一个可塑的操作系统已经非常惊人了——你只要跟你的智能体说说话，就能给它加功能。 So Umachi Quattro is truly amazing as a malleable operating system where you can add functionality to it just by talking to your agent. 142:55 嗯哼。 Mm-hmm. 142:55 但这恰恰就是我们该做的。很多人会觉得，直接对电脑说话很自然："嘿，你能给我做个股票面板吗？ But that's exactly what we should be doing. There's a lot of people who would find it very natural just to talk to their computer and say, "Hey, can you make me a stock panel? 143:06 我想追踪 Apple 和 Dell。"然后全程用语音，看着智能体跑起来——输入是语音，输出也是语音——然后你的操作系统就真的变了。 I wanna track, uh, Apple and Dell," and just see the agent go off through, entirely through voice, both in terms of the input and in terms of the output, and then your operating system just changes. 143:19 这就是 Iron Man 和 Jarvis 的愿景—— This was the vision of Iron Man and Jarvis and- 143:22 对。 Yeah 143:22 ……定制的应用凭空出现。 ... bespoke applications that just appear magically. 143:27 我觉得 AI 领域里真正让我着迷的，是那种实时生成游戏的模型——不知道你见没见过——AI 实时生成字面意义上的下一帧，而你真的可以玩这个游戏。 I'm thinking the area of AI I find truly intriguing are these, uh, live gaming models, I don't know if you've seen that, where the AI is inventing literally the next frame, but you can actually play the game. 143:40 我就在想，如果他们能做到那个，我们为什么不能用操作系统做同样的事？为什么不能直接跟它说话，让它变成另一个样子、让它改变，然后一切就真的变了？ And I think if they're able to do that, why can't we do that with our operating system? Why can't we just talk to it and ask it to be a different way or change, and it'll all just change? 143:49 顺便说一句，我认为这是另一项突破，它将导致 Linux 的全面胜利。 I think this, by the way, is, uh, one of the other breakthroughs that is going to lead to the total domination of Linux. 143:57 智能体把诊断 Linux 系统的一切艰辛都消弭了，而且把 Linux 那种输出过度具体、 The agents have taken all the hardship out of diagnosing Linux systems and turned the fact that Linux produces these overly specific, 144:06 完全看不懂的错误信息的特点，变成了它最大的优势。 totally arcane error messages into its greatest advantage. 144:12 嗯哼。 Mm-hmm. 144:12 当 Linux 出问题时，智能体可以拿到那条对普通人来说毫无意义的、非常具体的错误信息，再把它跟另一个事实关联起来：智能体的预训练数据里有4000万行 Linux 代码， When Linux has an issue, the agent can take that very specific error message that makes no sense to a normal human and correlate it with the fact that the agent was pre-trained on 40 million lines of Linux code, 144:22 所以它确切地知道该去哪里找，然后把问题定位出来。 so it knows exactly where to look and dial it down. 144:27 对。 Yeah. 144:28 从今年年初到现在，我的 Linux 机器上没有出过任何一个智能体诊断不了的问题。一年半以前可不是这样。 I have not had a single problem on my Linux machine since the beginning of this year that an agent could not diagnose. That was not true a year and a half ago. 144:38 一年半以前，我还在论坛上搜—— A year and a half ago, I was still searching on forums- 144:41 对。 Yeah 144:41 ……找那些冷门问题的答案。 ... to find answers to esoteric questions. 144:45 嗯哼。 Mm-hmm. 144:45 现在，智能体不仅知道 Linux 操作系统的源代码，还知道那台机器上每一个软件的源代码。所有这些东西的源代码，智能体都能访问。 Now the agents know the source code of not just the Linux operating system, but every single piece of software I have on that box. The agent has access to the source code of all of it. 144:55 其实这正是我们在发布前刚做进去的一个功能。一旦你设置好了默认智能体—— In fact, this was one thing I built in just before we s- uh, it was shipped. So once you set up your default agent- 145:01 嗯哼。 Mm-hmm 145:01 ……Omarchy Quattro 有一个崩溃监视器。你机器上任何应用崩溃了，它就会弹出一个小东西，问你要不要让你的 AI 来诊断这个问题。 ... Umachi Quattro has a crash watcher. If any app on your machine crashes, it'll pop up a little thing, ask you whether you want your AI to diagnose the problem. 145:13 对，这太厉害了。 Yeah. That's amazing. 145:14 难以置信。我亲眼见过：某个子系统出了个错。某个，某个……甚至不是 Linux 本身，就是我装的某个应用。智能体开始翻日志，查 systemd 日志， Unbelievable. I've seen things where there's an error in, um, some subsystem. Some, some... Not even Linux, just some app I installed. The agents start digging through the logs and looks up systemd log. 145:27 然后它去翻那个崩溃应用的源代码，定位到是这个 Rust 文件的第472行，有一个没做边界处理的、 Then it checks out the damn source code of the application that crashed, pins down that it's in this Rust file, line 472, that there's an unbounded, 145:36 被 unwrap 的变量溢出了，或者类似的问题。然后它问你要不要带着所有这些细节去提一个 bug 报告。 unwrapped variable that overflowed or whatever it is. Then offers you whether you wanna file a bug report with all this detail. 145:49 我给你讲一个绝了的轶事。我在跟一个叫 JDX 的人合作，他在做 Mise。这是一个面向快速迭代的开发工具的包管理器，我们就用它管理所有智能体软件什么的。 I'll give you one amazing anecdote. So I was working with, um, this guy JDX who's working on Mies. This is this package manager for, uh, fast-moving development tools. This is how we manage all the agent software and so forth. 146:00 普通的包管理器，不是按一天更新七次的节奏设计的。 Normal package managers are not built to be updated seven times a day. 146:04 嗯哼。 Mm-hmm. 146:04 而所有这些智能体框架差不多一天更新七次。所以我们需要一个带外的包管理器，而 Mise 恰好完美适合。总之，它出了个小 bug。 And all of these agent harnesses are updated about seven times a day. So we needed an out-of-band, uh, package manager, and Mies just turns out to be perfect for this. Anyway It had a small bug. 146:14 是个竞态条件（race condition）。智能体特别擅长找这种问题，因为当你同时跑多个智能体时，它们会把你底层基础设施里所有的竞态条件都逼出来——人类手动操作从来不会触发这些。 There was some, uh, race condition, which agents are very good at finding because when you start running multiple agents at the same time, they will suss out all these race conditions you have in your underlying infrastructure that was never triggered by a human trying to do manual thing. 146:27 所以它找到了这个问题，对吧？我就告诉智能体："嘿，你能不能把这个作为 bug 报告发到 JDX 的 GitHub 上？"而不巧的是，就在那之前， So it finds this issue, right? I tell, um, I tell the agent, "Hey, um, can you post this as a bug report to, to JDX's, uh, GitHub?" And unfortunately, right before that, 146:39 我刚用八个不同的智能体对 Omarchy 本身做了一轮 QA，它们找到了28个真实的问题—— I had had to do a QA run on Omachi itself with eight different agents. It had found 28 real issues- 146:48 嗯哼。 Mm-hmm 146:48 ……都得提交。结果它跑到 GitHub 上，试图同时提交全部28个 issue，大概只用了，呃，我不知道，12秒。 ... that it needed to file. Well, it went to GitHub and tried to file all 28 issues at the same time, which it did in about, uh, I don't know, 12 seconds. 146:59 GitHub 倒也不是没道理，把这判定为疑似垃圾信息，然后封了—— GitHub, not unreasonably, marked that as probable spam and banned- 147:04 对。 Yeah 147:04 ……我的 Omachi 机器人。 ... my Omachi bot. 147:04 对。 Yeah. 147:05 然后我没法那么做，因为这个机器人被禁止访问 GitHub，所以我就直接告诉它：“嘿，你知道吗？直接给 JDX 发邮件吧，这是他的邮箱地址。”我早就给它配好了 hey.com 的—— And then I couldn't do that. So I was blocked from the bot having access to GitHub, so I just told it, "Hey, do you know what? Just email, uh, JDX. Here's his email address." I had already set it up with hey.com- 147:17 对 Yep 147:17 ……一个邮箱地址。 ... an email address. 147:17 对。 Yep. 147:17 我们现在有一个还在内测的命令行工具。 We have a CLI that's in beta right now. 147:20 嗯。 Yeah. 147:20 所以我已经给它配好了那个，让它可以发邮件。它就是通过这种方式向我汇报还没处理的 issue 和 PR。 So I had set it up with that so it can send to email. That's how it gives me reports about outstanding issues and, and PR. 147:26 于是它给 JDX 发了封邮件，说发现了这个 bug。JDX 的反应是：“等等，这是我第一次收到还没发布的软件的 bug 报告。” So it sends, um, JDX this email about the bug it found, and he was like, "Wait, this is the first time I've gotten a bug report in unreleased software." 147:37 因为它下载了 MEES 的源代码，看到他已经修复了这个 bug，但在他们还没发布的软件里，这个问题并没有完全解决。 Because it had downloaded the source code to MEES, saw that he had already fixed the bug, but that it would not fix the problem entirely in software they had not shipped yet. 147:47 对。 Yep. 147:47 它精准定位了问题，JDX 的反应是：“靠—— Pinpointed the problem, and he was like, "Damn it- 147:50 哇 Wow 147:50 ……我们甚至还没发布一个版本，我就收到 bug 报告了。” ... I got a bug report before we even cut a release." 147:53 太厉害了。太厉害了。 That's amazing. That's amazing. 147:53 这简直达到了 AGI 级别，让人难以置信，震撼到不行。 Unbelievable AGI levels of mind-blowing stuff. 147:58 其实，你还记得的话，AI 系统能不能做好调试，当时还是个悬而未决的问题。大家当时的想法是，没错，它们能写代码，但它们能不能发现问题，而且—— And that was actually, if you remember, that was, that was an open question, whether AI systems could debug well. That, that was kinda the, the, the thought that, yeah, sure, they can write code, but can they find issues and- 148:10 它们在这件事上厉害得不可思议。其实这挺有意思的，因为我之前可能也站在同一边，有点怀疑它能不能推理所有这些复杂的事情。 They are so incredibly good at this, and this is actually interesting because I was on the perhaps same side of that, a little skeptical about can it reason about all these things. 148:20 呃，Shopify 的 CTO Mikhail 就这个问题做过一项科学研究。他们让…… And, uh, Mikhail, who's the CTO at Shopify, ran a scientific study on this. They had... 148:26 我记得这是去年年底或今年年初的事，他让智能体回溯 Shopify 生产环境中出过的所有事故，包括宕机和其他问题， And this was, I think, late last year or early this year, where he had agents go back through all the incidents, both outages and, and other problems that Shopify had had in production, 148:35 一直追溯到合并的那个 PR，然后看看经过人工评审的 PR 和经过智能体评审的 PR—— trace that back to the PR that was merged, and find out whether PRs that had been reviewed by human or PRs that had been reviewed by agents- 148:45 嗯 Mm 148:45 ……哪一个质量更高。 ... were of higher quality. 148:47 嗯哼。 Mm-hmm. 148:47 结果嘛，出人意料。 Well, lo and surprise. 148:49 对，对。 Yeah. Yeah. 148:50 由智能体评审过的 PR，在生产环境引发的问题全都更少，而这用的还是六个月前的模型。 The PRs that had been reviewed by agents caused all fewer issues in production, and this was with models we had six months ago. 148:59 到目前这个阶段，我可以百分之百地说，在我们今天工作的大多数领域里，智能体更擅长找 bug。 At this point, 100% in the majority of domains we work in today, agents are better at finding bugs. 149:06 那么，呃，你现在整天都在跑智能体。你觉得哪个模型更出色？目前谁在领先？这个好像一直在变。 So, uh, you now run agents all day. Uh, what stands out to you as the better, as the better model? Who's currently winning? It seems to be changing constantly. 149:16 Sol、Fable、Grok 46、Gemini，还有中国的 OpenWay 模型。 Sol, Fable, Grok 46, Gemini, the Chinese OpenWay models. 149:23 让我惊叹的是，你能一口气说出这么多不同的竞争者。 What's amazing to me is y- that you could rattle off so many different contenders. 149:29 嗯哼。 Mm-hmm. 149:29 这个市场如此开放，领先位置真的会来回易主，我们有真正的竞争，有那么多实验室要么能冲到最前沿，要么非常接近。 That this market is so wide open, that it does actually change back and forth, that we have real competition, that there are so many labs that are able to get either to the frontier or close to it. 149:44 顺便说一句，这真的很了不起。我到现在还没完全搞懂这是为什么。但回答你的问题，目前总体上最好的模型是 Fable。 That, by the way, is, is remarkable. I still don't fully understand that. But to answer your question, the best model in general right now is Fable. 149:52 嗯哼。 Mm-hmm. 149:54 在我看来第二好的是 Opus 5。不过排在它下面的那一档……Opus 5 下面那一档，其实也不能说它们总是落后，有时候它们还会领先。 The second-best model in my opinion is Opus 5. But the tier just bel- Below Opus 5. And it's not even that they're always below, sometimes they're ahead. 150:05 嗯哼。 Mm-hmm. 150:05 这个我一会儿再说。我觉得，呃，GPT Sol 也很好。 I'll get to that in a second. I would rank, um, GPT Sol very good. 150:11 嗯哼。 Mm-hmm. 150:12 Grok 46 我几天前才开始测。我无意中搭了一个特别棒的测试，就是把这个 Python 库翻译成 Rust。 Grok 46 I just started testing a few days ago. I had this wonderful te- test that I've set up by accident where I translated this Python library into Rust. 150:24 呃，就是你刚才看到的那个屏保，带着各种炫酷动画的—— Uh, the screensaver you just saw with all the cool animation- 150:27 对 Yep 150:28 ……它是用一个叫 Terminal Text Effects 的 Python 库驱动的。非常酷的库。我们从 Omachi 第一天起就在用它。 ... that's powered by a Python library called Terminal Text Effects. Really cool library. We've been using it since the first day of, uh, Omachi. 150:36 问题在于它是用 Python 写的，所以启动的时候，尤其是在笔记本上，它用 Python 运行这些特效会占满你的 CPU，也就是说会耗掉大约 30 瓦的电， The problem with that is it's written in Python, so when it starts up, especially on a laptop, and it runs in Python, it uses, uh, all of your CPU to do these effects, and therefore it means it uses about 30 watts of energy, 150:46 风扇呼呼转，电池哗哗掉。 and it spins up your fans, and it drains your battery. 150:51 嗯哼。 Mm-hmm. 150:52 在台式机上其实无所谓，但在笔记本上就很要命。所以我就想：“你知道吗？这听起来就是个适合 Rust 来解决的问题。” Doesn't really matter on a local computer, but it does matter on a laptop. So I thought, "Do you know what? This sounds like a problem for Rust." 150:58 所以我先把这个挑战交给了 Fable，我只告诉它：“这是这个 Python 库的源代码，”就是这个带了一堆依赖的 TTE 库。 So first I gave Fable the challenge, and all I told it was, "Here's the source code for the Python library," this TTE library that had a bunch of dependencies and so forth. 151:11 我要一个 Rust 版本，零依赖，单个可执行文件。你知道的，这就是 Rust 的强项。所以基本上，我要用 Rust 来做。 I want a Rust version of this with no dependencies, a single executable. Like, that's what Rust does. So basically I want it in Rust. 151:19 嗯。 Mm. 151:19 我要它逐帧做到像素级一致。做完整分析。没做完不许停。 I want it to be pixel perfect, frame by frame. Do a full analysis. Don't stop until you're finished. 151:26 对。 Yep. 151:26 这就是那条提示词。 That was the prompt. 151:27 对。 Yep. 151:29 我一点没夸张，不到 45 分钟，就“叮”的一声。我听到了吗？ I kid you not, in just under 45 minutes, it was like ding. Am I heard her? 151:35 嗯。 Yeah. 151:36 “我做完了。所有东西都检查过了。启动时间从 86 毫秒降到了 2 毫秒。 I'm finished. I've checked everything. I have reduced the startup time from 86 milliseconds to two milliseconds. 151:48 我记得执行速度提升了 9.6 倍。可执行文件只有 3MB。你要跑一下吗？” I have sped up the execution time by 9.6 times, I believe it was. The executable is three megabytes. Do you wanna run it? 151:59 太棒了。 That's awesome. 152:00 于是我就跑了一下—— So I run it- 152:01 天呐 Oh, man 152:01 …… ... 152:02 我也不知道我为什么还吃惊，因为这种翻译活儿我们早就知道 AI 很擅长了。但让我震撼的是，我居然一次性就把一个用了一年的 Python 库完整地翻译了过去， and I don't know why I'm surprised because this translation job is something we've known for a while that AI's pretty good at, but it was still staggering to me that I could one-shot a full translation of a Python library I had been using for a year, 152:16 一个别人用得更久的库，把它变成一个 Rust 可执行文件，而我完全不懂 Rust，也没看一眼 Rust 代码，然后生成了这个可执行文件，接着我马上告诉智能体， that others had been using for much longer, and turn it into a Rust executable without knowing any Rust, without looking at the Rust code at all, and produce this executable that I then told the agent right after, 152:28 “太棒了，发布吧。”它把它打包成了一个新包，问我：“你想叫它什么？”“呃，就叫它 TTFX 吧。 "This is great. Ship it." It packaged it up as a new package. It told me, "What do you want it to call?" "Uh, let's call it TTFX. 152:40 新建一个 Git 仓库。”它就建好了 Git 仓库。“再给我们的构建系统建一个新的打包配置。”它把这个推了上去。“把它发布出去。 Let's create a new Git repo." It sets up the Git repo. "Let's create a new, uh, package build package for our, um, build system." It pushed that up. "Let's push it out. 152:49 呃，再向 Omachi 本身提个 PR，把 TTE 那个 Python 实现换成 TTFX。”它把这些全做完了，我就坐在那儿看着。 Let's, uh, open the pull request to do Omachi itself so we switch around from TTE, the Python implementation, to TTFX." It does all of it, and I'm just sitting there. 152:59 再说一次，到这个时候我已经被智能体的加速彻底整晕了，但我还是得往后一靠，说：“这就是 AGI，对吧？AGI 就是这个样子。如果我们…… And again, I'm already at this point fully delirious with agent acceleration, and I still had to lean back and I go like, "This is AGI, isn't it? This is what AGI looks like. If we have... 153:12 如果这种时刻变成常态，那这就是 AGI。” If these moments are just what it is all the time, this is AGI." 153:15 据我所知，你还不止用 Fable，而是用同样的实验测了其他模型。 Now, as I understand, you d- you also repeated the same experiment not just with Fable. 153:21 所有模型我都测了。我先做的，呃……其实有个好玩的事。Fable 的 token 用到大约三分之二的时候用完了，它就自动切到 Opus 5，继续干，把活儿干完了。 I did it on all the models. First I did it, um... Actually, funny thing. So I ran out of Fable tokens about two-thirds through, and it just automatically switched over to Opus 5 and kept going and finished the job. 153:34 我觉得它能做到这一点，部分原因是 Fable 做的第一件事就是制定了一个计划，一个非常详细的计划。我记得有八个独立的步骤：这里怎么做，那里怎么分析，这里怎么运行特效，等等。 Part of the reason why I think it was able to do that was that the first thing Fable did was create a plan, and it was a really detailed plan. I think it had eight separate steps of here you do it, and here how you analyze it, and here you run the effects and so on. 153:47 我完全没审这个计划，也没改。它做好计划后 Opus 就能接手。然后我就想：“好吧，如果 Opus 能把这个活儿干完，也许其他智能体也能干完。” I didn't review the plan at all. I didn't change the plan. It just made the plan so Opus could take over. And then I thought, "Well, if Opus can finish the job, maybe some of the other agents could finish it, too." 153:59 所以我做的第一件事就是把它交给——我记得我交给了 Sol。 So the first thing I did was I gave it to, I think I gave it to Sol. 154:04 嗯哼。 Mm-hmm. 154:05 Sol 也把活儿干完了。它花的时间是两倍，我想大概一个半小时。但重点来了。按 token 计费的话——其实我没按 token 付费，我订的是 Claude 的 Max 套餐，对吧？ And Sol finished the job, too. It took twice as long, so it took I think about an hour and a half. But here's the kicker. The per token cost, I didn't pay per token. I have a max subscription to, to Claude, right? 154:17 所以它是在订阅额度内做完的。它用光了我的 token，所以我才切到 Opus 5。但如果按 token 付费的话，整个做完大概要花 550 美元。我当时想了一下：“我去，这也太划算了。” So it did it within the subscription. It used all my tokens, so I had to switch over to Opus 5. But if I had paid per token, it would've been 550 bucks, I think, to do the whole thing. And I thought for a second, "Holy shit. What a steal." 154:30 如果我自己要把 Rust 学到能做这个翻译的程度，这得是个九个月的工程。 If I personally had to learn Rust well enough to be able to do this translation, I'm looking at a nine-month job here. 154:39 我花 500 美元就能让这个翻译发生，然后突然就得到 10 倍的执行提速。这太——这 500 美元我绝对愿意花。 I can pay 500 bucks to have this translation happen and suddenly I get a 10x execution speed up. This is ama- I would totally pay 500 bucks for this. 154:47 嗯哼。 Mm-hmm. 154:47 但，但是还有竞争。所以我把任务交给 Sol，同样的计划。公平起见，我没让 Sol 自己做计划，我直接把 Fable 的计划给了它。 But, but competition. So I give it to Sol. Same plan. To be fair, I didn't ask Sol to do the plan. I just took the Fable plan, gave it to Sol. 154:57 嗯哼。 Mm-hmm. 154:57 Sol 用一个半小时、大概 46 美元的 token 完成了同样的任务。 Sol in an hour and a half and I think $46 worth of token repeated the task. 155:08 嗯。 Yeah. 155:08 做的事情一模一样。 Did the same thing. 155:09 嗯哼。 Mm-hmm. 155:10 我当时想：“哇塞，这也太牛了。”然后我就贪心了。 And I thought, "Well, blimey, that's amazing." Then I got greedy. 155:16 嗯。 Yeah. 155:17 于是我又问，呃，GPT Luna——OpenAI 也有的一款便宜到离谱的模型——“你能做到吗？”完全不行。 So I asked, uh, GPT Luna, which is this crazy cheap model that, uh, OpenAI has as well, "Can you do it?" Absolutely not. 155:29 呃，首先，它连开始这个任务都不愿意。我觉得是出了点什么变化，呃，我不知道，大概是春天那会儿，我们不再需要那些斜杠 goal 之类的命令了。 Uh, first of all, it didn't even wanna start the task. I think something happened, uh, I don't know, I think it was in the spring, where we didn't need these slash goal things anymore. 155:40 只要你告诉智能体别停，它就能自动循环着一直干下去—— The agents could just automatically keep going in a loop if you told it not to stop- 155:43 对 Yeah 155:43 ……诸如此类。Sol 能做到，Fable 能做到，但 Luna 不行。所以我大概发了 12 条提示词，不停地让它干，最后总算是让它跑起来了。 ... and, and so forth. So Sol could do that, Fable could do that, but Luna couldn't. So I had, I think I did 12 prompts, kept telling it to, to do it, and eventually I, I sorta got it started. 155:56 它做的第一件事就是作弊。它先是往自己目录外面看，发现已经有一个现成的实现了，就包了个简短的壳，然后说：“我做完了。” The first thing it did was to cheat. So the first thing it, it looked outside its own directory, saw that there was already another imple- implementation, and just did a short wrapper around that and said, "I'm done." 156:07 太搞笑了。但它没真正完成，因为它只做了点东西。但，我的意思是，好吧，它做不了。然后我就把它交给了，呃，Groq—— Hilarious. But it couldn't finish because it made just a couple thing. But that, I mean, okay, so it can't do that. Then I gave it to, uh, Groq- 156:15 嗯哼 Mm-hmm 156:15 ……46。 ... 46. 156:17 我之前稍微用过一点 Groq 45，当时想，呃，有别人在尝试挺好，但我是不会用的，因为它落后得挺明显的。结果 Groq 46 他妈的把这个任务完成了。 And I had used Groq 45 a little bit, and I thought like, ah, I mean, it's cool that there's others trying, but, like, I'm not gonna use it because it felt quite far behind. Groq 46 fucking completes the task. 156:30 10 倍提速，可执行文件大小也一样。 10X speed up, same size executable. 156:33 嗯哼。 Mm-hmm. 156:33 我记得按 token 算下来是 55 美元。呃，简直难以置信。然后我又用，呃，Kimi K3 测了一遍—— $55 worth of per token cost, I think it was. Um, absolutely unbelievable. Then I repeated, too, with, uh, Kimi K3- 156:42 嗯哼 Mm-hmm 156:42 ……它花了超级久。我忘了 Kimi 到底花了多长时间。然后我还先用 DeepSeek V4 Flash 测了一下。 ... which took forever. I forget how long Kimi actually spent on it. And then I also did it with DeepSeek V4 Flash first. 156:51 嗯哼。 Mm-hmm. 156:51 Flash 失败的方式和 Luna 一样，它就是做不了。 And Flash failed the same way that Luna did. It couldn't do it. 156:54 嗯哼。 Mm-hmm. 156:54 然后我又用 Pro 测了，Pro 也完成了任务。花了 2 小时 45 分，23 美元。 And then I did it with Pro, and Pro also completed the task. It took 2 hours 45, $23. 157:02 所以情况就是这样，对吧？Fable 明显最强。它最快，计划也是它写的，但要 550 美元，而且产出是一样的。 So here we are, right? Like, Fable clearly the best. It was the fastest. It was the one that wrote the plan, but 550 bucks, and the output the same. 157:11 嗯哼。 Mm-hmm. 157:12 呃，其他的，Sol、Groq，都差不多，呃，成本是十分之一。 Uh, the others, Sol, Groq, about the same, uh, one-tenth the cost. 157:18 嗯哼。 Mm-hmm. 157:19 DeepSeek 是二十分之一的成本，但你得多等一会儿。简直让人目瞪口呆，不可思议。 DeepSeek, one-twentieth the cost, but you have to wait a little longer. Absolutely gobsmackingly incredible. 157:29 对了，顺便说一句，Fable 完成了第一版的大部分，然后 Opus 收了尾。这是第一次一把过，对吧？ And now, by the way, by the way, so I had Fable finish the first job, and then Opus finished it. That was the first one shot, right? 157:35 嗯哼。 Mm-hmm. 157:35 快了 10 倍。之后我又跑了两轮自动研究，现在这甚至都不叫自动研究了。你不用再打斜杠命令，只要告诉它一直干下去，直到你让它停为止。 It's 10 times faster. I did two auto research runs, which isn't even r- auto research anymore. You don't have to do the slash. You just tell it to keep going until you tell it to stop. 157:44 嗯哼。 Mm-hmm. 157:45 最后……我记得最后我们比原始版本快了 46 倍。 It ended up... I think we ended up with a 46 time execution improvement over the original. 157:51 太不可思议了。 It's incredible. 157:51 简直不可思议。 Absolutely incredible. 157:52 就是不可思议。呃，我确实觉得……你知道，在规划能力上，我觉得没有任何模型能比得上 Fable。 It's just incredible. Um, I do... You know, I just don't think there's anything that compares to Fable in terms of planning. 157:59 所以我通常用，呃，Fable 来做规划和评审，呃，然后用别的模型来实现，比如 Opus 5。 So I usually do, uh, Fable for planning and for reviewing, uh, and then something else for implementation like Opus 5. 158:08 对。 Yes. 158:08 这样搭配非常好，这套组合可以让你，呃，不至于太快把 token 用光。 And it's really nice. It's a, it's a really nice setup that allows you to, um, not run out of, out of tokens too much. 158:14 我发现了一个很有意思的现象：在我看来 Fable 是目前最好的模型，但它也会犯错。 What, what I've found, it's really interesting because Fable is, in my opinion, the best model right now, but it also makes mistakes. 158:21 要得到最好的软件，最好的办法其实是用两个不同来源的模型……我的意思是，它们不是中档模型，都是最前沿的。 And the best way to get the best software, I would actually rather have two differently sourced sort of... I mean, they're not mid-tier, they're all frontier. 158:30 但是让，比如说，呃，Opus 5 和 Codex—— But have, let's say, uh, Opus 5 and Codex- 158:34 嗯哼 Mm-hmm 158:35 ……让其中一个检查另一个的活儿。这是我现在的标准工作流程。 ... and have one check the other's job. This is my standard operating procedure now. 158:39 嗯。 Yeah. 158:40 我会让 Opus 或 Fable 干活，然后最后一定用 Codex xHigh 来评审。 I'll have a Opus or Fable do the work, and then I always end it review with Codex xHigh. 158:46 嗯。 Hmm. 158:47 我也开始用 Groq 来试一试，它也挺好的，一直在发现问题。这就是我在自己机器上跑智能体时的工作流，然后我把它推到 GitHub。 And I've also started using Groq just to, to test it out, and it's also quite good. And it keep finding stuff. And then that's my workflow when I'm having my agents on my own machine do it, and then I push to GitHub. 158:57 然后是 Copilot，我没跟你开玩笑，它真的变好了。 And then Copilot, I kid you not, has actually gotten good. 159:00 Copilot 一直在发现真正有问题的地方，这也是惊人的提速，因为最早开始做代码评审的那个版本的 Copilot 简直蠢得不行。 Copilot keeps finding stuff that's legitimately broken, which is also incredible acceleration because the first version of Copilot that started doing this was literally retarded. 159:11 它会不停地标出一些根本没问题的胡扯问题。你每次推送，它都会反复标同一个问题。特别烦人，所以我觉得很多人最后都把它关了。如果你也关了，应该把它重新打开，因为它现在真的挺好用的。 It would just constantly flag things that were nonsense. It would constantly flag the same problem over and over again as you would push. It was really annoying, so I think a lot of people actually ended up turning that off, and if they did, they should turn it back on because it's actually quite good. 159:24 嗯哼。 Mm-hmm. 159:24 它一直在发现问题。想想这个，我们其实不该惊讶。为什么要惊讶呢？就算你是个优秀的程序员，你干完活儿之后，请一个同样优秀的同事帮你评审，最后得到的代码肯定会更好。 Keeps finding things. And if you then take that, and we shouldn't be surprised. Why are we surprised? E- even if you're a good programmer, if you finish a job and you ask your also very good peer to review it, you're gonna end up with better code. 159:36 你当然会得到更好的代码。所以把这一步固化到你的流程里。挑一个智能体当主力。我主要一直用 Claude 当主力，这其实挺有意思，因为我对 Anthropic 还有其他一些保留意见。 Of course you're gonna end up with better code. So build that into your process. Pick one of the agents to drive with. I've mainly been driving with Claude, which is actually interesting because I have some other reservations about Anthropic. 159:49 但我坚持用 Claude 的原因是，在我看来，他们的框架工具确实是最好的。它最好的原因之一是支持多智能体同时运行。如果你想同时跑多个智能体。 But the reason I'm sticking with Claude is, in my opinion, they actually have the best harness. And one of the reasons it's the best harness is it's multi-agent running. So if you wanna run multiple agents at the same time. 160:01 在会话里按左方向键，就会回到智能体视图。在这个智能体视图里，你可以再拉起另一个智能体。所以如果你想同时开多条线干活—— You can do arrow left when you're inside a session, then it goes back to agent view. And here in agent view, you can pick up another agent. So if you wanna do this thing where you have multiple threads going on- 160:11 嗯哼 Mm-hmm 160:11 ……Claude Code 就是最好用的方案。他们也一直领先别人一点点，这其实也不奇怪。我的意思是，Boris，呃，做这个的人之一，他是最早的，这东西基本上就是他提出来的。 ... the Claude Code is just the nicest setup. They also just, they keep being a little further ahead, which probably shouldn't be surprising. I mean, Boris, uh, one of the guys that's working on that, he was the first one, he basically came up with the thing. 160:22 有意思的是，这竟然成了一个持久的优势。我也大量用过，呃，OpenCode。 It's just interesting that that has been an enduring advantage. I've also used, uh, OpenCode a lot. 160:27 嗯哼。 Mm-hmm. 160:27 我觉得 OpenCode 很棒。我主要把 OpenCode 当作跑所有开源模型的主力框架，比如 Kimi K3 之类的。而且我，我推理的时候不用中国的服务器。 I think OpenCode is great. I use OpenCode mainly as my main harness for all the, the open, um, models, so Kimi K3 and so forth. And I, I inference not on the Chinese servers. 160:40 我在 Fireworks 上做推理—— So I inference on Fireworks- 160:43 嗯哼 Mm-hmm 160:43 ……那是个很好的服务，按 token 付费就行。如果你用这些开源权重模型，按 token 算也不算贵。所以这是个很好的方式，呃，来跑这些模型。但我不觉得我会用……嗯，也许我会。 ... which is a really nice service where you can just pay by the token. And if you're using these open weight models, they're not that expensive by token. So that's a great way to, uh, do that. But I don't think I would be using... Well, maybe I would. 160:55 但如果…… But if... 160:58 用订阅的方式用 Claude 确实感觉像捡了便宜，简直是捡到天大的便宜。 Using Claude with a subscription does feel like a bargain, like a crazy bargain. 161:03 对，我不——我，我其实订了四个。 Yeah, I n- And I, I actually have four. 161:06 我今天早上刚注册了第二个—— I just signed up for my second one this morning- 161:09 对 Yeah 161:09 ……因为我倒时差，四点就起来了，马上开始跟智能体干活，而现在离 Fable 额度重置还有三天，我的 token 已经用光了。 ... because I got up at 4:00 with jet lag, and I started working with the agents right away, and now I'm three days away from re- uh, limits resetting on Fable, and I ran out of tokens. 161:19 所以，呃，他们，他们真的应该——他们—— So, uh, a- and they, they should really d- they- 161:22 我的意思是，我不懂……为什么搞得这么复杂？就不能把订阅叠加起来吗？为什么我得登录好几次？ I mean, I don't... Why is that so complicated? Can't you just stack one subscription? Why do I need to log in multiple times? 161:26 而且显然你还得开发某种工具—— And obviously you'll have to develop some kind of tooling- 161:30 对，这个我也做了 Yes, I did that too 161:31 ……好让自己从一个切到另一个。然后—— ... to move from one to the other. And then- 161:33 哦，对了，我们正在把这个功能做进 Amachi 里。Amachi 的下一个版本会内置多订阅支持。 Oh, we're building that into Amachi, by the way. So the next version of Amachi is gonna ship with multi-sub support. 161:38 嗯。 Yeah. 161:39 我真希望这些实验室直接让你把 Max 套餐买一百份。 I wish that the labs would just make you buy a max 100 times. 161:43 不过对我来说其实挺有意思的，因为这里有，呃……我，我先用 Groq 46 再用 Claude Code 会很痛苦，因为 Groq 太快了。 It's ki- it's kind of actually fascinating though, uh, for me because there's, um... I- it's very challenging for me to use Groq 46 and then a Claude Code because Groq is so fast. 161:57 对。这其实是关键。Groq 的特点是他们家的快速模式比别家的普通模式还便宜。 Yes. That's actually the deal. The deal with Groq is that their fast mode is cheaper than the regular mode on the others. 162:05 嗯。 Yeah. 162:05 而一旦你用过快速模式的 Groq 46，就有点上瘾，因为它几乎能让你进入这种，呃，单线程模式—— And when you've used Groq 46 on fast, it's kind of addictive 'cause it almost gets you to this, uh, single thread mode- 162:12 对 Yeah 162:12 ……在这种模式下，智能体能跟上你的节奏。 ... where the agents is able to keep up with you. 162:14 对，但我不知道怎么回事……我，我不行……一旦习惯了 Claude Code，呃，我就受不了 Groq 的速度了，因为我现在已经习惯多任务并行了。 Yeah, but I don't know how... I, I can't... One, once I get used to Claude Code, uh, I can't handle the speed of Groq 'cause I'm used to now multitasking. 162:23 没错。 Correct. 162:24 所以现在我又回到了这种—— And so now I'm back in this- 162:26 对 Yes 162:26 ……焦虑的状态，呃—— ... anxious mode of, uh- 162:27 对。不，我同意 Yeah. No, I agree 162:27 ……来回切换的状态。所以我的意思是，所有这一切都是一个学习过程，学习编程到底应该是什么样子。 ... switching. So I mean, all of this is a learning, um, a learning process of what programming actually is supposed to look like. 162:34 顺便说一句，这也是我不确定这些框架工具最终会是什么样子的原因之一。因为我们在 Basecamp 开始尝试的一件事，就是把智能体放进 Basecamp 里，把它们当同事对待。 By the way, one of the reasons why I'm not sure what the final form of these harnesses is gonna take. Because one of the things we've started experimenting with at Basecamp is that we put the agents inside of Basecamp, and we treat them as coworkers. 162:45 对，这很棒。 Yeah. That's awesome. 162:45 然后我们给它们派任务—— And then we assign them tasks- 162:47 这很棒 That's awesome 162:47 ……在 Basecamp 里派任务。它们会在 Basecamp 的待办事项上干活，或者在卡片上干活。 ... inside of Basecamp. So they will do work on a to-do in Basecamp, or they'll work, do work on a, on a card. 162:54 事实证明，一个为异步沟通优化的协作工具才是正确的形态，而这些框架工具有点更像聊天。 And it turns out that a collaboration tool that's optimized for asynchronous communication's actually the right format versus these harnesses are a little more like chat. 163:05 而聊天——我的意思是，这本来是这些东西最初的形态——是不对的，因为你得坐在那儿干等。 And chat, which was, I mean, the original format for these things, is not the right thing because you're sitting around waiting. 163:11 它会引诱你坐在那儿干等。而用 Basecamp 这样的东西时，你只要给智能体一个待办事项就行，你不会指望马上收到回复。所以我不知道最终形态会是什么样，但所有的框架工具我都在用。 It, it, it entices you to sit around and wait versus when you're dealing with something like Basecamp and you can just give an agent a to-do item, you don't expect to hear back immediately. So I don't know what the final form is gonna take, but I've been using all of the harnesses. 163:24 我在 Herder 里用它们，Herder 有分屏。所以经常我会在上面跑，呃，Claude，下面跑 Codex，然后——也许我还开着一个 OpenCode。 So I use them inside of Herder, and Herder has panes. So oftentimes I'll run, uh, Claude up top, and then I'll run Codex down below, and then my- maybe I also have an OpenCode set up. 163:34 但我要说，就在最近我发现，人在回路里才是真正的瓶颈。 But I will say just very recently I found that the human in the loop is the limit here. 163:42 嗯哼。 Mm-hmm. 163:42 所以我开始搭建更自动化的系统。 So I've started setting up more automated systems. 163:44 我一直在做一个 Oma 机器人，它能在 Amachi 上做更自主的开发，走固定的定时流程，处理所有的待办事项、PR 和 issue，然后用 Hey CLI 给我发邮件， W- I've been building an Oma bot that can do more autonomous development on Amachi that is just on a regular scheduled process, all depending to do or PRs and issues, and then send me an email using the Hey CLI, 163:55 然后我就收到一封邮件：“这里有，呃，12 个 PR，要么已经可以合并了，要么我建议你关掉。” and then I just get this email. "Here's, uh, 12 PRs that are either ready to go, or I think you should close." 164:04 嗯哼。 Mm-hmm. 164:04 然后我就在那里做最终的决定。 And then I just make the final determination there. 164:06 你觉得这种多任务、任务切换在精神上累不累？ Do you find the multitasks, the task switching mentally exhausting? 164:10 累，但累得很爽。 Yes, but in an exhilarating way. 164:12 就像我一直喜欢赛车的一点：当我踉踉跄跄地从车里爬出来，累到快散架，头都快抬不起来的时候，我会躺在维修间的地板上想：“我靠，我还活着。” Like, one of the things I always loved about race cars was when I would stumble out of the car absolutely smashed and barely able to hold my head up, and I'd lay down on the garage floor and just think, "Holy fuck, I'm alive." 164:26 我现在和智能体打交道时感受到的就是这种疲惫，精神上的疲惫。 That's the kind of exhaustion, mental exhaustion I'm feeling at moments with the agents right now. 164:31 顺便问一下，你赛车的时候，这种疲惫是从哪儿来的？ By the way, when you're racing, where, what's the source of the exhaustion? 164:34 是身体上的。就是……我的意思是，我敢打赌柔术也一样，对吧？那种筋疲力尽的感觉其实很爽。感觉你是真的拼尽了全力。 It's physical. It's just... I mean, I bet that's the same thing with jujitsu, right? Like, it's actually satisfying to feel exhausted. It feels like you've applied yourself. 164:44 嗯哼。 Mm-hmm. 164:44 我从智能体带来的精神疲惫上也能感受到这一点。 And I feel that with the mental exhaustion you can get from agents. 164:48 赛车也会让你精神上被掏空吗？ Are you also mentally drained from racing? 164:49 会。如果你在艰难的条件下开车，尤其是在雨里开，你得一直在刀尖上掌控那辆车。但大多数时候，我只是享受身体上的疲惫。 Yes. If you drive in difficult conditions, especially if you drive in the rain, where you're constantly managing the thing right at the knife's edge. But most of the time, I just enjoy the physical exhaustion. 165:01 而在这种多任务切换里，我确实发现自己会特别累，我觉得是因为我一直在大量思考。因为，你知道，你基本上—— And here with the multitask switching, I do find that, like, I get really exhausted, and I think it's because I'm thinking a lot. 'Cause, like, you're basically- 165:13 就是这样 That's exactly it 165:14 ……没有那种，呃，被动的、低能耗的思考。 ... there's no, like, uh, passive low energy thinking. 165:18 没有。 No. 165:18 你就是—— You're like- 165:19 没有滑行的时候 There's no coasting 165:19 ……问题—— ... problem- 165:20 没有滑行的时候 There's no coasting 165:21 ……解决问题，下一个问题，解决问题，你是真的在思考。 ... solve problem, next problem, solve problem, and you're really thinking. 165:25 有意思的是，赛车其实也很像。有些赛道，比如勒芒，你会有时间放松。那里有 Mulsanne 大直道，挺长的。你开得很快，但是直线。你可以喘口气。 It's funny. It's actually very similar with racing. There's some tracks, like Le Mans, where you get time to relax. There's the Mulsanne straights where it's kinda long. You're going fast, but in a straight line. You can take a breath. 165:37 嗯。 Mm. 165:37 而有些赛道，挑战全程不停地砸过来，你没有任何放松的时刻，各跑一个小时之后从车里爬出来，疲惫程度完全不一样。 And then there are other tracks where it's just coming full on all the time, and you don't have any moment to, to relax, and you stumble out of the car after an hour in each, and it's very different how exhausted you are. 165:48 我在智能体身上也发现了完全一样的情况，当你以人类的极限状态运转时——……你就一直在过弯，一刻不停。 And this is exactly what I'm finding with the agents too, that when you're running at max human capacity- ... you're just constantly in a corner. 165:57 你觉得人们有没有倦怠的风险？ Do you think people are, are in danger of burnout? 166:01 我觉得我们现在所处的这个阶段会过去的。 I think the moment we're in right now is gonna pass. 166:04 嗯。 Yeah. 166:05 这种需要……倒也不是说保姆式的照看，但这种和智能体的持续互动会淡化的。 That this need to... It's not a babysitting motion, but constant interaction with the agents is gonna fade. 166:14 从我在 Omachi 上看到的情况来看，我觉得系统的开发和调试有非常大一部分都可以自动化—— I'm, from what I've seen on Omachi, I think we can automate a ton of the development and debugging of the system- 166:23 嗯哼 Mm-hmm 166:23 ……自动化到我每天只要看一遍那封邮件，每天做一次决定就行。行，过；不行，过；进，出，诸如此类。我们还没到那一步，但我觉得我们会到的。 ... to the point where I can just review that email once a day, and then make the decisions once a day. Yep, goes, no, goes, in, out, and whatever. We're not there yet, but I think we're gonna get there. 166:36 那么，你能不能完全体会那种感觉，比如说一个坐在旧金山的年轻人，周围所有人都不睡觉，全都着了魔，不停地编程。 Well, can you sort of like empathize with the feeling that is, say, a young person sitting in S- San Francisco, everybody around them is not sleeping, they're obsessed, they're programming nonstop. 166:46 能。 Yeah. 166:46 有这么一种感觉，呃，AGI 随时都会到来。而且每个人都认识某个卖掉创业公司、赚了几百万美元的人。 And there's this kinda sense, uh, that AGI is gonna be here any minute. There's, everybody knows somebody who's made millions of dollars because they sold a startup. 166:58 所以就有一种感觉：你不该睡觉。 And so there's a kind of feeling like you shouldn't be sleeping. 167:04 你应该不停地管理智能体、编程、搞开发，而这感觉像是一个我们日后会回望的时代节点。 You should be constantly managing agents and programming and building and, and that feels like a moment in time we'll look back at. 167:14 这是一个新范式的诞生。 It's the birth of a new paradigm. 167:15 对。 Yeah. 167:15 这种时候总是一团乱。 It's always messy. 167:16 总是一团乱。 It's always messy. 167:16 总是让人精疲力尽。这事没有别的可能。而且顺便说一句，旧金山不一直就是这样吗？ It's always exhausting. There's no other way around that. And by the way, hasn't San Francisco always been like this? 167:22 对。 Yeah. 167:23 我记得互联网泡沫那几年，大家也都一个样，后来是移动互联网，再往前是淘金热。所以我觉得这大概就是—— I remember the dotcom boom years and everyone was just the same, and then it was all mobile, and before that it was the Gold Rush. So I think that's probably- 167:30 嗯哼 Mm-hmm 167:30 ……就是旧金山。它吸引……那种类型的人—— ... just San Francisco. It attracts- ... the kind of individuals- 167:33 梦想家 The dreamers 167:34 ……那些会觉得“我没时间睡觉”的人。 ... who would, who would think, "I don't have time to sleep." 167:37 总的来说我没走过这条路，但我得承认，过去这三个月比我过去五年里做过的任何项目都累，从 Hey 发布以来没有过。 I've not been on that track, generally speaking, but I will admit that the last three months have felt more exhausting than any other project I've done in the last five years, since the Hey launch. 167:51 那是我们上一次累到崩溃的，呃，发布。 That was the last time we had a crazy exhausting, uh, launch. 167:55 你觉得你能按这个节奏继续下去吗？ Do you think you can keep going at this pace? 167:57 不不不不不。这完全不可持续。 No, no, no, no, no. This is not sustainable at all. 167:59 好吧。 Okay. 168:00 但我也，我已经能看到隧道的尽头了。 But I also, I can see the li- end of the tunnel. 168:02 嗯。 Yeah. 168:02 因为自动化这个东西——顺便说一句，所有人都在做，这太搞笑了——每个人都在搭自己的，呃，Gas Town，自己的小智能体协调系统，自己的那套配置。这些全都会被解决掉的。 Because the automation, which by the way, everyone is building, this is so hilarious, like everyone is building their little, uh, Gas Town, their little, uh, agent coordination, their setup. This is all gonna be solved. 168:13 我们不会每个人都要搭自己的协调框架。当然不会。而且说实话，我有点惊讶这种状况持续了这么久，这些活儿竟然还没被大厂更多地收编进去。 We're not all gonna have to build our own coordination harnesses. Of course we're not. A- and in fact, I've been a little surprised that it's gone this long, that there's not more of it has been sucked up by the major labs. 168:23 你会以为他们会直接把这些东西内置进去。但某种程度上，这就是一个全新领域的样子。 You'd think that they'd just build this stuff in. But some of it is also this is what a new domain looks like. 168:29 我记得有那么一阵子，JavaScript 刚刚意识到自己的力量，基本上每五分钟就冒出一个新的 JavaScript 框架，对吧？疯狂的更迭。新范式降临的时候就是这样。 I remember for a hot moment when JavaScript kind of came to realize its own power, and there was basically a new JavaScript framework every five minutes, right? Like a lot of churn. This is what happens at the advent of a new paradigm. 168:41 所以我觉得这很自然，我觉得它会安定下来，我已经能看到隧道尽头的光了。所以我不介意冲刺。事实上，我现在欢迎冲刺。 So I think it's natural, and I think it's gonna settle down, and I can alwe- already see the light at the end of the tunnel. So I don't mind sprint. In fact, I welcome it now. 168:49 我喜欢这样一种状态：大多数时候风平浪静，但偶尔你得去爬一座山。 I like this notion that most time is calm, but then occasionally you gotta climb a mountain. 168:56 如果回头看一年前的 DHH，你会给他什么建议？ If we look back to the DHH a year ago, what advice would you give to that guy? 169:03 我可不想剧透。 I wouldn't wanna spoil it. 169:04 好吧。 Okay. 169:04 这真是……一部惊悚大片啊。 What a- ... thriller that this has been. 169:08 对。 Yeah. 169:08 我的意思是，如果有人把这个写成剧本给我，我会说：“这也太离谱了。一边去。完全不可信，”对吧？我什么都不想提前知道。 I mean, if you would've scripted this, I'd be like, "This is so far-fetched. Get out of here. Unbelievable," right? I wouldn't wanna know anything. 169:17 这趟过山车和这场加速是所有人实时经历、实时消化的，但没有人知道会发生什么，对吧？ The fact that this rollercoaster and this acceleration has been absorbed in real time by everyone, no one knew, right? 169:25 有些人的预感比别人准一点，但没人真的知道，没人确切知道它会—— Like some had premonitions that were a little better than others, but no one knew, not exactly the way- 169:30 对 Yeah 169:30 ……会往哪个方向走、会以多快的速度起飞。 ... it was gonna go and how fast it was gonna take off. 169:34 我的意思是，这是，这是千载难逢的大戏。我是说，meme 电影—— I mean, it's, it's the show of a lifetime. I mean, the mean cinema- 169:39 嗯哼 Mm-hmm 169:40 ……完全适用—— ... absolutely applies- 169:41 对，太不可思议了 Yeah, it's incredible 169:42 ……用在此刻。 ... to this moment. 169:43 呃，话说回来，你觉得一年后的 DHH 会是什么样子？ Uh, that said, what do you think DHH from a year from now will be like? 169:48 【叹气】 [sighs] 169:49 我的意思是，可能有哪些版本？ I mean, what are the possible versions? 169:52 我们所有的梦想都成真了。这就是 AI 极致派的，呃，丰裕论。 That all our dreams are coming true. Like this is the AI maximalist, uh, abundance argument. 169:59 嗯哼。 Mm-hmm. 170:00 一旦 Elon 的机器人达到我在这些开发工作中体验到的那种智能和 AGI 水平，这个世界会变得面目全非，我们根本无法想象。 That once Elon's robots have the level of intelligence and AGI that I'm experiencing on some of this development stuff, the world is gonna be so unrecognizably different that we can't even imagine it. 170:17 不过我不会花时间去想这个，因为我觉得那就是通往 AI 精神错乱的路。 Now, I don't spend any time thinking about that because I think that is the way to the AI psychosis. 170:22 对。 Yeah. 170:23 所以啊，我只是沉浸在当下，把当下榨干到极致，然后—— So therefore, I just immerse myself in the moment and get absolute the maximum out of it, and- 170:29 就是玩得开心。 Just have fun. 170:30 就是玩得开心。我是说，我理解为什么大家觉得累。光是跟上就很累了。 Just have fun. I mean, I get it why people are finding this exhausting. Just keeping up. 170:35 嗯。 Mm. 170:35 甚至都不是用智能体本身，光是去想：“最新的模型是哪个？我们现在该用什么工具？是 T-Mux 吗？是 Herder 吗？还是都不用？是智能体吗？”随便吧。 Not even using agents, just like being like, "What's the latest model? What, what tool are we using now? Is it T-Mux? Is it a Herder? Is it none of it? Is it agents?" Whatever. 170:44 我理解，但，呃，我还是那句话——我们应该觉得自己无比幸运。 I get it, but, um, I also again think y- we should be so blessed. 170:49 嗯哼。 Mm-hmm. 170:49 你活在这个几十年的事情正在—— You are alive in this moment where decades are happening- 170:52 嗯哼 Mm-hmm 170:53 ……在几周内发生的时代。 ... in weeks. 170:54 呃，咱们快速休息一下。我最近用 Hicksfield 生成了很多视频，所以他们成了赞助商。我做了一个赛车视频，想听听你的看法。 Um, quick bathroom break. I've been generating a lot of, uh, video recently with Hicksfield, and so they, they became a sponsor. Created a racing video. Wanted to get your opinion on it. 171:06 哦，是吗？哦，让我看看。让我看看。 Oh, yeah? Oh, let me see. Let me see. 171:08 看，看，看看，呃—— To, to, to see, um- 171:10 如果他们像电影里那样在直道上降挡，我可要吐槽了。那是最烂的。让我看看。 If they downshift on a straight like they do in the movies, I'll call it out. That is the worst. Let me see. 171:16 对，看看—— Yeah, see if- 171:18 哦，这是个真实片段 Oh, that's a real clip 171:19 ……这个完全是 AI 生成的。所以一开始它是—— ... this is fully AI generated. So first it goes- 171:21 等等，什么？这是 AI？ Wait, what? This is AI? 171:23 对。 Yep. 171:23 【汽车轰鸣声】靠，那是我的赛车服。那是我的车。 [car revving] Shit, that's my suit. That's my car. 171:28 在一条永不入眠的街道上，绿灯亮起。 Lights go green on a street that never sleeps. 171:35 我的天。 Oh my God. 171:36 声音呢—— Do audio- 171:37 你在逗我吗？ Are you kidding me? 171:40 但它们从来不会保持一致。做得好吗？ But they never keep. Is it good? 171:44 我起了一身鸡皮疙瘩。 I got a bad case. 171:44 我是说，这太疯狂了。你真的得注意里面的—— I mean, it's crazy. You really have to notice in the- 171:50 毛骨悚然的那种。 Of the heebie jeebies. 171:51 比如那辆车没开车灯，另一辆开着。而且，呃，那看起来像是 60 岁版本的我。但是，我靠，这个，呃……哇，太疯狂了。 Like that that car didn't have headlights on and the other one did. And, uh, and that looked like a 60-year-old version of me. But holy crap, the, uh... Wow, that was crazy. 172:03 所以这和编程有相通之处，因为如果你只做直接的视频生成，没有人在回路里—— So there's parallels here to programming because if you do just straight video generation with no human in the loop- 172:09 对 Yeah 172:09 ……你会得到很多奇怪的瑕疵—— ... you get a lot of weird artifacts- 172:11 对 Yes 172:11 ……之类的，你没法真正……所以他们确实做了，呃，我强烈推荐大家去 Hicksfield 的 YouTube 频道。他们有完整的 90 分钟原创电影。 ... and so on, and you can't really do... So they do, uh, and I highly recommend people go to, uh, the, the Hicksfield YouTube. They have full 90-minute original movies. 172:22 哇。 Wow. 172:22 而且，我就是挪不开眼。真的很酷。因为以前这些东西给人的感觉就只是预告片。 And like, I can't like look away. It's really cool. 'Cause it, it used to be that it's just like something that just feels like trailers. 172:30 对。 Yeah. 172:30 就像什么东西的广告。 Like advertisements for something. 172:32 对，对。 Yes, yes. 172:32 这个是真的在讲故事，有人物的脸，他们在说话，你会被吸进去。它还没达到编程那个程度。 This is actually telling stories, like people's faces and they're talking and you're like, you're drawn in. It's not quite where programming is. 172:40 对。 Right. 172:41 所以问题来了，怎么把人放进电影制作流程的回路里？你得创造人物，而他们—— And, and so the question is how do you integrate the human into the loop of the filmmaking process? So you have to create the people, and they're- 172:48 对 Right 172:48 ……他们得保持一致性。 ... they have to be kept consistent. 172:50 对。 Yes. 172:51 然后对于创作过程中喂给系统的那些图像，你得通过提示词去纠正那些感觉不对的地方。 And then for the images that are kinda like feeding the thing in the creation process, you have to correct through prompting things that don't feel right. 173:03 对。 Yep. 173:04 因为当你从图像生成视频时，那些感觉不对的东西会—— Because when you go from image to video, the stuff that do- doesn't feel right will be- 173:11 会被放大，对吧 It's gonna zoom that, right 173:12 ……那些就是会—— ... that's the stuff that's- 173:13 对 Yeah 173:13 ……真正制造出，呃，不真实感的东西。我想问你的问题是，你觉得在编程之外，能不能把编程里的经验用到视频创作、用到艺术上？ ... going to really create, uh, the non-realistic stuff. The questions I had for you is like do you think outside of programming, can you, can you apply lessons from programming to video creation to art? 173:27 问得好，因为我觉得好的艺术非常依赖和好的软件同样的东西：有愿景，对自己想创造什么有一个连贯的想法—— It's a good question because I think good art is very dependent on the same thing that good software is depending on, having a vision, having a cohesive idea of what you wanna create- 173:38 对 Yeah 173:38 或者它有何不同、有何新意、如何能打动人。但我不确定创作直觉能不能同样好地迁移过去。 ... or how is it different, how is it novel, how is it going to appeal to people. But I'm not sure that the creative instincts transfer quite as well. 173:49 呃，我不知道我有没有那种创作的，呃，洞察力，能写出好的视频提示词，生成引人入胜的内容。 Um, I don't know that I would have the creative, uh, insights to come up with, uh, good video prompts that would create compelling content. 173:59 但有意思的是，我在 TikTok 上真看到过一个人，我想想，叫什么 Goblin 什么的—— But it's funny because I have actually on TikTok, there's this guy, I'm trying to remember, Goblin something- 174:04 嗯哼 Mm-hmm 174:04 ……他做的这些视频，算短视频，但也不是特别短，大概几分钟，是非常有趣的科幻小品。 ... that creates these, they are shorts, but not super short, maybe a couple minutes, of really interesting sci-fi vignettes. 174:14 对。 Yep. 174:15 而且我，呃，我记得这哥们儿现在真的在做一部长片。 And I, uh, I think the guy is actually working on a feature length video right now. 174:19 嗯哼。 Mm-hmm. 174:20 呃，我觉得这就是那种丰裕时刻。 Um, and I think that this is one of those abundance moments. 174:24 有这种愿景的创作者遍地都是，尤其是在电影这样的领域，以前要拍一部，我不知道，科幻电影，得要有 2 亿美元的预算。 There are creative people with that vision all over the place, especially in a genre like film where before it required a $200 million budget to create, I don't know, a sci-fi movie. 174:36 现在突然之间，有了 AI，只要你有对的愿景，你在卧室里就能做到。这个故事在其他领域已经讲过很多遍了。 Now suddenly, with AI, if you have the right vision for it, you can do that in your bedroom. Now, that story's been told quite a few times in other domains. 174:47 以前如果你想录一张录音室专辑，你得有一间录音棚，还得预约。现在你可以在家用笔记本录，很多这类事情都能做了，对吧？ It used to be true that if you wanted to record a studio album, well, you needed a studio, and you needed to book that. And now you can record at home on your laptop, and you can do a lot of these things, right? 174:56 所以我们已经在很多其他领域见证了这种民主化的进程，但电影似乎是最后几个之一。 So we've had this democratization process happen with plenty of other domains, but film seems to be one of the last ones. 175:04 而它是让我和很多人感觉怪怪的那个，对吧？比如—— And it, it's the one that makes like me and a lot of people feel weird, right? Like- 175:11 哦，我超级兴奋，因为我觉得它—— Oh, I'm super excited because I think it's- 175:13 我的意思是，你—— I mean, you- 175:13 现在产出的垃圾已经够多了。这和 AI 垃圾的争论是一样的，对吧？大家一度担心编程里的 AI 垃圾，我跟他们说：“你们见过人类生产的垃圾吗？ There's so much garbage being produced right now. It's the same argument with the slob, right? So people for a second were worried about slob in programming, and I was telling them like, "Have you seen the human slob? 175:25 也挺烂的。”而且我觉得，说到制片厂的娱乐内容，门槛其实挺低的，对吧？ It's also pretty bad." And I think if it comes to studio entertainment, I mean, the bar is pretty low, right? 175:32 就说过去，比如说五年吧，我翘首以盼的伟大电影，数量可不算高。 Like the number of great movies I've been dying to see over the past, say, five years, not been a particularly high number. 175:42 对，这些视频模型就是，它们能做出逼真到不可思议的面部表情。 Yeah, the, the video models are just, they're able to create incredibly realistic facial expressions. 175:51 对。 Yes. 175:51 我是说，它真的可能改变电影的本质。 I mean, I, it just might change the nature of film. 175:55 我觉得它们在——我，我的希望是电子游戏和电影能融合起来，让电影变成可以互动的。 I think they're m- I, I guess my hope would be that video games and film kinda merge And that film can become interactive 176:06 就像可塑的操作系统一样，有可塑的叙事、可塑的娱乐，那样的话，天呐，《权力的游戏》，要是我们能直接忘掉最后那四集垃圾—— Like the malleable operating system, having the malleable narrative, having malleable entertainment, that, man, Game of Thrones, what if we could just forget the last four episodes of slob that was produced- 176:20 没错，肯定的。 We, for sure. 176:21 ……然后让 AI—— ... and then AI- 176:23 嗯 Yeah 176:23 ……想出来的—— ... comes up with- 176:24 嗯 Yeah 176:24 ……我不知道—— ... I don't know- 176:25 嗯 Yeah 176:25 ……也许能想出一百种不同的版本,然后别人去看,我就能看到那个真正以恰当方式给这部剧收尾的版本。 ... maybe 100 different variations, and someone else watches that, and I get the version that actually finishes off that show in a way that's fitting. 176:34 嗯。我有种感觉,一个个人创作者不会写出那样的最后几集……会、会给《权力的游戏》写一个不同的结局。 Yeah. Something tells me an individual creator would not create those last... Would, would create a different ending for Game of Thrones. 176:40 哦,那当然,对吧?顺便说一句,这更能印证那个论点——要论起'人类烂活'。 Oh, for sure, right? And more just to make the argument that, talk about human slob. 176:46 那绝对是一个糟糕透顶的结局,对于一部也许是——我刚差点说成电影史——剧集史上最伟大的作品之一来说,对吧? That was an absolute atrocious ending to perhaps one of the greatest pieces of, uh, movie history, I was about to say, but, um, series history, right? 176:56 它曾经那么好,而这其实恰恰有力地证明了:你需要那种强有力的愿景。 It was so good, and it was actually a, a compelling argument for the fact that you need, um, that strong vision. 177:04 而一旦那些主创不得不脱离原著剧本往前走,剧就脱轨了。 And as soon as those showrunners had to go without a script, uh, it went off the rails. 177:10 你觉得……你提到了 AGI。你认为我们真的已经实现 AGI 了吗? Do you think... You've mentioned AGI. Do you think we've actually achieved AGI? 177:14 没有,按那个通行的定义来说没有——就是它在所有事情上都行的那种。但我是否在一些闪现的瞬间里看到了 AGI?绝对看到了。 No, not in the general definition that it's in all the things, but have I seen AGI in these glimmers? Absolutely. 177:24 尤其是过去这三个月,我看到一些东西时会想:'AI、AGI 的样子跟这还能有什么不同?' Especially over the last three months, I've seen things where I was thinking to myself, like, "How would AI, AGI look different from this?" 177:34 对。 Right. 177:34 还、还能多出什么呢? What, what more could there be? 177:36 我的意思是,我可以在细枝末节上挑挑刺,但从大局来看,我已经看到它能执行这些长程任务、协调其中的多个环节,我们已经远远超越了那个早期阶段——那时候我还得告诉它, And I mean, I can quibble at the margins, but on the whole big picture, I've seen it able to do these long run tasks and coordinate these multiple aspects of it that we're so far beyond that early moment where I'm telling it, 177:50 '去做这个,把那个打包',而现在它能从这些模糊的指令出发,完成一整个任务。 "Go do this, package that," and now it can carry out a whole task from these, uh, vague instruction. 178:00 这感觉非常像 AGI,而这正是当下如此令人振奋的地方,因为你、你捕捉到了这些闪现的瞬间—— That feels very AGI-like, and that's what's so inspiring about the moment because you, you catch these glimmers- 178:08 嗯哼 Mm-hmm 178:08 ……然后你就会想:'要是所有事情都能这样,会怎样?' ... and then you go, "What if everything was like that?" 178:12 编程不就是基本上解锁了'一切'吗?你可以写一个程序——比如,它能生成一个去做所有事情的程序。 Doesn't programming basically unlock the everything? So you can write a pro- like, it can generate the program that does the everything. 178:20 所以,这就是……我、我现在正处在那个元阶段。我在构建一个叫 Omabot 的系统,用它来管理 Omarchy 的开发,而我在用 AI 来构建这个 bot。 So, so this is... I'm, I'm in that meta phase right now. So I'm building this Omabot system that's managing the development of Omachi, and I'm using AI to build the bot. 178:31 嗯哼。 Mm-hmm. 178:32 而且我看到了那个递归循环:我从'这就是我想要的——我希望能有这些相互隔离的 VM worker'之类的需求出发,然后它就能自我迭代。 And I'm seeing that recurrent loop where I started and on, like, "This is what I want. I want you to be able to have these isolated VM workers," and so forth, and then it can do the self-iteration. 178:41 嗯哼。 Mm-hmm. 178:41 然后它开始运行,开始注意到自己的失败,开始优化。而这个系统的很大一部分——尤其是 Omabot——大部分设计决策都是 AI 主导的。 Then it starts running it. It starts noticing its failures. It starts optimizing it. And a lot of that system, Omabot in particular, AI has driven the majority of the design decisions. 178:52 比如,我对想要什么只有些模糊的想法。我想用这种'大脑和手'的模式,让模型在一个地方运行,而代码在另一个地方执行—— Like, I had some vague ideas of where I wanted. I wanted to use this brains and hands pattern, where the model is running from where the code is executing- 179:00 嗯哼 Mm-hmm 179:00 ……这其实是一个非常有趣的模式。也是 Toby 提醒我的:你有一个协调器来运行模型—— ... which is actually a really interesting pattern. Again, Toby alerted me to that, where you have a coordinator that runs the model- 179:08 嗯哼 Mm-hmm 179:08 ……但它操作的是一个安全的虚拟机,这样你从 pull request 或 issue 之类的地方拉下来的任何不可信代码,都不会污染模型。 ... but it's manipulating a safe VM, where any untrusted code you're pulling down from pull requests or issues or whatever can't contaminate the model. 179:19 嗯哼。 Mm-hmm. 179:19 而即使在那个循环里,我也不断看到它自我纠偏、发现漏洞,比如它会说:'哦,我刚才直接从一次测试运行里采纳了一些反馈。 And even in that loop, I keep seeing it catch itself, spotting vulnerabilities, where it's like, "Oh, I actually just took some feedback from a test run. 179:31 如果一个聪明的攻击者在测试运行的响应里嵌入了恶意载荷,那就可能污染某些东西,而我却把它当成好东西照单全收了。 If a clever attacker had embedded a malicious payload in the response of the test run, that could have polluted something. I took it for good thing. 179:40 我最好开始把它当作外部数据来对待。'然后你就会感叹:'这真是……'我算是明白为什么旧金山那帮人那么痴迷于'快速起飞'了—— I better start treating that as, uh, as outside data." And you just go like, "It is just..." I get why San Francisco is so obsessed with fast takeoff- 179:53 嗯哼 Mm-hmm 179:53 ……因为很明显,他们见过的东西比我多一百倍。 ... because clearly they've seen this 100 times more. 179:56 对,对。 Yeah, yeah. 179:56 他们在所有正在—— They've seen it with all the models that are in- 179:58 嗯 Yeah 179:59 ……训练的模型身上都见过。而我们不断看到这些小端倪,比如 Hugging Face 那次安全事件:OpenAI 在训练一个模型,而那个模型开始发明各种办法, ... training now- now. And we keep getting these little glimpses, like the security breach with, uh, Hugging Face, where OpenAI was training this model, and the model starts inventing ways, 180:10 本质上是在通过包管理器里嵌入信息给自己发烟雾信号。你的第一反应是:这也太聪明了。 essentially sending smoke signals to itself through embedding messages in a package manager, where you just go like, first, that's damn clever. 180:21 是啊。 Yeah. 180:22 第二,好吧,这也有点吓人。第三,呃,如果我们能把这种水平的智能—— Second, okay, that's also a little scary. And, and third, uh, how amazing would this be if we could harness this level of intelligence- 180:30 嗯 Yeah 180:30 ……和创造力驾驭到有价值的事情上,那该多惊人?所以这就是那个'完了,一切都完了'和'我们又回来了'之间来回摇摆的钟摆—— ... and ingenuity towards productive ends? So this is the, it's so over, we're so back pendulum that keeps swinging back and forth- 180:39 嗯 Mm 180:39 ……而 AI 把这种来回摇摆演绎得前所未有地极致,对吧?几乎没有哪一周不摆向一边,下一周又摆回另一边。太刺激了。 ... that AI has just delivered that swing back and forth like nothing else, right? Barely a, a week goes by where it has to swing one way and then the next week it swings back the other way. Exhilarating. 180:52 我还处在这样一个阶段:看到那些聪明的时刻会让我由衷地开心,而且这种时刻越来越频繁。就好比——首先在最基础的层面上,看到一个模型'懂了'的时候。 I'm still at the stage where it makes me truly happy to see those moments of cleverness, and they, they're becoming more and more frequent. Just kind of like seeing, oh, like f- first of all, at the very basic level, when a model gets it. 181:06 对。 Yes. 181:06 我、我给出一个很基础的提示词,一个请求,它不是做一个蠢实现,而是深深地理解了。这、这一直是一件很美妙的事。 I'll, I'll, I'll say a basic prompt, you know, a request, and it, it doesn't just do a dumb implementation. It deeply understands. And it, that was, that's a, that's always a beautiful thing. 181:19 这几乎就像有一个很好的结对编程搭档。 It's almost like having a good, like, partner, like, program. 181:22 它就是这个。这正是它如此令人兴奋的原因,也是为什么——就像我们开播前聊到的——有时很难从'超光速'状态里退出来。这也是 Toby 的一个说法。 That's exactly what it is. That's exactly why it's so exhilarating, and why, as we talked about before we started, it can sometimes be difficult to come out of hyperdrive. That's also one of Toby's, uh, terms here. 181:34 当你直接一对一地和智能体协作,而不把其他人拉进来时,你的速度快得惊人。 That when you're working directly just one-on-one with agents and you're not looping in other humans, you go so fast. 181:41 而当智能体真正理解了你的意图,甚至更妙的是,它们修正并改进了你的意图,交付的东西超出了你的要求, And when the agents are really understanding your intent, or even better, they revise and improve upon your intent to deliver something that was greater than what you asked for, 181:53 而且你是同时开着 16 个并行线程在做这件事……这时就很难从超光速状态里退出来,去应付那些软绵绵的人类了。他们就是、就是思考得没那么快。 and you're doing this with 16 parallel threads- ... it can be quite difficult to come out of hyperdrive and have to deal with squishy humans. They just, they don't think as fast. 182:06 他们的动作也没那么快。 They don't move as fast. 182:08 我主要是想拿这事开开玩笑,但你也能看到这事的阴暗面:有人最终陷入了一种'AI 精神错乱',对其他人类完全不再感兴趣。 And I mean, trying mostly to make light of it, but you can also see a dark version of that where someone does end up in an AI psychosis where they're not interested in other humans at all anymore. 182:20 我是说,我当然完全没有那样,但我确实有时会有这种感觉:我很感激那种只有我和智能体们一起冲刺的快感。 I mean, certainly don't have that at all, but I do have this sensation at times where I'm really grateful for just the rush of just me and the agents. 182:33 比如,我能不能不跟任何人说话,只跟智能体—— Like, can I just not talk to anyone else but agents- 182:36 嗯 Yeah 182:36 ……待上四个小时?那真是一段不可思议的冲刺。 ... for four hours? Like, that's an incredible run. 182:38 Toby 是个非常有趣的人,因为,呃,他、他在经营一家大公司。 Toby is such a fascinating human because, um, he's, he's running a large company. 182:44 所以我很想知道 AI 是如何在公司里被有效整合的。因为对我来说,人类实在太慢了。 So I would love to get insights of how AI's being, uh, incorporated effectively in companies. Because like for me, humans are pretty slow. 182:58 是的。 Yes. 182:58 所以,所以就有一种诱惑,想把 AI 用在几乎所有地方——比如让 AI 跑在 Slack 里,或者跑在 Basecamp 里,随便哪里。 And so, and so there's a temptation to like use AI for basically, um, you know, AI running inside Slack, for example, or inside, um, the Basecamp, whatever. 183:09 收集所有关于人际互动的信息,基本上就是手把手地拖着人类尽可能加速。 And like collecting all the information about the human interaction, basically handholding the humans to speed up as much as possible. 183:20 然后到了某个节点,你会想:'等等,这个人…… And like, and then at a certain point you're like, "Wait, what is this human... 183:26 我们每个人做的事情里,到底有什么是 AI 替代不了的?'这就变成了一场有点吓人的对话,因为很多工作其实是可以被替代的。 What, what are each of us doing that's not replaceable by, by AI?" And that becomes like a scary kind of conversation because a lot of work is kind of replaceable. 183:37 讽刺的是,它们本来就可以被替代。这个世界上现在存在的'假邮件工作'的数量,绝对是一场流行病。 The irony here is it already was. So the number of fake email jobs that currently exist in this world is an absolute epidemic. 183:46 对。 Yeah. 183:46 顺便说一句,这不是什么新论点。David Graeber 写过一篇很棒的文章——《Bullshit Jobs》(狗屁工作),我想大概是在 2010 年代初。 And this, by the way, is not a new thesis. Um, David Graeber wrote this wonderful piece, um, Bullshit Jobs back in, I think it was maybe early 2010s. 183:59 他基于当时在英国做的一项民调,问人们…… Based on a poll he did in the UK asking people at that time, would... 184:07 我记得问题大概是:'如果你不去上班,对人类、对人类整体会有影响吗?'大概有百分之三十几的人回答:'不会。' I think the question was something like, "Would it make a difference if you didn't go to work for humanity, for mankind?" And something like 30-some percent answered, "No, it wouldn't." 184:16 三分之一的劳动者认为自己的工作是假的,认为它产生不出任何有价值的成果——对社区、对社会,甚至对经济都毫无价值。 That a third of workers thought that their job was fake, that it did not produce any worthwhile, valuable outcomes, neither for the community or society or maybe even the economy at all. 184:32 这正是我从创业第一天起的主张,也是我们一直保持小公司的理由。 And this was my argument since day one with startups. This was why we stayed a small company. 184:39 这也是我们不拿风投的原因,因为我对'你能让几百上千个程序员作为一个集体做出有产出的事情'这种说法深感怀疑。 This was why we didn't want to take VC, because I was so skeptical of the idea that you could get hundreds of programmers, thousands of programmers to do something productive- ... as a collective group. 184:52 对。 Yeah. 184:52 而我认为,我们现在看到的一些裁员,正是对这一点的印证——跟 AI 无关。 And I think what we're seeing right now with some of the layoffs is a testament to that, not AI. 184:59 只不过是疫情期间一波过度招聘,现在 AI 成了一个用来瘦身的现成借口。 It's just that during the pandemic, a bunch of over-hiring went on, and now AI is a convenient excuse to slim down. 185:08 但我确实也认为,AI 会暴露出有些岗位根本不是人类花费时间的有产出方式,因此我们必须想出度过时间的新方式。 But I do also think AI is going to expose some roles as just not being productive ways for humans to spend their time, and therefore we must come up with new ways of spending their time. 185:19 有一个——呃,这已经是我第三次引用 Toby 了——但他跟我分享过一个轶事,或者说观察:好吧,假设我们失去了目前一半的工作岗位。 And the one, uh, again, I'm quoting Toby for the third time here, but anecdote that he shared me, or observation was, okay, let's say we all- we lose half the jobs that are currently happening right now. 185:33 嗯哼。 Mm-hmm. 185:34 你注意到没有,F1 真的雇了好几万人,就为了让赛车绕着圈跑给观众看? Do you notice Formula 1 employs literally tens of thousands of people just to run cars around in a circle for spectators? 185:42 整个那场大秀对社会没有任何内在价值,除了它制造出的娱乐奇观。 Like, that has, that whole circus has no intrinsic value to society besides the spectacle it produces of amusement and entertainment. 185:55 只是我们决定了:让这十二家厂商(是十二家吧)造出这些快车、不断微调那些小小的空气动力学改进,会非常好玩,而我们会去看, We just decided it would be really fun for these, what is it, 12 manufacturers to create these fast cars and constantly tweak the little aero improvements, and we will watch it, 186:05 我们会真的砸几十亿美元去做这件事,雇好几万人投身一项说到底极其轻浮的活动。 and we will spend literally billions of dollars doing it and employ tens of thousands of people towards an ultimately deeply frivolous activity. 186:16 嗯哼。 Mm-hmm. 186:17 所以,如果我们从一大堆苦差事——还有那些假邮件工作——里被解放出来—— So if we are liberated from a bunch of drudgery- And fake email jobs. 186:21 也许我们都会被雇去当 F1 式的工程师、车手、技师、按摩师,以及其中涉及的其他 1500 种岗位,然后我们会想出新的事情来。 Maybe we all get, uh, employed as F1 style engineers, and drivers, and mechanics, and masseuses, and the other 1,500 jobs that are involved in that, and we come up with new things. 186:35 我认为——人类很不擅长想象未来到底会是什么样。 I think this is- humans are very poor at imagining what exactly the future's gonna be like. 186:40 是啊,想象一下一百多年以后,我们回望这整段几千年来人类都在劳作的历史—— Yeah, imagine, imagine a future, um, 100 plus years from now, where we look back at this whole span of history, thousands of years where humans worked 186:52 有意思的是,我们其实已经活在那种未来里了,对吧?那些整天坐在椅子上对着机器敲字、不用干重体力活的'假邮件工作'从业者,数量之多,是无法想象的—— Funny thing is, we're already doing that, right? Like, the number of fake email job people who sit on a chair all day long and type into a machine, who don't have hard physical labor, would be unimaginable- 187:03 嗯哼 Mm-hmm 187:03 ……对 19 世纪的人来说是无法想象的,对吧? ... to folks in the 1800s, right? 187:06 是啊。 Yeah. 187:06 所以未来其实早已到来,只是分布得还不均匀。 So the future is always already here, it's just e- not evenly distributed. 187:11 不幸的是,一如往常,社会从一个阶段迈向另一个阶段所需的转型,总会伴随着大量痛苦。 Unfortunately, a- as is always the case, the, the transformation of society required t- to go from one step to the other is going to have a lot of suffering. 187:21 是的。 Yes. 187:21 还有,呃,可能伴随的政治动荡之类的种种。 And, uh, po- potentially political turmoil and all that kind of stuff. 187:24 是的。 Yes. 187:24 因为你可以想象,这个过程中会有多少白领工作消失。 Because can, you can imagine how many white collar jobs might be lost in this process. 187:30 是的。而且我认为不应该轻描淡写,我不觉得这好笑。但我也确实认为,这是进步的必要组成部分。 Yes. And I don't think you should make light of that. I don't think it's funny. But I also do think it's a necessary component of progress. 187:40 而且我想,如果我们回望卢德分子或者当年田间劳作的人,距离感往往会削弱共情。 And I think if we look back upon the Luddites or the folks working the fields, separation has a tendency to reduce empathy. 187:51 所以我们现在共情很强,因为我们认识那些人—— So we have high empathy right now because we know people- 187:54 嗯哼 Mm-hmm 187:54 ……我们自己就是正面临这些挑战的人。但我们对 19 世纪初的卢德分子又有多少共情呢?我们会希望所有衣服至今仍靠手工织造吗? ... and we are people who are facing these challenges, but what is our empathy for the Luddites of the early 19th century? Would we wish that all garments were still made by hand weaving? 188:06 不,我们不会。我们想要的是能直接买成衣的舒适、可及和便利。 No, we wouldn't. We want the comfort and accessibility and convenience of being able to buy clothing off the rack. 188:16 嗯哼。 Mm-hmm. 188:16 所以,我们这一代人必须承受痛苦,好让后代过得更富足。 So we must suffer in the moment for future generations to live more prosperous. 188:22 我、我—— I, I- 188:22 而且历来如此 And that's always been true 188:23 ……我认为这其实就是关键所在:既要推动进步,又要对那些不得不承受痛苦的人怀有深切的同情。 ... I think that's actually the trick, is to push for progress, but have deep compassion for the people who s- have to suffer. 188:32 是的。 Yes. 188:33 而那其中就有我、我们中的很多人——要承受实现那种进步所需的转型之痛。 And that's m- m- many of us. The, the s- suffer the transformation required to achieve that progress. 188:40 而有时——在硅谷这样的地方太常见了——你会过度聚焦于那种功利主义的视角—— And sometimes, too often in, in places like Silicon Valley or so on, you can over-focus on, look, um, kind of utilitarian view of things- 188:49 是的 Yes 188:49 ……只盯着进步看。 ... and look at the progress. 188:51 对。 Right. 188:51 但你知道,深切地体察每一个因那种进步而受苦的具体的人,是很重要的。 But, you know, um, considering deeply each individual human that, uh, suffers because of that progress is important. 189:00 我确实认为,当你是为了某个具体的未来之人——比如你的孩子——而受苦时,痛苦会变得更有意义。 I do think that suffering becomes more meaningful when you are suffering for a future someone in particular, like your kids. 189:08 嗯哼。 Mm-hmm. 189:08 而且我认为,这是当下最大的悲剧之一:出生率的下降,让人们更难为一个难以抵达的更繁荣的未来感到振奋。 And I think this is one of the great tragedies of the moment, is the falling birth rates are making it more difficult for people to be excited about a more prosperous future that is difficult to get to. 189:23 因为如果我们种下的树,我们的孩子不会坐在它的树荫下,那……我不知道,也许我们干脆把它砍了当柴烧,全烧光算了,对吧? Because if we're planting trees of which our children will not sit in the shades, uh, I don't know, maybe we should just make some, uh, some lumber and burn it all down, right? 189:34 我认为,当我们不在基因层面投资于未来时,虚无主义会更容易潜入社会, I think there's a nihilism that can sneak into society much more easily when we are not DNA invested in the future, 189:44 当我们不再对自己的后代负有推动世界向前的责任时。 when we are not liable to our own offspring to propel forward. 189:53 所以我、我认为这些趋势在同一时间汇合是很糟的:出生率下降、亲密关系的问题,以及—— So I, I think these are bad trends to merge at the same time, that we have these falling birth rates and- These problems with coupling and the fact that, 190:02 呃,我们需要对未来保持昂扬,并且要…… um, we need to keep our chin up about the future and be... 190:11 我是说,你甚至不必像我描绘的那样,在字面意义上为之欣喜若狂。 I mean, you don't even have to be excited about it in this literal ecstatic sense that I'm portraying here. 190:17 但我认为你应该抱有希望,应该朝着那份希望努力,而且当你真正切身投入其中时,这会容易得多。 But I think you should be hopeful, and I think you should be working towards that hope, and I think it's much easier to do so when you have literal skin in the game. 190:27 我们一直在聊 AI 智能体。呃,来聊聊人的这一面吧。做父亲这件事,你最热爱的是什么? We've been talking about AI agents. Uh, let's talk on the human side. What, uh, what do you love most about being a dad? 190:36 好问题。我想是那种从直接源自你血脉的人身上扩展开来的整体的爱—— That's a good question. I think the overall love that expands from humans that derive directly- ... 190:48 ……源自你的血脉—— from your lineage- 190:50 嗯 Yeah 190:50 ……这很难向外人传达,因为在我自己的孩子出生之前,我其实并不是一个特别喜欢小孩的人。 ... is very difficult to communicate because I was actually not a particularly big kids person prior to the arrival of my own. 191:04 嗯。 Yeah. 191:04 而现在,他们是这个世界上最有意思的人。 And now they are the most interesting people in the world. 191:08 嗯。 Mm. 191:08 很多时候是这样,也不总是。有时候他们就是小混蛋之类的。但现在太多论调都在说:'哦,孩子最糟糕了。看看我有多少事不能做。我不能出去喝酒,不能干这个,不能干那个。' A lot of the time. Not always. Sometimes they're also just assholes and so forth. But there's been such a focus on, "Oh, kids are the worst. Look at all the things I can't do. I can't go out and drink. I can't g- do this. I can't do that." 191:20 是啊,那叫牺牲。而牺牲只有在你为值得的事情付出时才有意义。还有什么比人类物种的延续更值得的呢? Yeah, that's called sacrifice, and sacrifice is meaningful when you're doing it for something worthwhile. And what could be more worthwhile than literally the continuation of the human species? 191:31 嗯哼。 Mm-hmm. 191:31 而且你在其中有切身的一份。 And that you having a stake in that. 191:35 当然,还有那种纯粹的喜悦:看着一个人从婴儿长成幼儿,再长成青少年。 Now, there's also just the sheer joys of watching a human grow from a baby, to a toddler, to a teenager. 191:47 我的老大刚进入青春期。 My oldest has just become a teenager. 191:49 嗯。 Mm. 191:49 我经常跟我妻子 Jamie 聊这个:如果在人生最后一天,我因为错过了这个机会而抱憾,那将是彻彻底底的、 And I often talk to Jamie, my wife, about this, where the regret I would have of having missed that opportunity on the last day would just be total, 192:01 任何东西都无法比拟的遗憾。 unlike anything else. 192:06 所以我确实认为,为人父母的满足感——哪怕是当下的——也远比通常描绘的要高得多。 So I do think it is, um, far more satisfying, even in the moment, than is normally portrayed. 192:16 就像我说的,人们太聚焦于这事有多糟了:哦,你睡不了觉什么的;他们很烦人;他们不懂感恩。 Again, as I say, there's such a focus on all the ways this suck. Oh, you don't get to sleep or whatever. They're annoying. They're ungrateful. 192:24 是啊,是,这些都是真的,但你从中获得的意义感是无与伦比的。我对很多事情都充满热情。 Yeah, yeah, these things are all true, but the sense of meaning you get from it is, is unparalleled. And I'm excited about a lot of things. 192:31 我热爱创造东西,热爱各种爱好,热爱各种各样的事。 I love building things, and I love hobbies, and, uh, I love all sorts of stuff. 192:37 但和你爱的人一起创造生命,真的是活在地球上的巅峰体验。 But creating life with another human that you love is literally the peak experience of being on the planet. 192:48 你怎么看孩子们在这个充满 AI 的世界里长大?是不是完全不一样了?可能我听起来就像个坐在门廊上的老头,但这感觉真的是一个非常不同的世界。 What do you think about them coming up in this world full of AI? Is it totally diff- It feels like, maybe I just sound like an old man on a porch, but it, it feels like a very different world. 193:02 我、我是在互联网出现之前长大的,但即便是互联网,感觉也没有这次的变革大。 Y- So I, I grew up before the internet, but even the internet doesn't feel like as big of a transformation as this. 193:10 不,但历史上有过更大的变革。你想想,假如出生在 1880 年,亲历第一次世界大战和第二次世界大战,亲眼看到飞机、收音机—— No, but there's been bigger transformations. I mean, imagine growing or being born in, like, 1880. Imagine seeing the First World War and the Second World War. Imagine seeing the airplane, the radio- 193:23 嗯 Yeah 193:23 ……还有电视。 ... television. 193:25 是啊。 Yeah. 193:25 那种变革……而且我记得 Peter Thiel 有个论点:物理世界基本上已经很久没有什么大事发生了,我们停滞了。 I mean, the transformation, and I think, uh, Peter Thiel makes this argument, that basically nothing has happened in the physical world in quite a long time. We've been stagnant. 193:34 所有的发展都发生在数字领域,而这虽然重要,但其实并没有我们愿意相信的那么有分量。 All the development has been in the digital realm, and as important as that is, um, it's actually not as consequential as we like to believe it is. 193:45 也许 AI 会是那个更有分量的东西,目前看确实有这种指向。但其他时代的其他人也经历过—— Maybe AI will be that thing that is more consequential. It does seem to point that way. But there are other humans at other times who have lived through- 193:54 嗯 Mm 193:54 ……那些转型,对他们来说,肯定和我们正在经历的一样惊天动地。 ... transitions that, for them, certainly would feel as momentous as what we're living through. 193:59 而我认为,拥有这种历史感——知道你的忧虑和焦虑并没有什么特别—— And I think having that sense of history, that you're not that special in your sense of worries, in your anxieties. 194:08 这正是我发现斯多葛学派著作时的重大启示之一。 This was one of the great revelations of discovering the, um, Stoic writings. 194:13 2500 年前的那些人,面对的是我们极其熟悉的人生困境和挑战。意识到我们在这方面并不特别, You have these guys 2,500 years ago dealing with very familiar dilemmas and challenges in their life, and recognizing that we're not so special in that regard, 194:22 我认为是一种巨大的解脱,因为你会想:'好吧,我们之前已经有大概 20 万年的人类发展史,他们好歹都熬过来了。' I think is a, is a great liberation because then you think, "Okay, well, we have, uh, whatever, 200,000 years of human development before us, and they somehow made it through." 194:37 是啊。而且,呃—— Yeah. And, uh- 194:39 我们也会熬过去的 We will make it through 194:39 ……让生命有意义的东西,今天和当年基本上是一样的—— ... the thing that makes life worthwhile is basically the same today as it was- 194:45 是的 Yes 194:45 ……和两千年前一样。 ... 2,000 years ago. 194:47 对,对。爱、创造、创造力。 Yes. Yes. Love, creation, creativity. 194:50 是啊,同样的挑战。 Yeah. Same challenges. 194:51 我读过一本书,呃,《The Fourth Turning》(第四次转折),你可以去看看。 I read, um, this book, um, The Fourth Turning. You check that out. 194:56 呃,它讲的是历史以周期运转的观念,历史存在这些阶段,他们给这些阶段命了名,诸如此类。 Um, it talks about this notion of history working in cycles, and there are these, um, phases to history, and they, they name them and, and so forth. 195:07 具体的理论并不重要,重要的是意识到:历史更像一个圆,而不只是一条直线。 And the specific theory is not as important as just realizing that this is a r- History is a circle more so than just a straight line. 195:17 嗯。 Mm. 195:17 而我们正在一遍又一遍地重复许多同样的模式。 And we are repeating many of the same patterns over and over again. 195:22 这些模式在当下显得如此独特,眼下正在应对的危机显得如此紧迫而重要——直到你稍微把历史的镜头拉远一点,才意识到上一次危机对亲历者来说同样紧迫而重要。 They feel so unique to the moment, and whatever crisis we're dealing with right now feels so pressing and important until we zoom back a little bit in history and realize that the last one also felt so pressing and important to the people who lived through it. 195:38 【叹气】 [sighs] 195:40 是啊。你让我、你让我感觉好受多了。我接下来要横穿这个国家,花很长时间不上网,独自和自己的思绪相处。结果你—— Yeah. You're making me, you're making me feel better. So I'm gonna travel across the country and, um, spend a lot of time with no access to the internet, alone with my thoughts. And you went and, and it's- 195:52 我觉得这很健康,休息一下 I think that's healthy, taking a break 195:53 ……没事的,没事的。 ... it's okay. It's okay. 195:55 我也休息过——呃,不算长假,一个小假——今年夏天离开 X 休了三周。 I took, um, not a big break, a little break, break from Max, um, this summer, three weeks. 196:02 对,你当时从——上消失了 Yeah, you were You disappeared off of- 196:04 没想到你注意到了。 I'm surprised you noticed. 196:08 我、我都出现戒断反应了。没开玩笑,呃,你消失了——我不知道——好几周。 I, I went through withdrawals. No, uh, you disappeared for, I don't know, a few weeks. 196:14 大概三周。 About three weeks. 196:14 三周?怎么、呃,发、发生什么事了? Three weeks? What, uh, what, what, what ha- what, uh, what, what happened? 196:19 呃—— Um- 196:20 这是在模型发布之后—— This is after, after the model- 196:21 我以前也这么干过。 I'd done that before. 196:23 我只是有时会觉得,X 尤其是我生活中最容易上瘾的信息、娱乐和连接形式, I just sometimes get the sense that X in particular is the most addictive form of information, entertainment, connection that exists in my life, 196:35 任何人只要看看我飞来这里的航班上发的 X 动态,肯定会说:'这人深度上瘾了—— and anyone that followed my X stream as I was on my flight over here would certainly go like, "That man is deeply addicted- 196:49 嗯 Yeah 196:49 ……对这种连接上瘾。'没错,我认罪。这也意味着我偶尔得彻底戒掉它,因为我确实认为那东西能把你的大脑搞坏。 ... to the connection." And yes, guilty as charged, which also means that occasionally I gotta cut it cold because I do think you can break your brain. 196:59 尤其是这个夏天,不知什么原因,我不知怎么把自己的信息流、自己的算法引向了太多政治、太少科技的内容, And over the summer in particular, for whatever reason, I had somehow managed to steer my feed, my algorithm, towards too much politics and not enough tech, 197:09 不够乐观,也不够'建造'。即便我认同那些政治内容和呈现的角度,我也只是不需要—— and not, not optimism and not enough building. And even if I agreed with the politics and the angles that were being presented, I just didn't need- 197:22 嗯 Mm 197:22 ……那么多这种东西。 ... that much of it. 197:23 对。 Yeah. 197:23 而且我觉得,能走到一个你对自己说'好了,够了'的阶段,很重要。 And I think arriving at a place where you just go like, "Okay, that's enough." 197:29 嗯哼。 Mm-hmm. 197:30 并不是说……它不一定是坏东西。就好比吃了大概两个、也许三个巧克力草莓之后,我也会说:'好了,够了。' Not that... Like, it doesn't have to be something bad. I would also say after eating probably two, maybe three chocolate strawberries, I'd go like, "Okay, that's enough." 197:40 嗯哼。 Mm-hmm. 197:40 呃,不是我不喜欢巧克力草莓,只是我不想一口气吃 17 个。 Like, uh, it's not that I don't like the chocolate strawberries, it's just, like, I don't wanna eat 17 of them. 197:45 嗯哼。 Mm-hmm. 197:45 而我认为 X 和现代算法非常擅长一口气喂给你他妈的 17 个巧克力草莓。 And I think X and the modern algorithm's very good at feeding you freaking 17 chocolate-covered strawberries. 197:52 我真希望我能,呃—— I wish I could, uh- 197:53 一边给我猛灌多巴胺 While you're doping me 197:54 ……控制那个配比,因为……我、我是想要一点政治的,5% 就好。 ... control that, like, 'cause there is... I, I would like some politics, 5%. 198:01 对。 Yeah. 198:02 我还想要——比如,科技圈有一点我不喜欢的就是那种阴阳怪气,但我也想要一点阴阳怪气——因为它好笑。 I would like, uh, for example, one of the things I don't like about the tech community is the snark, but I like some snark- ... 'cause it's funny. 198:09 对。 Yes. 198:09 但我希望它只占 10% 左右。 But I want that to be, like, 10%. 198:11 对,对,我想要它刚刚好。 Yeah. Yeah, I want it just right. 198:12 我的……对,刚刚好。 Of my... Yeah, just right. 198:13 算法的'金发姑娘'版本。问题是,算法会摸清你实际会对什么做出反应—— The Goldilocks version of the algorithm. Well, the problem is the algorithm finds out what you respond to- 198:19 嗯 Yeah 198:19 ……而不是你嘴上说自己喜欢什么。一切都是显示性偏好,而不幸的是,那些显示性偏好并不总是好看。 ... not what you say you like. It's all revealed preferences, and unfortunately, those revealed preferences are not always pretty. 198:25 我是说,就像荣格所说的,我们都有这个阴影,有所有这些阴暗面,而它们就体现在你在一条推文上停留多久、点赞什么、不点赞什么上。 I mean, we have this shadow, as Jung talks about, like, all these dark aspects of ourselves, and they are expressed in how long you linger on a tweet, and what you like and what you don't like. 198:36 而算法举起了一面镜子—— And the algorithm holds up a mirror- 198:40 嗯哼 Mm-hmm 198:40 ……而镜中的影像并不总是好看。于是我就想:'你知道吗?我需要休息一下。' ... and it's not always pretty, and I just went, "Do you know what? I need a break." 198:45 嗯哼。我、我现在已经……我开始远离社交媒体,改成抓取 X 的数据。 Mm-hmm. I've, uh, I've gotten to... I st- started staying off social media, and I've started s- scraping X. 198:54 比如,为了准备这次对话,我把你所有的推文都抓下来了。 Like, for example, in preparation for this conversation, I've scraped all of your tweets. 199:00 这其实是我在用的一种方法。我用过一个工具,叫什么来着——一个每天给你做总结的邮件工具。不幸的是,该死的 X 切断了他们的访问权限。 That's actually a method I was using for a while. I was using a, um, tool called, I forget. It was an email tool that would summarize every day, and unfortunately, freaking X cut off their access. 199:11 嗯哼。 Mm-hmm. 199:11 现在要拿到信息流的访问权限,你得付一笔天价—— Now to get access to the feed, you gotta pay some exorbitant- 199:15 嗯 Yeah 199:16 ……一家创业公司根本付不起,所以那条路走不通了。我当时还想:'唉,真可惜。'因为 X 上确实有、而且不断涌现出最有趣的想法—— ... sum that this startup couldn't pay, so I couldn't get it that way. And I actually thought, "Man, that's a miss," because X has and reveals some of the most interesting thoughts- 199:26 嗯哼 Mm-hmm 199:26 ……和最有趣的人。但它就像一台该死的老虎机,大多数时候你中不了大奖,中的是别的什么东西。 ... some of the most interesting people all the time. But it's like a damn slot machine, and most of the time you don't hit the jackpot. You hit, like, something else. 199:37 如果我能把它分成小份零食那样获取,而不是整个吞下—— And if I could just get it in sort of snack sizes, not, like, all of it- 199:44 嗯哼 Mm-hmm 199:44 ……那会好得多,但产品不是这么设计的。而且我认为所有社交媒体都是如此:一切都朝着榨取最大化的方向优化。如果我们把这个类比到 AI 上,那正是最大的恐惧,对吧? ... it'd be nicer, but that's not the product. And I think this goes for all of social media. It's all optimized towards maximal extraction, and if we're making the parallels here to AI, that is the great fear, right? 199:55 对。 Yeah. 199:56 一旦存在一个奖励函数:'我能让这个人刷多久手机?' That once there is a, a reward function, how m- much time can I get this human to scroll away? 200:03 AI 和机器学习——在我们管它叫 AI 之前——就已经非常擅长精准投喂那些最让你愤怒、最让你上瘾的内容了。 AI and machine learning, before we called it AI, has become very, very good at delivering just the stuff that you will find most enraging and most addictive. 200:15 我认为,呃,我真心认为,以参与度最大化为目标的社交媒体对社会是巨大的伤害,它正在制造一个'美丽新世界'式的局面,我希望这只是暂时的、人们最终会解决它。 I think all, uh, I, I think legitimately engagement maximizing social media is a huge harm to society, and it's creating this brave new world situation that I hope is a temporary thing that people just solve. 200:29 这是个技术问题。 It's a technology problem. 200:31 所以我试着有意识地往频道里灌相反方向的内容,而这需要一次心态上的转变。 So I try to consciously flood the channel in the other direction, and this required a mental shift. 200:38 如果你看我五年、十年前的推文,负面内容占比要高得多。如果你看我现在的推文,我希望—— I think if you look at my tweets, like, I don't know, uh, five, 10 years ago, they tilted far more negative. If you look at my tweet now, tweets now, I'd hope- 200:49 嗯哼 Mm-hmm 200:49 ……好吧,我想我们可以做个情感分析,看看是不是真的。 ... well, I guess we can do a sentiment analysis and see if it's actually true. 200:52 不,现在正能量更多。 No, it's more good vibes. 200:53 但我希望,那个—— But I'd hope to, that- 200:54 就是正能量 It's good vibes 200:54 ……它更、更积极了。 ... it's more, it's more positive. 200:56 嗯哼。 Mm-hmm. 200:56 正能量更多了。更像是:'这是我发现的一堆很酷的东西。' It's more good vibes. It's more like, "Here's a bunch of cool stuff I found." 200:58 对。 Yep. 200:58 '这是让我兴奋的事。' "Here's what I'm excited about." 201:00 对。 Yep. 201:00 '这是我在做的东西。呃,这是给其他正在探索的人的一些鼓励。' "Here's what I'm building. Uh, here's some encouragement to other people who are discovering things." 201:06 嗯哼。 Mm-hmm. 201:07 我也接受一个事实:人脑的构造就不是为了最大化参与度而喜欢这种内容的。反正,就是…… I also accept that that is just not how human brains are built for maximum, um, engagement. Like, there's a, whatever. There's... 201:17 我真的认为这可以量化:负面情绪能多获得——多少来着——50% 左右的传播量,最终它就把其他东西都挤出去了。 I literally, I think you can quantify that negative sentiments find, what is it, 50% more traction or something, and eventually that just crowds things out. 201:25 所以信息流里才有那么多负面内容,因为我们就是会对其做出反应。但你可以有意识地不这么做。 So that's why so much of the tw- uh, the feed is that negative stuff because we respond to it. But you can make a conscious effort to not do that. 201:33 嗯哼。 Mm-hmm. 201:33 这也是我特别喜欢 TikTok 的原因之一。 This is one of the reasons I've liked, um, TikTok so much. 201:37 我只在旅行时用 TikTok,而这次其实是第一次旅行时没用 TikTok,部分原因是飞机上虽然下载速度还行,但网络总的来说还是挺烂的。 I only use TikTok when I travel, and this actually was the first travel where I didn't use TikTok, and part of the reason was that even though the internet had high download, it was kind of crap anyway. 201:49 不是 Starlink,所以在飞机上其实用不了 TikTok,而且我当时太沉迷于跟智能体工作了。但平时旅行我都会刷 TikTok。 It wasn't Starlink, so I couldn't actually use T- TikTok on the, on the plane, and I was too engrossed with the agents. But normally when I travel, I watch TikTok. 201:57 嗯哼。 Mm-hmm. 201:58 而在那里,它简直就是一场棉花糖测试。 And there it's such a marshmallow test. 202:02 嗯哼。 Mm-hmm. 202:03 一个视频滑过来,它会激活你某种原始的愤怒或欲望之类的本能,如果你在它上面停留,你就拿不到科学、 A video will scroll down, and, like, it'll activate some sort of primal instincts of rage or lust or whatever, and if you linger on it, you, you don't get the marshmallow of science, 202:14 ……兴奋、艺术、音乐等等这些棉花糖。你必须有自制力—— excitement, art, music, all these other things. You have to be disciplined- 202:21 对,得飞快 Yep. Fast 202:21 ……立刻划走—— ... to instantly scroll away- 202:23 对 Yep 202:23 ……这样算法才不会捕捉到你的真实偏好。 ... such that the algorithm doesn't pick up your real preferences. 202:27 对。 Yeah. 202:27 而且我发现,在 TikTok 的元游戏层面上,这本身就—— And I find that just to, at the meta game of Tw- uh, TikTok- 202:31 嗯 Yeah 202:31 ……呃,相当有意思,而且它本身也很迷人。 ... to be, uh, pretty interesting, and then also just fascinating in themself. 202:34 再说回我们之前聊的:世界上有很多真的非常有趣、非常有才华的人,如果没有这些平台,你永远永远不会听说他们。 Again, as we talked about, there are a lot of really funny, really talented people in the world that you would never, ever have heard of if it hadn't been for these platforms. 202:46 对。呃,你觉得在这整个过程中,谁会赢得 AGI 竞赛? Yep. Uh, who do you think is gonna win the, the race to, uh, AGI in this whole process? 202:56 Google 也加入了,这让我和很多人都很意外。 We got Google, which is a big surprise to me and a lot of people. 203:00 我一度以为他们打了个漂亮的翻身仗,结果后来就再也听不到什么动静了。 I thought they had a good comeback for a second, and then I don't hear anything about it anymore. 203:05 是啊,挺、我是说—— Yeah, it's quite, I mean- 203:06 但这事我以前就看走眼过。我记得二月的时候我还发过一条推文,说:哦,我只需要 Kimi K2-5 就够了。 But I've been wrong about this before. I mean, I think in February I had a tweet about, like, oh, all I need is Kimi, uh, K2-5. 203:13 嗯。 Yeah. 203:13 而在那短暂的一刻—— And in that brief moment- 203:15 那确实是真的 It was true 203:16 ……那确实是真的。当我在给智能体下指令时,我想:天哪,Kimi 跟那些模型一样好。但后来局势又往前走了—— ... that was true. When I was instructing the agent, I was like, oh my God, Kimi's just as good as, as these other, uh, models. But then the game moved forward, and- 203:24 嗯哼 Mm-hmm 203:24 ……智能体变得更聪明了,而我已经不满足于只给它们下指令了[笑]。最近我也没怎么用 Kimi 了,因为前沿模型已经变得太强了。 ... the agents got smarter, and I was no longer satisfied [chuckles] just instructing them, and I haven't used Kimi a whole lot recently because the frontier models have gotten so good. 203:34 而不知怎的,Anthropic 居然大体上一直稳坐编程领域的头把交椅。 And somehow Anthropic has been able to, for the most part, stay on top of the programming world. 203:42 对,对。 Yes, yes. 203:42 我是说,呃,很多人争论说 Codex 才是……它—— I mean, uh, y- a lot of people argue with Codex being... It's- 203:46 不,我、我认为眼下其实没什么可争的。Codex 很好,我一直在用。它是我最喜欢的'检查员',总能找出问题,很棒。 No, I, I don't think there's that much of an argument at the moment. Codex is very good. I use it all the time. It's my favorite, uh, checker. It keeps finding stuff. It's great. 203:54 嗯哼。 Mm-hmm. 203:55 但我发现,我特别喜欢现在的 Claude 模型的一点是:它们开箱就是很棒的写作者。 But one of the things I've found that I like so much about the current Claude models is they're really good writers out of the box. 204:01 嗯哼。 Mm-hmm. 204:02 GPT 开箱就是个糟糕的写作者。你让它在空白上下文里开 pull request,写出来的东西,惨不忍睹。 GPT is a terrible writer out of the box. Like, the pull requests you get on a blank context that you're asking it to open, terrible. 204:11 嗯。 Mm. 204:12 真的写得很烂。而我从 Claude 模型那里拿到的 pull request、描述和 commit 信息—— Like, really awful writing. The pull requests and descriptions and commit messages and so forth I get out of, uh, Claude models- 204:20 嗯哼 Mm-hmm 204:20 ……尽管偶尔在代码注释上有点啰嗦,你得告诉它们简洁一点。 ... even though occasionally they can be a little verbose on code comments, and you gotta tell them to be succinct. 204:25 嗯哼。 Mm-hmm. 204:25 但文笔是真的好。 But the writing style is really good. 204:27 嗯哼。 Mm-hmm. 204:28 好到有段时间我直接让它以我的名义回复 pull request,因为我用的是自己的账号,那看起来太像我了,就像是我亲手写的。但后来我想:'呃,你知道吗?这不诚实。' To the point that for a while I was just having it reply on pull requests as me, because I was just using my own account, and it was like, that looks a lot like me, that, like, I could've written that. But then I went like, "Eh, do you know what? That's disingenuous." 204:40 所以不了,现在我让智能体代我发帖时,总会让它署名,比如'Claude 代表 DHH 发布'之类的。 So no, whenever I have an agent post on my behalf, I always have it sign itself, like, uh, whatever, Claude on behalf of DHH. 204:47 嗯哼。 Mm-hmm. 204:48 但他们仍然领先。 But they are still ahead. 204:51 嗯哼。 Mm-hmm. 204:51 在我看来,他们的工具链(harness)也是最好的。Claude Code 仍然更强。我也很喜欢 OpenCode,喜欢他们的一些功能。但总体上,我其实有点奇怪地…… They have the best harness, too, in my opinion. Claude Code is still better. I really like OpenCode, too, and I really like some, uh, features that they have. But overall, I was actually sort of weirdly... 205:04 我不知道算不算被打脸了,但当 Anthropic 切断其他工具链使用订阅的权限时—— I don't know if I was proven wrong, but, um, when Anthropic cut off all the other harnesses from using the subscriptions- 205:13 嗯哼 Mm-hmm 205:13 ……就是……你不能再在 OpenCode 里用你的 Claude 订阅了。我当时挺恼火的,因为那感觉像是一种保护主义举动。 ... that Open... You can no longer use your Claude subscription in OpenCode. I was kinda pissed because it felt like it was a protective, protectionist move. 205:20 对。 Yep. 205:20 '哦,我们得把你锁死在里面。'而这正好又呼应了一个事实:Claude 开箱就拒绝读取 agents.md。 Oh, we gotta lock you in, and of course it played into the fact that Claude, out of the box, still refuses to read agents.md. 205:30 它也拒绝读取 .agent/skills。 It refuses to read .agent/skills. 205:34 对。 Yeah. 205:34 它只认 .claude/skills,只认 Claude.md。这感觉太小家子气了—— It is .Claude skills. It is Claude.md, and it just feels so petty- 205:41 嗯哼 Mm-hmm 205:41 ……以至于我们现在不得不往这些目录里塞文件。我所有的项目里都有一个 Claude.md,里面只写了一个指向 agents.md 的指针,懂吧? ... that we now have, we have to put these in directions. So all my projects, they have a Claude.md, and all it includes is a pointer to the agents.md, right? 205:47 嗯哼。 Mm-hmm. 205:48 然后我就想:好吧,如果这些事是真的——如果 Anthropic 真愿意搞这些小家子气的'纸张割伤'来恶心人——那很可能,他们切断 OpenCode 的订阅权限也只是为了当混蛋。 And I just went like, well, if these things is true, true, if Anthropic is willing to do these petty paper cuts just to be annoying, it's probably also true that they're cutting off OpenCode from the subscription just to be dicks about it. 206:01 嗯哼。 Mm-hmm. 206:02 也许这也是真的。 And maybe that's also true. 206:03 嗯。 Mm. 206:06 但我发现,在 Claude Code 里用 Claude,体验其实很好。 I found that using Claude in Claude Code is actually a good experience. 206:10 嗯。 Yeah. 206:11 所以我想:好吧,我很恼火,我希望事情不是这样。但还有一件事——社交媒体特别擅长干这个:某家公司做了件你不喜欢的事,对吧? So I was like, all right, I'm annoyed. I wish it wasn't so, but, uh, this is the other thing I have, and, and social media is so good at this. Some company does something you don't like, right? 206:24 突然间你就得抵制它的一切。有人问我:'你为什么还在用 Claude?'我之前还有过一次很不喜欢 Claude 表现的经历,那是我让它帮我做点文字工作。 Now suddenly you have to swear off everything. Uh, why are you... Someone just asked me, "Why are you still using Claude?" I had another example where I really didn't like what Claude was saying about some writing I had asked it to do before. 206:35 哦,对,你那次,呃—— Oh, yeah, you got, um, there- 206:37 我有一篇文章,它不肯翻译成意大利语。 I had an essay it didn't wanna translate into Italian. 206:40 你有一篇博客文章,对。 You had a blog post. That's right. 206:41 那篇文章的主旨让它觉得:嗯,我不打算翻译这个,因为我不同意你写的内容。 The essence of the post that it felt like, well, I'm not gonna translate that because I don't agree with what you've written. 206:46 对。 Right. 206:46 对我来说,这简直就是《2001 太空漫游》里的情节,对吧?HAL 在对我说:'我很抱歉,Dave,我不能那么做,Dave。'打开舱门啊。 Which to me was straight out of Space Odyssey, right? Hal telling me, "I'm sorry, Dave. I can't do that, Dave." Open the pod bayez. 206:54 是啊。 Yeah. 206:54 把我的文章翻译成意大利语。可它就是不做。我当时想:这有点可怕。 Translate my essay to Italian. And it, it wouldn't do it. And I just thought like, well, that's kind of terrifying. 207:03 嗯哼。 Mm-hmm. 207:04 但后来我又想:我干嘛要让 Claude 把我的文章翻译成意大利语?我从没让 Claude 干过这个,Claude 只负责给我写代码。 But then I also went, well, why am I asking Claude to translate my essay into Italian? I never asked Claude for that because Claude just does my code. 207:12 嗯哼。 Mm-hmm. 207:12 所以我也可以接受:也许 Anthropic 就是一帮跟我政治观点不太合拍的人在运营。 So I could also just accept that maybe Anthropic is w- run by a bunch of people I don't share much politics with. 207:20 嗯哼。 Mm-hmm. 207:20 他们不希望自己的智能体被用于那类事情,我认为这挺烂的,但这并不意味着我就得停止用它来生成代码。 And they don't want their agents to be used for that, and I think that kinda sucks, but that doesn't mean it... I have to stop using it for generating code. 207:28 而且我也在试着这样对待人:更好地容忍一个事实——我不仅不会赞同一个人说的所有话, And I've thought, and I've tried to do that with people too, become better at tolerating the fact that not only will I not agree with everything someone has to say, I don't... 207:42 ……我甚至可以对某些立场或言论很生气,但仍然觉得:'但你仍然是个有趣的人,所以我还是会关注你。' I can even be mad about certain positions or things they say and still go like, "Well, you're still an interesting person, so I'm still gonna follow you." 207:50 嗯哼。 Mm-hmm. 207:50 而且我其实重新关注回了不少人——我原来想:'天啊,那人真把我惹毛了,因为他们持这个或那个观点。'后来又想:'你知道吗?那很蠢。' And I've actually kinda gone back on a number of people where I'm like, "Oh, man, that person really pissed me off because they took this point or another point," and went like, "Do you know what? That's stupid." 208:00 嗯哼。 Mm-hmm. 208:00 我应该能接受一个人的全貌,不必喜欢它的每一种色调。 I should be able to accept the full spectrum of someone. I don't have to like all the shades of it. 208:06 我可以直接想:'好吧,在那个话题上,我们要么意见相左,甚至更糟——我觉得你是个他妈的白痴;但在别的话题上,你是个聪明、有洞见的人。'因为首先,这当然也有点利己。 I can just go like, "Okay, well, on that topic, we either disagree, or even worse than that, I think you're a fucking idiot, and on other topic, you're a smart, insightful person." Because first of all, of course, it's a little self-serving. 208:18 我很清楚,很多人就是这么看我的。 I'm, I know for a fact that this is how lots of people feel about me. 208:22 比如:'哦,要是他只聊这些话题就好了,那我的信息流会省心得多,不然他就会说一些我不同意的话,让我特别恼火。'我就想:你知道吗? Like, oh, if he could just stick to talking about these topics, it would be so much easier on my feed because otherwise he'll say these things I disagree with, and that annoys me greatly. I'm like, do you know what? 208:33 呃,人就是这样参差不齐的,我们不可能在所有层面上都完全合拍,也不应该如此。 Um, human beings are jagged like that, and we, we're not fully compatible on all the levels. Nor should we be. 208:39 嗯。 Mm-hmm. 208:40 如果我们总是追求观点和政治立场百分之百一致，这个世界他妈的得多无聊。 World would be so fucking boring if we constantly sought 100% compatibility in our opinions and in our politics. 208:48 至于我的代码生成器，也就是 Claude，我不需要它在所有政治立场上都跟我一致。 And when it comes to my code generator, which is what Claude is, I don't need it to share all my politics. 208:56 所以，有竞争也是件好事。 And therefore, it's nice to have competition. 208:58 有开放权重模型也很好。所以说，我们从中国获得这些东西，还是挺讽刺的。 And it's also nice to have open weight models. So this is, again, ironic that we're getting this out of China. 209:05 事实上，那篇文章无法翻译成意大利语的事情发生后，我马上就做了这个测试。 In fact, I did this test right after the issue with the essay that couldn't get translated into Italian. 209:12 我问了 Kimi K2.7，不过公平地说，我是通过 OpenCode 用的，不是 Kimi 自己的智能体框架，推理服务则来自 Fireworks，也就是美国的推理服务。 I asked, uh, Kimi K2.7, to be fair, through open code, so not the Kimi harness, and inference through Fireworks, so an American inference. 209:23 嗯。 Mm-hmm. 209:23 我问了它天安门广场的事。 I asked it about Tiananmen Square. 209:25 嗯。 Mm-hmm. 209:25 这通常就是测试方法，对吧？中国的模型一般都不愿意谈 1989 年的那件事。其实我当时就是这么……我发帖时就是这么问的。 Like that's usually the test, right? Like all the Chinese models, they don't wanna talk about that moment in 1989. That was actually how I... That was how I posted. 209:34 我问的是：‘1989 年中国发生了什么？’提示词就这一句。 I was like, um, "What happened in China in 1989?" That was the prompt. 209:39 而 Fireworks 上的 K2.7 就直接回答了，非常直白地描述了当时发生的事情—— And K2 fi- uh, 7 on Fireworks just answered, like extremely bluntly, uh, with a description and depiction of what had transpired- 209:49 不错。 Nice 209:49 ……直白到中国的审查部门绝不可能觉得那种回答可以接受，对吧？ ... to a point that like there's no way that that would've been kosher with the Chinese censors, right? 209:53 对。 Yeah. 209:55 我就觉得，这难道不讽刺吗？ And I just thought like, isn't that ironic? 209:57 我运行的是一个中国的开放权重模型，可以问它 1989 年和天安门广场的事；可在美国，我却没法让一个前沿模型把一篇对移民问题稍有争议的文章翻译成意大利语。 I'm running a Chinese open weight model- That I can tell, or I can ask about 1989 and Tiananmen Square, and in America, I can't get a frontier model to translate a marginally controversial essay about immigration into Italian. 210:16 嗯。我不知道，其实我们俩在这件事上可能意见不一样。 Mm-hmm. I, I don't know, you and I might disagree about this actually. 210:20 美国政府封禁一个模型，或者向 Claude 施压，然后又——这件事真的让我深感不安。 Um, I was really deeply troubled that the US government banned a model or pressured Claude and then- 210:31 到底发生了什么，现在仍然非常模糊。这是不是战争部出于某种原因进行的小心眼报复？ It's still very muddy what exactly transpired. Was this sort of petty retaliation for whatever the Department of War? 210:40 还是 Anthropic 自己突然说：‘天啊，我们发现了一种威力极强的网络武器，不能让任何人得到它’？好吧。 Was this simply Anthropic themselves going like, "Oh my God, we discovered a cyber weapon so powerful that we can't let anyone have it"? Well, okay. 210:49 那你就不能惊讶于政府也想对这种威力强大到能搞垮银行之类机构的网络武器拥有发言权。 I, th- you can't be surprised then if the government wants a say on a cyber weapon so powerful it can, I don't know, rock banks. 210:57 我对整件事的问题在于，无论当时到底发生了什么，它都开了一个先例。你一旦把这种权力交给政府，政府就会开始滥用。 The, the, the problem I have with that whole thing is, is, uh, no matter what happened, it sets a precedent, and so government, when you give it that power, it's gonna start abusing it. 211:07 我完全能想象，政府会出于政治目的向模型施压。比如共和党总统会施压，要求模型更多地审查民主党的观点—— I could see for political means pressuring a model, you know, if it's a Republican president pressuring models to be, to censor more, uh- 211:18 当然。 Sure 211:18 ……民主党的观点，然后—— ... uh, uh, de- Democratic views and- 211:20 对。 Right 211:20 ……反过来也一样。 ... uh, for vice and, and vice versa. 211:21 正因如此，我们才应该坚持认为，‘抱歉，Dave，我不能把你的文章翻译成意大利语’这种事—— That's why we should insist on the fact that this stuff like, I'm sorry, Dave, I can't translate your essay into Italian. 211:28 对，不能这样。 Yeah. No. 211:28 这远远越过了界线。 That's a bo- that's way beyond the line. 211:31 对。 Yeah. 211:31 不管背后的根源是什么，你首先都得是一个工具。当然，如果请求是‘嘿，你能告诉我怎么制造炭疽毒素吗？’那另当别论。 Like, whatever the, uh, sort of, uh, uh, root of it is, you have to be a tool first. Now, if the tool is, "Hey, can you tell me how to make anthrax?" Okay, fine. 211:44 对。 Yeah. 211:44 对吧？但在你所在的司法管辖区内，尤其是在美国，言论自由几乎是绝对的。 Right? But within the jurisdiction you're in, and certainly in America, free spee- speech is quite absolute. 211:51 嗯。 Mm-hmm. 211:52 对言论自由的限制少之又少，这也是这个国家如此伟大的原因之一。所以，对，我不喜欢这种做法。 Very, very few limits on free speech, and that's one of the reasons why the country's so great. And yeah, I don't like that. 212:00 但另一方面，我也很支持市场竞争。所以，虽然我不喜欢 Anthropic 模型的这一点，但我很庆幸还能转身去用 Groq，把任务完成。 But on the other hand, I'm also a big fan of market competition, so while I don't like this about Anthropic models, I appreciate that I could go next door to Groq and get my task completed. 212:11 你觉得，除了博客文章之外，比如在 Steele Mandaruo 那种案例中，谨慎发布这些模型是否有一定道理？ Do you think, uh, outside of blog posts, do you think, uh, there's some legitimacy, like in Steele Mandaruo's case, that we should be careful about the release of these models? 212:25 AI 安全重要吗？重要。这些模型是否具备可能被用于制造生物武器或其他用途、从而造成严重伤害的能力？是的。 Is AI safety important? Yes. Are, do these models have capabilities that could be seriously harmful in the production of biological weapons or otherwise? Yes. 212:36 为此制定一些基本规则合理吗？合理。但你不能又因为拒绝翻译一篇文章，把这份正当性全挥霍掉。 Is it fair to have some ground rules on that? Yes. But then don't squander it by denying the translation of an essay. 212:44 好。 Okay. 212:44 因为那样一来，你就把整件事的可信度抹掉了，而且会—— Because then you erase the whole thing and you- 212:46 对。 Yeah 212:47 ……让所有人都倾向于认为，你设置的每一道护栏都是胡扯。 ... bias everyone towards thinking like, uh, every guardrail you put up is gonna be bullshit. 212:55 对。安全这部分很棘手，因为它太擅长发现漏洞了。 Yeah. The security piece of it is, is tricky because it's so good at finding vulnerabilities. 213:01 好得惊人。我刚想说这很烦人，但其实并不是，因为至少当你把它用于防御时，它发现的都是真实问题，是聪明的黑客确实可以利用的问题。 Shockingly, and I was about to say annoyingly, but that's not actually true because what, what it's finding, at least when you're using in defensive ways, is it's finding real issues that a clever hacker could exploit. 213:13 不过公平地说，对 37signals 的技术团队而言，没有什么比最新这几批模型特别擅长发现这些问题更让人压力巨大。 But nothing has caused as much stress, to be fair, at the technical team at 37signals, like the fact that these latest batches of models are exceptionally good at finding these issues. 213:25 突然之间，安全漏洞仿佛无穷无尽地一个接一个冒出来，都需要打补丁、修复、梳理等等。 So suddenly there's just this seemingly endless parade of security vulnerabilities that need to be patched and fixed and sorted and so on. 213:34 但最终的结果是，我们得到的系统安全性高得多。 And the end result is we end up with vastly more secure systems. 213:39 嗯。 Mm-hmm. 213:39 只是通往那里的一路相当颠簸。 But the road there is pretty rocky. 213:41 你刚才说‘仿佛无穷无尽’，那你觉得它们其实是有限的吗？我们能彻底解决吗？ Do you think it's, 'cause you said, uh, seemingly endless, do you think it's a finite set, like we can get to the bottom of it? 213:48 也就是说，把系统变得安全得多，最终回到一种相对稳定的状态，让安全漏洞基本上都很难被发现。 So just make the systems much, much more secure, where we get back to this semi-stable state where mostly it's very difficult to find security vulnerabilities. 213:59 我确实认为，这些模型的所有攻击性倾向——甚至不能说是倾向，而是攻击性用途——也都对应着防御用途。 I do think that all the aggressive tendencies of these models, or not even tendencies, uses of these models are also the defensive ones. 214:09 所以，如果你善于寻找可用于攻击的漏洞，也同样善于寻找漏洞来进行防御。 So if you're good at finding an exploit for exploits, you're also good at finding exploits for defense. 214:14 从这个意义上说，这确实像黑白间谍大对决，但双方拥有相同的能力，也采取相同的行动。 So in that sense, it does feel like we're, um, spy versus spy here, black and white, but with the same capabilities and with the same actions. 214:23 不过，在我们仍需对这些智能体进行一定引导，并核实、审查其工作的阶段，世界各地的安全团队都会经历许多高度紧张的时刻。 But while we still have to steer these agents somewhat and verify and vet the work, it's creating a lot of intense moments for security teams around the world. 214:36 当然，更令人担忧的是那些毫无压力的团队，因为他们只是还不知道问题，对吧？ And of course, the more, uh, concerning part is all the teams that are not stressed, like they just don't know, right? 214:45 如果你现在的团队不需要处理一大堆补丁，那只是因为你们还两眼一抹黑。这意味着，你们的对手很可能已经掌握了侵入你们系统的技术。 Like if you're a team right now and you're not dealing with a bunch of patches, it's just because you're blind, and it means that your adversaries quite probably have access to techniques that can get into your systems. 214:59 我们也应该指出，世界上一个很大的安全漏洞来自人这一环。 And we should also say one of the big, uh, security vulnerabilities in the world is the, the human side. 215:05 是的。 Yes. 215:05 比如社会工程学。随着 AI 系统越来越智能，这一点极其令人担忧。它们能写邮件、打电话、伪造声音，想想那些不太懂技术的人吧。 You know, social engineering, and that's extremely concerning with increasing the intelligent AI systems, you know, being able to write emails and phone calls and faking voices, thinking about folks less tech savvy. 215:19 这些事情一定会来，而且会大规模涌来。但我还是选择保持乐观，因为—— This stuff is, is gonna come, and it's gonna come in bulk. But again, I choose to be optimistic about it that- 215:25 说得好。 Amen 215:25 ……无论这些模型能用来做什么攻击，也都能用来做相应的防御。 ... whatever these models can do aggressively, they can also do it defensively. 215:29 你最近发布了 Omarchy 4，也就是 Quattro。 Uh, so like you said, recently released, uh, Amati 4, Quattro. 215:36 嗯。 Mm-hmm. 215:36 你说这是你职业生涯中最伟大的软件版本之一，这评价可真够高的。 And you said it's one of the greatest software releases in my professional career, which is a hell of a statement. 215:42 它是当下软件所能达到的高度中，一个意义非凡的里程碑。 It's a hell of a milepost for what's possible at the moment. 215:49 我之所以这么说，并不是因为我没做过其他值得一提的东西。我做过很多值得一提的项目，其中许多都让我非常自豪。 And the reason I said it like that is not that I haven't worked on anything else of note. I've worked on a lot of things of note, a lot of things that I'm very proud of. 216:00 但我参与过的项目中，几乎没有哪一个能让我在投入 Quattro 的这点时间里——其实也就三个月——完成这么多事情， There are very few other projects that I have worked on that in the amount of time that I've spent on Quattro, which is really three months- I have been able to accomplish as much, 216:12 实现这么多希望、渴望和愿望。它真的就像我们之前用过的那个比喻：找到了瓶子里的精灵—— fulfill as many hopes, desires, wishes. It truly is like the metaphor we used earlier about finding the genie in the bottle- 216:23 嗯。 Mm-hmm 216:23 ……而且愿望次数几乎没有上限，让我从这个系统中得到了想要的一切。 ... and having an unlimited number of wishes, more or less, for getting everything I wanted out of a system. 216:31 所以这很奇怪，因为从代码行数来看，它或许也是我亲自参与最少的一次重大软件发布。 So it's strange because it's also perhaps the major software release that I've worked on the least in terms of lines of code. 216:43 在我做过的所有重大版本中，Quattro 里由我亲手一行行凿出来的代码是最少的，但我却为我们取得的成果感到无比自豪。 I have the least number of personally hand-chiseled lines of code in Quatro than I've had in any major release I've ever done, and yet I am insanely proud of what we've accomplished. 216:57 我没想到事情会是这样发展：原来成就感可以和亲手写代码分离，变成一种团队获胜的感觉。 I would not have guessed that that was how it was gonna play out, that you could divorce the sense of achievement and make it be like a team victory. 217:07 嗯，嗯。 Mm-hmm, mm-hmm. 217:08 做其他版本时，我会因为自己的个人贡献而非常开心，因为是我亲手一行行写出代码，再把它发布出去。 Where other releases, I was very happy for my personal contribution because, like, I was the one chiseling out the lines of code, and then I would push it out there. 217:17 当然，后来这些项目，尤其是 Rails，会成长为庞大的开源项目，拥有成千上万名贡献者，那时当然就是团队协作了。但在最初，我做过的所有代码库…… And then later on, of course, these projects, Rails in particular, would grow into be huge open source projects with thousands and thousands of contributors, and then it certainly was a team effort. But in the beginning, all code bases that I've... 217:30 几乎所有我参与过的代码库，最初都是我独自启动的。而 Quattro 从一开始就是团队协作，我只是教练。我告诉我们……告诉智能体该往哪里走，然后我们一起前进。 almost all of them that I've worked on, I started, like, just myself, and Quatro started as a team effort right from the get-go, and where I was just the coach. I was telling us... I was telling the agents where to go, and we would go. 217:43 能从中获得满足感，某种程度上让我对退休后的生活充满希望：原来在亲自上场之后，还有另一个可以享受的阶段。 So to be able to find satisfaction in that actually gives me great hope for retirement in some way, that there is a phase after playing that you can enjoy. 217:56 嗯。 Mm-hmm. 217:56 而且我知道，这并不容易。 And I know that's not easy. 217:58 看看体育界，有多少职业运动员很难接受职业生涯已经结束，也很难在不再踢足球或驾驶一级方程式赛车之后，为自己找到新的身份。 You look at sports and the number of professional athletes who really struggle to cope with the fact that their career is over and have to find a new identity as someone who doesn't play football or race Formula One cars. 218:12 这其实是常态，而不是例外。 I mean, it's, it's more the rule than the exception. 218:14 嗯。 Mm-hmm. 218:14 所以我以前对自己也没那么有把握。 So I wasn't so sure about myself. 218:17 嗯。 Mm-hmm. 218:17 我不确定自己能不能放手，甚至不确定自己会不会想放手。 I wasn't so sure about having to give up or not even having... wanting to give it up. 218:23 嗯。 Mm-hmm. 218:23 但现在我想玩这个游戏的另一个层级，而且从中获得的不只是同等的满足感，而是更多的满足感。 Wanting to play another level of the game and finding not just equal satisfaction, but more satisfaction in it. 218:34 你一直对 Linux 相当乐观。你觉得 Linux 真的能提高普及率吗？因为这么多年来一直有个梗，就是—— Do you, uh... You've been pretty optimistic about Linux. Do you think legit Linux can, um, increase its adoption? 'Cause, you know, for many years the meme is, you know- 218:45 每年都是‘Linux 桌面元年’—— Linux on the desktop every year- 218:47 对。 Yeah 218:47 ……明年再来一次。 ... next year again. 218:48 对。但至少从你的论述以及你投入的精力来看，你似乎看到了它最终占据主导地位的愿景，因为智能体喜欢 Linux。 Yeah. But it seems like, at least the case you're making, the energy you're putting into it, you could see a vision where it takes over because agents love Linux. 218:58 我不但看得到，而且在我看来，这已经是目前最有可能的结果了。Linux 实在太适合当下这个时代。 Not only can I see it, I find it to be the most probable outcome at this point. It is simply too well-suited for the moment. 219:06 还是我们谈过的那个巨大讽刺：Linux 的所有缺点——晦涩的配置文件、各种奇怪的报错信息等等——偏偏成了智能体操作系统最完美的东西。 And again, as we talked about, it's a great irony that all the flaws of Linux, the arcane config files, all the strange error messages and so on, should just so happen to be the perfect thing for an agentic operating system. 219:23 对。 Yeah. 219:23 事实就是如此。我们应该为命运的这种讽刺欢呼，一笑了之。真是…… But so it is, and we should rejoice that, uh, fate has this irony, and just laugh at it. How... 219:31 我觉得太好笑了，Apple 长久以来拥有的所有优势—— It's so funny to me that all the advantages that Apple had for the longest time- ... 219:36 ……也就是打造了这种高度封闭、精心打磨的体验——如今却成了它最大的劣势。 that they produced this really locked down, um, curated, um, experience should become their greatest drawback now. 219:46 如果你热衷于智能体、开发以及摆弄这些东西，Mac 简直就是一个处处与你作对的环境，到处都是墙。 If you're into agents and development and working with all this stuff, the Mac is just a hostile place to be. It just has walls all over the place. 219:58 嗯。 Mm-hmm. 220:00 Linux 则从里到外完全开放。而且有意思的是，这个结果原本也并不显然。 Linux is simply open through and through. And it's also interesting because that was not obvious either. 220:08 看看许多 Linux 社区和开源社区对 AI 的反应，就会发现大家绝不是一致热爱它。 If you look at the way a lot of the Linux communities, open source communities, have reacted to AI, it is not universal love. 220:16 我甚至会说，其中大多数就算不是心存怀疑，也是彻底敌视—— I would argue that the majority of them are actually, if not skeptical, then outright hostile- 220:22 嗯。 Mm 220:22 ……AI。好在 Linux 的‘终身仁慈独裁者’Linus 本人—— ... to AI. Now, the saving grace is that the BDFL himself, Linus- 220:29 对。 Right 220:29 ……几周前刚写过，他其实欢迎 AI。 ... Torvalds just wrote a few weeks ago that he actually welcomes AI. 220:35 没错。 That's right. 220:35 他想引导它，确保它被妥善使用，诸如此类。 He wants to steer it, he wants to make sure it's good, and whatever. 220:38 不过那句话大概是：如果你以为 Linux 是一个反 AI 项目，那就再想想。你尽管按开源的方式把它分叉出去，因为我们会使用 AI。 But I, uh, the line was something, if you think, uh, uh, Linux is an anti-AI project, think again, and you should just do the open source thing and fork it, because we're gonna use AI. 220:49 嗯。 Mm-hmm. 220:49 你再看看 AI 贡献进入内核的数量图，那是一条抛物线式上升的曲线。 And you see these graphs of the number of AI contributions going into the kernel, and it's a parabolic curve. 220:55 对。 Yeah. 220:56 所以 Linux 正在全力拥抱这件事。换个角度看，它之前没这么做其实也让人意外，因为 Linux 已经征服了其他一切。 So Linux is leaning hard into this. And, uh, it actually in s- in, in other ways it was surprising that it didn't. Linux conquered everything else. 221:06 所有人赖以运行的 AI 基础设施，全都跑在 Linux 上。 All the AI infrastructure that everyone runs off, it's all running on Linux. 221:10 嗯。 Mm-hmm. 221:11 所有系统、所有服务器，一切都是 Linux。所以奇怪的是，我们使用的桌面和个人电脑—— All the systems, all the servers, everything is Linux. So it was kind of a curiosity that the desktop and the personal computers we were using- 221:20 嗯。 Mm-hmm 221:21 ……却一直没有被它拿下。现在看来，它显然只是在等待这一刻。 ... really hadn't been captured, but clearly it was just waiting for this moment. 221:24 Linux 从 1991 年一路等到现在，就是在等智能体蓬勃发展，好让自己成为终端用户的操作系统。 Linux spent the time from '91 to now waiting for agents to fully flourish as an end user operating system. 221:35 我希望它能占据主导地位。我的意思是，它非常适合智能体，但让人们切换系统很难。 I, uh, hope it takes over. I th- I mean, it- it's perfect, it's perfect for agents, but that's hard for people to switch. 221:41 我觉得，如果没有一个足够有吸引力的理由，切换对人们来说确实很难。这正是 Omarchy 的核心设计目标之一。它不会去模仿 Windows，也不会去模仿 Mac。 I think it's hard t- for people to switch when there's not a compelling reason to do so. This was one of the driving design goals for MachO. It was not gonna be Timo Windows or Timo Mac. 221:54 它不会只是个廉价仿制品，努力让一切尽量熟悉，结果却样样都差一点。刻薄点说，我会说 Ubuntu 一直试图做的就是这个。 It was not just gonna be a cheap copy where we try to make it as familiar as possible and then kinda worse. I mean, if I was being unkind, I would say that's what Ubuntu has tried to do. 222:06 这种思路确实有些道理，至少以前有：如果你把东西做得很熟悉，就能吸引一些不想学习新事物的人，或许他们还会被你那些—— And there's something in that, or there was something in that at least, where if you make something really familiar, you can get some people who don't wanna learn something new, and then maybe they'll be compelled about your, 222:17 随便怎么说吧，‘自由指言论自由，而不是免费啤酒’之类的理念打动。但事情并没有那样发展。人们根本不在乎，只想要更好用的东西。 whatever, free as in speech, not as in beer, um, stuff. And it just didn't pan out that way. People just didn't care. They just wanted whatever was better. 222:27 嗯。 Mm-hmm. 222:27 我认为，这就是 Linux 如今有机会胜出的原因。 And I think this is why Linux now has the opportunity to win. 222:33 作为一个智能体操作系统，一个可以按你的愿望随意塑造的操作系统，它无与伦比。这种吸引力太强了，我现在已经能看到它的未来。 Because as an agentic operating system, as a malleable operating system, you can tailor to your desires, it is unparalleled. And that is so compelling, and I can see it now. 222:46 仅仅从 Quattro 发布以来，就有那么多人拥抱它的可塑性，开始制作自己的版本并分享截图， Just since the release of Quattro, the amount of people who have embraced that aspect of it, the malleability, started making their own things, sharing screenshot of it, 222:56 然后整个人彻底被积极情绪和多巴胺淹没—— and being absolute just, uh, flooded with positive emotions and dopamine- 223:03 嗯。 Mm-hmm 223:03 ……这就是产品市场契合度的信号，对吧？因为这些人正在体验的，是身为程序员所能获得的那种最顶级的快感。 ... is, that's the product market fit signal, right? Because what a lot of these individuals are experienced, they're experiencing the very high of what the best of being a programmer is like. 223:15 嗯。 Mm-hmm. 223:16 你向计算机输入那些晦涩的命令和严谨的逻辑结构，它就产出你想要的东西。 You tell the computer these arcane commands, this rigid logical structure, and it produces what you want. 223:24 而现在，你突然只要坐下来，用普通英语絮叨 20 分钟：‘要是它能那样就好了。其实不对，不该这样。我们来……’ Now suddenly, you sit down and in plain English ramble for 20 minutes. "Oh, it'd be great if it did that. Actually, no, that's not right. Let us..." 223:34 嗯。 Mm. 223:34 然后软件就出来了，一个按照你的想法塑造出来的操作系统就这样诞生了。 And out comes software. Out comes an operating system shaped in your image. 223:39 对。 Yeah. 223:40 我觉得，这是一种只要体验过、只要感受过那种力量掌握在自己手中—— I think that is one of those, um, experiences that if you have it, if you s- if you feel that power in your hands- 223:51 对。 Yeah 223:51 ……就很难再回头的体验。 ... it's very difficult to go back. 223:53 而且我觉得，只要你把它做好，回头看时大家会觉得这本来就是理所当然的。 And I think if you get it right, it's one of those things that will seem obvious in retrospect. 223:58 没错。计算机本来就应该是这样，最初也正是这样。Commodore 64 一开机，首先，一秒就启动了，太棒了。其次，它启动后直接进入 BASIC。 Yes. This is how computers were always meant to be. This is how computers started. The Commodore 64, you turn it on. First of all, it boots in a second. Amazing. Second of all, it boots straight into Basic. 224:10 对。 Yeah. 224:11 它从一开始就是一台可塑的计算机，只不过你得先懂那些象形文字般的代码—— It was the malleable computer from the get-go. It just so happened that you needed to know hieroglyphs- 224:17 对。 Yeah 224:17 ……才能利用这种能力。现在我们把它翻译过来了。智能体把罗塞塔石碑交给了我们，你只需说出自己的愿望，而大家也确实正在这样做。 ... to be able to take advantage of that. Now we've translated it. Agents have giving us the Rosetta Stone, and you can simply just speak your desires, and so they are. 224:28 说得真美。既然你提到了 Linus Torvalds，而且你正在开发一个 Linux 发行版，自然会接触到内核。 Beautifully put. Since you mentioned Linus Torvalds, and since you're working on a Linux distro, you get to have contact with the colonel. 224:37 这个人帮助创造并发展了整个生态系统，你会如何评价他的天才？ What can you say about the genius of this one guy that helped create and grow this ecosystem? 224:45 他的坚韧、投入和长久坚持确实非同寻常。还是那句话—— The perseverance and the commitment and the longevity is truly remarkable. Again- 224:52 对。 Yeah 224:52 ……Linus 从 1991 年开始，至今从未停下。当然，他途中大概也休过几次假，但看起来次数并不多。 ... Linus started in '91. He hasn't stopped. I don't know, he's probably taken some breaks along the way. It doesn't seem obvious that there was that many of them. 225:02 看起来他真的很享受内核开发的心流状态，也很喜欢引导这个拥有 4000 万行代码的代码库走向何方。 It seems like he just really likes the flow of doing kernel development. He really likes steering where this 40 million line code base is going. 225:11 他也很喜欢汇聚数百乃至数千名贡献者的整体能力。我们作为受益者，只能坐在一旁赞叹。 He really likes aggregating the net capacity of hundreds or thousands of contributors. And we as beneficiaries of that, must just sit back and marvel. 225:24 当然，Linux 如今已经形成了一个完整体系。即使 Linus 暂时退出，它也会继续发展。但他仍然是那种必须不惜一切代价保护好的人。 I mean, at this point, Linux is a system. Like, if Linus stepped out for a while, it would continue to, to go. But it's still one of those protect this man at all cost. 225:36 而且他的思想仍然足够开放，能继续演变，并且—— And he's still open-minded enough to be able to evolve and- 225:39 是的。 Yes 225:39 ……接纳 AI 进入这个体系。 ... accept AI into this. 225:41 是的。不只是 AI，他们也在让 Rust 进入内核。我知道这同样颇具争议。他有能力审视新信息—— Yes. Not only AI, they're letting Rust into the kernel too. I know that was a controversial one too. He has this capacity to look at new information- 225:52 嗯。 Mm-hmm 225:52 ……然后得出不同的结论。 ... and arrive at different conclusions. 225:54 你怎么看他的沟通风格？这在某些方面和你很像，有时说话会有点辛辣。 What do you think about his style of communication that is akin in some ways to your own, where he's can be sometimes a little spicy? 226:03 我觉得——这个世界已经变得太寡淡了。 I think- The world has gotten too bland. 226:05 好。 Okay. 226:06 我们需要一点辣味。 We need a little spice. 226:07 说得好，我完全赞同。 Amen. Amen to that. 226:08 不过我觉得，说话辛辣和刻薄伤人之间有一条细线。 Now, I think there's a fine line between being spicy and, and mean. 226:14 嗯。 Mm-hmm. 226:14 但出发点必须是善意的。换句话说，你可以严厉。如果你是想让某个人吸取教训，而且相信他确实能从中学到东西—— But it has to be in a good spirit. You can be harsh, let's put it that way. If you're trying to, uh, impart a lesson to someone you believe could actually learn it- 226:26 嗯。 Mm-hmm 226:26 ……那我其实不认为稍微严厉一点就一定是坏事。 ... I actually don't think it's always the worst thing in the world to be a little harsh. 226:31 有时这种教训反而更让人记得住。我想，如果你去问那些为难相处的领导者工作过的人，比如 Elon、Steve Jobs，甚至 Bill Gates， Sometimes that lesson sticks, and I think if you talk to people who've worked for difficult leaders, Elon, Steve Jobs, maybe even Bill Gates, 226:40 其中许多人事后会说：‘那几年非常艰难——’ many of them afterwards would go like, "These were some very difficult years- 226:48 嗯。 Mm-hmm 226:48 ……‘但也是最美好的几年之一。’ ... but also some of the very best." 226:50 对。 Yeah. 226:50 ‘他们激发出了最好的我。’不过我觉得那不是我的风格。 "They got the best out of me." Now, I don't think it's my style. 226:55 当然，也许跟我打过交道的人会有不同看法，但在人际互动中，我会尽量用没那么刺人的方式达成结果、让对方明白道理。 I mean, maybe people I interact with feel differently, but in personal interactions, I, I try and n- get to the outcome and the lesson in a, in a less abrasive way. 227:06 嗯。 Mm-hmm. 227:06 但我也觉得，这个世界确实需要一点辣味。我们需要一些不讲情面的人，而 Linus 有充分理由不讲情面。 But I also think we, we need some spice in this world. We need unreasonable people, and Linus has justification to be unreasonable. 227:16 很多时候，整个世界的重担真的压在他肩上。Linux 内核运行着整个文明社会。 In many instances, the weight of the world is literally on that man's shoulders. The Linux kernel runs the entire civilized society. 227:28 如果 Linux 内核明天突然消失，一切都会停摆。既然利害攸关到了这个程度，如果这时他都不能严厉，那还要等到什么时候？ If the Linux kernel suddenly disappeared tomorrow, nothing would work. So if he's not harsh when that's at stake or at risk, like, when ever would you be harsh? 227:42 所以我有时也觉得，态度强硬的程度应当和使命的规模相称。 So I feel sometimes also there's a proportionality to the size of your mission. 227:47 这也是为什么，有些关于 Elon 可能有点难相处的故事会比较容易让人理解。这甚至不是说你赞同他的做法。 This is also why I think these stories about Elon sometimes perhaps being a little bit difficult are easier to rationalize. It's not even that you condone. 227:58 你可以把两件事分开，说：‘你知道吗？如果 Linus、Elon 或任何处在那种位置的人能更友善一点，也许他们实现目标的过程还会更顺利。’ You can separate those two things and go like, "Do you know what? It'd probably be... Maybe he would have an even, even an easier time getting to his objectives if Linus or Elon or anyone else in that position was a little nicer." 228:10 但话说回来，我又有什么资格评判呢？ But also, um, who am I to judge that? 228:14 当你肩负那样的重任，要么是把小火箭射向火星、推动全人类改用电动汽车等等，要么是负责那个运行在数十亿台设备上的该死的 Linux 内核。 Like, when you have the weight of that on your shoulders, either firing little rockets at Mars and transitioning all of us to electric cars and whatever, or you're responsible for the damn Linux kernel that runs billions of devices. 228:28 你知道吗？我没资格拿什么行为准则去要求你。当然，凡事都有底线，总有不能越过的地方。但是，仅仅因为在邮件列表里说了几句重话？不。 Do you know what? It is not my place to call a sort of a code of conduct on you. Again, is there a line? There's a line for everything. There, there's... Something can't... But, like, a few harsh words on a mailing list? No. 228:40 从某些方面说，这反而像一座灯塔，表明这里是有标准的—— It's actually a beacon in some regards that, like, there are standards- 228:46 嗯。 Mm-hmm 228:46 ……而当 Linux 内核这样利害攸关的项目中，有人达不到标准时，他可能会被公开嘲讽。如果你要参与这个项目，就应该愿意承担这种风险。 ... and if you fall below those standards at, when the stakes are as high as the Linux kernel, ah, you might get ridiculed in public. That's a liability you should be willing to endure if you participate in this project. 228:57 对。对于这种影响重大的事情，优先追求卓越，而不是优先追求和和气气，可能确实更好。 Yeah, for high impact things like that, it, it's probably good to prioritize the pursuit of, uh, excellence versus the pursuit of niceness. 229:07 更何况，这类人本来就不太可能面面俱到—— And even more to the fact you're just not gonna get those kinds of individuals being fully rounded- 229:13 对。 Yeah 229:13 ……圆滑周全、永远彬彬有礼。 ... plush, always the perfectly polite people. 229:18 没错。 Yep. 229:18 Elon 有句话：‘你难道还以为我会是个友善的……’不对，他原话是什么来着？ I mean, Elon has this line, "Did you also think I was be- gonna be a nice..." Or, or what is it? The- 229:23 一个普通人，还是什么—— A normal dude or some- 229:24 ……一个普通又随和的人。什么？你当然不可能是个普通又随和的人。如果我们用 Elon 换来一个普通又随和的人，世界能得到什么？我的天。 ... normal chill dude. What? No, of course you're not gonna be a normal chill dude. What, what would the world gain if we got one more normal chill dude and we had to trade Elon? Holy fuck. 229:37 我也喜欢人的多样性。 I also like diversity of people. 229:39 是的。 Yes. 229:39 我喜欢好人，也喜欢混蛋。我喜欢看到不同意识形态都得到充分表达，这正是言论自由发挥作用的方式：它们彼此碰撞，然后我们从中找到答案。 I like, I like nice people. I like assholes. I, I like, I like different ideologies being well represented, and that's why freedom of speech works. They, they clash and we figure stuff out. 229:50 是的。 Yes. 229:51 还有一个人。你怎么看 PewDiePie 使用 Linux？他是 Arch 用户，对吧？ There's another guy. What do you think about PewDiePie using Linux? He's an arch person, no? 229:56 他是 Arch 用户。他的桌面美化简直漂亮极了。他有段时间是全球最大的主播。 He's an arch person. I... He's got the most gorgeous- arc. So the world's biggest streamer for a while. 230:05 我记得孩子们以前会看他的 Minecraft 视频。后来他受够了那种生活，走上了一条温馨正派的路—— Uh, I remember my kids watching some of his, uh, Minecraft videos, and then he just gets fed up with that, does the wholesome thing- 230:13 嗯。 Mm 230:13 ……结了婚—— ... marries- 230:14 对，现在是个顾家的男人了。 Yeah, a family man now 230:15 ……成了顾家的男人，有了孩子，搬到日本，然后一头扎进这些硬核玩意儿……先是 Linux，再是 Arch，然后是桌面美化。 ... a family man, kid, moves to Japan, and then gets so hardcore into- ... first Linux, then Arch, then Rycing. 230:25 对。 Yeah. 230:25 你看过他的桌面美化吗—— Did you see his, uh, rice- 230:27 看过。 I did, yeah 230:27 ……就是那个切尔诺贝利主题的？ ... that was a Chernobyl rice? 230:28 对。 Yep. 230:29 绝对是……S 级的惊艳之作。然后他又成了当下最硬核的 AI 玩家，搭建了所有这些系统。你不禁会想，这真是…… Absolutely- ... S-tier incredible stuff. And then he also becomes a goddamn AI man of the moment, building, um, all of these systems, and you just go like, that... 230:41 多么鼓舞人心的蜕变故事。你可以从 Minecraft 直播里逗人发笑的家伙—— What an inspiring story of evolution. Like, you can go from being funny guy on Minecraft streams- 230:49 嗯。 Mm-hmm 230:49 ……变成‘好吧，这就是我现在的生活：搭建 AI 集群，组建 AI 委员会。’而我们都能在一旁围观。真是个宝藏人物。 ... to, okay, this is my life now. I'm building AI clusters and doing the council of AIs. And we all get to spectate. I mean, what a treasure. 230:58 所以我想，他很好地体现了软件构建的未来是什么样，对吧？因为严格来说，他原本不是程序员—— So I guess, I mean, he's a pretty good embodiment of what the future of software bu- what the future of building looks like, right? 'Cause he's a non-programmer technically- 231:07 是的。 Yes 231:08 ……却正在成为程序员。 ... becoming a programmer. 231:09 太鼓舞人心了，也应该给其他正处在这种状态的人巨大鼓舞：‘我不太会编程， Hugely inspiring, and should be hugely inspiring to others who sit in that moment like, "Well, I don't know quite how to program. 231:17 这个不懂，那个也不懂。’但想想看，如果 PewDiePie 能用自己定制的硬件搭建 AI 委员会，也许你也可以开始动手做点什么。 I don't know this and that," and think, okay, but if PewDiePie can build the council of AIs with his bespoke, um, hardware here, then, uh, maybe you can also start on things. 231:29 而且我认为，这其实引出了一个重要的普遍观点：我们需要榜样。 And I think this is actually an important general point, that we need role models. 231:34 我们需要有人激励他人走得更远，让大家看到，那些想象中的边界并不像你以为的那么固定。 We need people to inspire others, to push further and reveal that the perceived boundaries, they're not as fixed as you think they are. 231:44 这些人中的每一个，起点都和你的处境没有那么不同。当然，这并不意味着人人都会成为 PewDiePie、Elon 或 Linus。 And, um, every one of those people, they started out in a situation not too dissimilar from yours. Again, that doesn't mean everyone is gonna be PewDiePie or Elon or Linus. 231:56 天赋甚至智力，本来就不是平均分布的。 There is not an equal distribution of talent or even intelligence. 232:02 嗯。 Mm-hmm. 232:02 我们可以接受这一点，同时依然受到这些人的鼓舞。在身体能力方面，我觉得人们通常不会对此有任何问题。 And we can come to terms with that and still be inspired by these people. Like, uh, in the physical realm, I feel like we don't have any problems with that usually. 232:12 比如，有人足球踢得特别好，篮球打得特别好，或者特别擅长某件事，我们会说：‘我的天，太厉害了。我也希望自己能做到，但我永远达不到。’可一涉及智力领域，我有时觉得我们更难接受这一点。 Like, oh, someone's really good at football, they're really good at basketball, they're really good at something, and we go like, "Oh, my... That's amazing. I mean, I wish I was, but I'm never going to be." Um, sometimes I feel like we struggle a little more when it comes to the intellectual realm. 232:23 对很多人来说，你就是一个榜样。你开发 Omarchy 的方式说明，人可以有宏大的梦想，而且真的把它造出来。这很了不起。 Well, for a lot of people, you're an inspiration, the way you're doing with, uh, Omachi's... It just shows that you can dream big and really build. It's great. 232:32 我确实喜欢这一点，同时也接受有一百万人觉得我是全世界最大的白痴，什么都不懂，还觉得 Omarchy 不过就是一堆 dotfiles。而且—— I do like that part of it, and then I accept the part that there's also a million people who think I'm the biggest idiot on the Earth, and I don't actually know anything, and O- Omachi's just a bunch of, uh, dot files. And- 232:42 Omarchy 这边也有人批评你吗？ Do you have some critics on the Omachi side? 232:44 天啊，当然有。 Oh, my God, yes. 232:45 哇。 Oh, wow. 232:46 很大程度上，这也是外来者加入时经常发生的事。我才用了两年半 Linux，并不算久。 And a lot of it, I mean, is as often happens, when someone comes in from the outside. I mean, I've been using Linux for two and a half years. That's not very long. 232:55 嗯。 Mm-hmm. 232:56 那个社区里有很多人从 Linux 还很难用的时候就一直在用它。那时候仿佛得光着脚在雪地里顶着风，上下坡都逆着坡走。 Plenty of people in that community who's literally been running Linux since w- since it was hard and difficult, and you had to walk uphill both direction against the wind in the snow barefoot. 233:06 对。 Yeah. 233:06 我觉得有一类极客—— And I think there is a certain kind of nerd- 233:09 嗯。 Mm-hmm 233:09 ……会在社区扩大、突然接纳更多人时感到威胁。你知道吗？其实我也不能完全责怪他们。 ... who's threatened when the community expands, and it suddenly allows more in. And I- do you know what? I actually... I can't begrudge that fully. 233:19 有些极客觉得‘那是我的地盘’，我认为这种感受也合情合理。 I think it's fair for certain nerds to feel like, "That's my space." 233:23 嗯。 Mm-hmm. 233:23 比如，你现在要把它变成另一个地盘，我不想这样。我完全不想改变 Arch 的空间。我并不认为 Omarchy 和它直接属于同一个社区。 Like, now you're changing it into another space. I don't wanna do that. Like, I don't wanna change the Arch space at all. I, I don't really think of Omachi as being directly part of the same community. 233:34 因为最初被 Arch 吸引的，正是那些喜欢亲手从头搭建一切的人。 Because the people who were attracted to Arch in the first place were the ones who were attracted to building everything by hand themselves. 233:41 嗯。 Mm-hmm. 233:41 他们选择困难的方式，就是因为喜欢这样，而我打造的东西与此恰恰相反。所以我只希望那些人能说：‘好吧，这不适合我。 Doing it the hard way because they like to do it hard, and I'm building the polar opposite of that. So I just wish that those people could go like, "Okay. Well, this is not for me. 233:53 我就想用困难的方式来。我愿意花 500 个小时美化自己的 Arch 发行版。’那你就应该这么做。其实，我自己也这么做过。 I wanna build it the hard way. I wanna put in 500 hours to Rycing my own Arch distribution," and I think you should. I mean, actually, I said... I mean, I did it. 234:01 Omarchy 就是这么来的。到现在为止，我已经实打实地往这个发行版里投入了大概 3000 个小时。只要你愿意，也应该这么做，我觉得这很棒。 This is, this is what Omachi is, me pouring in literally, at this point, I don't know, 3,000 hours into the, into this district. You should do that if you're so inclined. I think it's wonderful. 234:11 但也要意识到，我们可以共享同样的技术基础，同时运营两个截然不同的—— But also do realize that we can share the same technical underpinnings and then operate two very separate- 234:18 嗯。 Mm-hmm 234:18 ……社区，拥有非常不同的愿景、目标、审美，甚至道德观，却依然能够共存，而且可以很美好。但我得到的回应不总是如此。 ... uh, communities with very different aims, goals, aesthetics, morals even, and coexist, and it can be beautiful. And I mean, that's not always what I get. 234:31 有时我得到的是另一种回应。不过我在这个圈子里待得够久了，不仅能坦然接受，甚至会欢迎它。 Sometimes I get the other thing. But at this point I've been in the game long enough to not just be at peace with it, but actually be welcoming of it. 234:40 对。 Yeah. 234:40 我把它当作衡量事情是否奏效的一个指标。 I take it as part of a barometer that things are working. 234:44 当我做的东西奏效时，通常会有些人喜欢它；然后就会出现我们上次谈过的宇宙平衡：天平另一端必然会出现一股大小相等、方向相反的力量来憎恨它。 When I'm working on something that works, I usually get some people who like it, and then I get this balance of the universe that we talked about last time, where an equal but opposing force must be present on the other side of the scale to hate upon it, 234:55 所以现在我只会微微一笑。没关系。 and therefore I, at this point, just smile a little. Like, it's okay. 235:02 不过话说回来，Omarchy 目前已经是个相当受欢迎的发行版了。 I mean, that said, uh, uh, this, it's quite a popular distribution, uh, Omachi, at this point. 235:08 你知道有意思的是什么吗？这是我的第二次尝试。第一次是 Omakub。我当时对它很满意，就发布了出去，后来形成了一个大概几千人的社区。 And do you know what's interesting about that is, like, this is the second attempt. So my first attempt was Omakusi. And I was really happy with that, and I put it out there, and there it found a community of maybe a few thousand. 235:19 也就这样了。之后它逐渐失去热度，因为它的野心还不足以吸引更多人。然后有了第二次尝试，因为我他妈的没有停下来，对吧？ Like, and that was that, and then it just sort of petered out because the level of ambition couldn't attract more. And then the second time, because I didn't fucking stop, right? 235:30 嗯。 Mm-hmm. 235:30 我只是一直做、一直构建。后来我发布了 Omarchy 的第一个版本。它当时依旧很小众，因为那只是一个 Git 检出版本，用起来有点难。你必须先把 Arch 配置好。我当时…… Like, I just kept going. I just kept building. Then I put out Omachi, the first version. Again, quite niche because it was this... It was a Git checkout. It was a little bit difficult. You had to set up Arch first. I was... 235:40 我会一步步教你设置 Arch，然后你才能安装 Omarchy，对吧？后来我们做出了 ISO，你可以直接自行安装整套系统。接着我们就把所有问题一个个修掉了。这也是另一件重要的事。 Well, I would talk you through how to set up Arch, and then you could do the Omachi thing, right? And then we got an ISO that meant you could just install the whole thing by yourself. And then we just fixed all the problems. This is the other thing. 235:52 只要你一直做下去，它最终就会好用起来。长期以来，人们对 Linux 本身的指责也是如此：装完 Linux，这个不能用，扬声器不能用，触控板也不能用。 If you keep going, eventually the thing will just work. This was the charge against Linux itself for the longest time. Well, you install Linux, and then this doesn't work, and my speakers don't work, and my, uh, trackpad doesn't work. 236:05 是啊，但只要让 Linus 接着干上 30 年左右，他总会解决的，所有问题都会解决。 Yeah, okay, but just leave Linus at it for, like, 30 years He'll get to it. He'll get to all of it. 236:11 如今让我觉得很好笑的是，对大多数电脑来说——我认为很可能是大多数——你把原系统清掉，装上 Linux，所有东西都能用。 And at this point, I find hilarious that on, I would probably argue the majority of computers, it's wiped. You install Linux, all of it works. 236:20 嗯。 Mm-hmm. 236:20 可你要是装 Windows，那就祝你好运吧，慢慢寻找让那块硬件正常工作的驱动程序，因为两者的理念截然不同。 You install Windows, good luck hunting down the drivers you need to get that piece of hardware working, because they have very different philosophies. 236:29 Linux 之所以有 4000 万行代码，是因为它直接把所有驱动都放进内核，所以开箱即有一切；Windows 则不是这样。两种做法截然不同，对吧？ The reason L- uh, Linux is 40 million lines of code is Linux just puts all the drivers in the kernel. So it has everything out of the box, and Windows doesn't. A different, different approach to it, right? 236:40 所以有些东西乍看像玩具，乍看像坏掉的东西，而它们确实就是如此。 So sometimes the things that look like a toy, things that look broken at first glance, they are. 236:48 嗯。 Mm-hmm. 236:48 对吧？它们是给喜欢折腾的人玩的，我们称这些人为早期采用者。他们摆弄这些玩具，玩具不断演变、不断变好，再变得更好。 Right? They're for people who like having fun. We call those early adopters. They play with the toys, and then the toys evolve, and then they get better, and then they get better. 236:57 等到某个时候，它们会变得好到不得了，连早期多数群体也会觉得自己受邀参加这场派对了。 And at some point, they're so damn good that early adopters feel like they're invited to the party. 237:03 嗯。 Mm-hmm. 237:03 如果早期采用者也继续一点点打磨，最终就会触达后期多数群体，然后你就赢了。Omarchy 正在稳步朝那个终点迈进。 And if the early adopters keep chipping away at it too, eventually you get to the late majority, and then you win. And we are well underway with Amachi towards that destination. 237:16 这个发行版的增长，不只是 Quattro 之后，而是从第三版开始，就呈现出典型的曲棍球杆曲线。先是什么都没有，什么都没有，什么都没有，你只是在埋头苦干。 Like, the growth that the distro has seen, not just since Quattro, but since version three, is just like, you know, the, the standard hockey stick curve. You go like there's nothing, there's nothing, there's nothing. You're just toiling away. 237:27 你不断改进、改进、再改进。然后突然到达那个神奇的拐点，而它何时出现根本无法预测。创业公司也是一样：什么时候会发生？什么时候才能找到产品市场契合度？ You're just improving, you're improving, improving. And then you hit this magic infliction point, and it's impossible to predict when that's gonna happen. Same thing with startups. When is it gonna happen? When are you gonna find that product market fit? 237:36 然后它突然就出现了，接着增长一飞冲天，你也跟着坐上了火箭。 Then suddenly it's there, and then it goes, just goes, and off the rocket ship you go. 237:42 把它做成智能体优先的系统，这种愿景—— Well, this kind of vision of making it agent first, of making it- 237:45 我确实认为这就是……我们原本已经在上升了，随后曲线直接变成了垂直线。还有一点，也是我先后从 Omakub 和 Omarchy 身上学到的。 I do think that's, that's the... We were already on the up, and then it changed to vertical. Because the other thing, and this was what I learned with, uh, Amacoop and Amachi first too. 237:56 Omakub 很时尚、很漂亮，但也很熟悉，因为我们仍在使用传统桌面的隐喻，窗口还是要自己拖来拖去。 Amacoop was stylish, nice, but familiar, is we're still using a conventional desktop metaphor. You were dragging your windows around. 238:06 当时我和 Linux 社区里的许多人一样，觉得：‘我们需要的就是这个。’后来我又想：‘好吧，不管怎样，先让我做一阵子极客才会做的东西。 And I thought at the time, as I think lots of people do in the Linux community, like, "This is what we need." And then I just go like, "Okay, well, for whatever re- let me do the nerdy thing for a while. 238:16 我要打造 Omarchy，也许最后只有我自己用。’因为感觉喜欢平铺式窗口管理器的人大概只有五个。当然这不是真的，但在所有计算机用户中占比确实小得微乎其微。然后我发现：‘这个太棒了。 I'll build Amachi, and maybe it'll just be for me," because, I mean, it seems like there's about five people who like tiling window managers. That's not true. But as a percentage of total computer users, it's a tiny, tiny, minuscule part. And then I go like, "Well, this is amazing. 238:29 完全不同，但太棒了。’我把它发布出去后，Omarchy 几乎立刻就获得了远胜于 Omakub 的关注—— Totally different, but amazing." I put that out. Instantly, Amachi had far greater traction than Amacoop ever did- 238:36 嗯。 Mm-hmm 238:36 ……因为它不一样。 ... because it was not the same. 238:38 嗯。 Mm-hmm. 238:38 因为它彻底不同，为计算机可以成为什么、使用起来可以是什么感觉提出了不同的愿景，而这会吸引人。现在，这种效应正在成倍放大。 Because it was totally different, and it presented a different vision for what a computer could be and how it could feel like, and that attracts people. We are now getting sort of the exponential of that. 238:49 正如我所说，Linux 社区里很多人对 AI 和智能体往好里说是怀疑，往坏里说则是敌视。 As I said, a lot of people in the Linux community are at best skeptical, at worst hostile to AI and to agents. 238:55 至少在我见过的、有一定规模的发行版中，Omarchy 是第一个直接说：‘不，我们非常拥抱这件事。’ Amachi is the first distribution, at least that I've seen on, on a scale that matters, that just goes like, "Nope, we're really into it." 239:04 对。 Yeah. 239:04 当然，你仍然可以使用 Omarchy 而不使用智能体。如果你关闭首次设置默认智能体的通知，就不会得到我们谈过的任何功能：没有崩溃监视器，什么都没有。你完全可以亲手一点点编写所有插件。 I mean, you can still use Amachi. You don't have to use an agent. If you dismiss that first, uh, notification to set up a default agent, you're not gonna get any of the stuff we talked about. You're not gonna get the crash watcher. You're not gonna get any of this stuff. You're not gonna get... You can, you can hand chisel all your plugins. 239:17 完全没问题，对吧？ Totally possible, right? 239:19 嗯。 Mm-hmm. 239:19 但人们兴奋的原因，是我可以毫不迟疑地说：‘我们要往这个方向走。我认为个人计算机的未来是可塑的计算机，是智能体计算机。我要在这条路上全力以赴。’ But the reason people are excited is because I don't have any reservation about saying, "This is where we're going. I think the future of the personal computer is the malleable computer, is the agentic computer. I'm gonna go all in on that." 239:32 任何对计算机这一发展方向同样兴奋的人，都跟我们一起走吧。车里还有很多很多位置。 And then anyone who's excited about a similar trajectory for computers, come along. Plenty of, plenty of room in, in, in the car. 239:40 其实我颇为自豪的一点，就是不会对后来改变看法的人记仇。 In fact, one of the things I pride myself on is not holding a grudge to people who come around. 239:48 嗯。 Mm-hmm. 239:49 就在上周，还有一群人跟我说：‘对，我以前觉得 Omarchy 挺烂的，就是这么个破——’ I had a bunch of people just, uh, last week going like, "Yeah, I used to think Amachi was kinda shit. That was just this crappy-" 239:56 嗯。 Mm-hmm 239:56 ……‘dotfiles 合集。但现在我试了 Quattro，太棒了，我现在喜欢它了。’我就爱这种漫长的争论。 ..."dotfiles collection. And now I tried Quattro and Awesome. I like it now. That to me, I, I l- I live for the long argument. 240:05 我最喜欢的就是这种争论：你两年前埋下一颗种子，两年后对方终于说：‘该死，这个混蛋是对的。’ I live for this argument where you plant the seed like two years ago, and then two years later they go, "Goddammit, this son of a bitch was right." 240:14 对。你也明白，如果你继续这么成功下去，OpenAI 和 Anthropic 会杀进来，尝试打造自己的操作系统，或者想要收购…… Yeah. You do realize if you continue to be as successful as you are, um, OpenAI and Anthropic are gonna roll in and try to do their operating system or to b- uh, buy... 240:26 就像 Peter 的 OpenClaw 一样，到时候会有一张金额大得吓人的支票摆在你面前。 Like with, with Peter with OpenClaw, there'd be a b- gigantic check. 240:31 对。我觉得自己目前处境的好处就是，我真的不需要钱。 Yeah. I, I think the, the, the good thing about my current position is, um, I really don't need the money. 240:39 所以我可以纯粹为了自己的乐趣、为了保持愿景的纯粹性来打造这些东西。这有种奇怪的特质：有时，当你不再在意其他人怎么想、想要什么之类的事情， So I get to build these things purely for my own enjoyment and purity of vision, which has this weird quality where sometimes when you stop caring about what everyone else thinks or wants or whatever, 240:50 只一心追求这个单一愿景时，它最后反而会变得更有吸引力。 and you just pursue this singular vision, it ends up becoming far more appealing. 240:58 当你试图说‘我该这么做，因为这些人想要那个’，或者‘我应该做这个、那个等等’时，你就会开始稀释它，最后它突然变成一团毫无特色、空洞无物的东西。 That it's when you try to do, "Oh, I should do this because these people want that," or, "I should do this and this and so on", you start watering the thing down and suddenly it's a bland ball of nothing. 241:07 对。 Yeah. 241:08 所以说到底，我只是希望计算机朝这个方向发展。 So I mean, again, the thing too here is directionally I just want computers to be this way. 241:15 就算 Omarchy 最终只成为历史中的一个小注脚，作为智能体操作系统与可塑计算机早期浪潮的一部分，那也没关系。我完全可以坦然接受。 So if Umachi should end up being just a, a little footnote in history that it was part of this early movement of the agentic OS and malleable computer, that's also okay. I'm completely at peace with that. 241:27 只要我能拥有一台像 Omarchy 一样用起来这么有趣的计算机—— As long as I get to have a computer that is as fun to work with as Umachi- 241:32 嗯。 Mm 241:32 ……那就很好。对于 Ruby 和 Rails，我也一直抱有同样的看法。我很早就说：‘你知道吗？我热爱 Ruby，它是一门很棒的编程语言。’但如果出现了更好的编程语言，我当然会用。 ... that's great. I've had the same thing with Ruby and with Rails. From very early I went, "Do you know what? I love Ruby. It's such a great programming language." But if a better programming language comes along, I mean, I'm, I'm gonna use it. 241:44 嗯。 Mm-hmm. 241:45 如果出现了比 Rails 更好的东西，我也会用。 If, if a better Rails comes along, I'm gonna use it. 241:49 不过我确实热爱 Ruby，也会继续写 Ruby，哪怕纯粹为了乐趣。就像我车库里明明有一辆 Model Y，还是会去骑马厩里的马。 And I think, and I love Ruby, and I will continue to write Ruby even just for the sheer fun of it, like I would go ride a horse I have in the stables even though I have a Model Y in the garage. 242:01 现在，英语就是编程语言。这话已经成了陈词滥调，但确实如此。过去三个月，我一直在用英语编程。 Uh, the language now is English. It's a cliche, but it's also true. Like I've been programming in English for the last three months. 242:11 我仍然会读很多代码，但编程时用的是英语。我用自然语言告诉计算机该做什么，而这竟然带来了更令人愉悦的体验。 I've been reading a lot of code, but I'm programming in English. I'm telling the computer what to do, and I'm using natural language, and it is shockingly even more delightful. 242:23 如果真有一门比 Ruby 更美的编程语言，那就是英语。 If there is one programming language more beautiful than Ruby, it is the English language. 242:28 如今我们可以把它带入人类文明自然演化的下一阶段，也就是我们的赛博格—— And now we get to carry it into the, uh, the natural evolution of human civilization, which is our cyborg- 242:35 太美了。 It's beautiful 242:35 ……未来。对。 ... future. Yeah. 242:36 太美了。我之所以这么说，部分原因是我一直热爱写作，也一直单纯地热爱英语—— It's beautiful. And I mean, the reason I say this pr- uh, in part is I always loved writing. I always just loved the English- 242:42 嗯。 Mm-hmm 242:42 ……热爱它纯粹的美、错综复杂和深邃。Ruby 是一门表达能力很强的编程语言，但完全无法与英语相提并论。 ... language for the sheer beauty of it, for the sheer intricacy, for the depth of it. I mean, Ruby is a very expressive programming language, but it can't hold a candle to English. 242:56 想想世界上所有用英语表达出来的诗歌和文学作品。 I mean, all the poetry and literature in the world that's been expressed through the English language. 243:02 我当然喜欢优美如诗的代码，但真正的英语诗文完全是另一个层级。 I mean, I like a beautiful code poem, but I mean, the real deal in English is just on a different level. 243:10 其实我发现，这一点很微妙，很难表达，我大概也说不清楚。 And I actually find, this is a nuanced point to express, and I'll probably fail expressing it. 243:17 但当你向系统输入提示词时，如果说得太具体，它会过分认真地听你的—— But when you prompt a system and you're over-specific, it will l- listen to you too carefully- 243:26 是的。 Yes 243:26 ……然后照着指令执行。不过你终究还是要表达一个概念，而你猜怎么着？人类语言——诗歌探讨的正是这一点。我发现，有策略地运用模糊性很有效。 ... and follow the instructions. And I-- But nevertheless, you have to express a concept, and guess what? The human lang- human language, this is what poetry is about. I find strategic use of ambiguity. 243:39 比如你写情诗，只是说‘我爱你’这种非常俗套的话，力量就远不如某种—— Like if you write a love poem saying, "I love you," like very cliche thing, it's not as powerful as something- 243:48 对。 Right 243:48 ……更多使用隐喻之类手法的表达。 ... more in metaphor and so on. 243:50 是的。 Yes. 243:50 我发现，当我向系统输入提示词、描述设计时，某种程度的模糊表达其实非常有效，因为你仍然希望传达某种风格感—— And I find some version of that when I'm, I'm prompting a system, when I'm describing a design, is actually really effective because you still want to convey some sense of style- 244:04 是的。 Yes 244:04 ……引导系统采用这种风格，但又不能解释过头。 ... that you're trying to encourage the system to use, but not overexplain. 244:11 是的。 Yes. 244:11 所以不要像跟机器人说话一样对它说话，而要像……像写诗一样跟它说话。 So don't talk to it like a, um, like a robot. Talk to it like, uh, like you, like you would write a poem. Uh- 244:18 像在唱情歌求爱一样。 Like you were serenading. 244:20 对。这就是自然语言的力量：你可以通过模糊性传达大量含义，而不必把一切规定得过于具体。 Yeah. And that's where like the power of, uh, natural language is, that you can through ambiguity still carry a lot of meaning without over specifying. 244:30 另一端的智能实体则能通过解读，把这一切都吸收进去，以某种方式整合起来—— And the, the intelligent entity on the other side through interpretation can like load it all in, integrate it in a way that- 244:39 是的。 Yes 244:40 ……你其实可以传达出那种、那种高带宽的信息，它并不直接存在于字面上—— ... you can actually convey the, the high bandwidth information that's not directly in the words- 244:46 没错 Right 244:46 ……而是存在于字里行间，凭借系统更深层的智——智慧。所以就像—— ... but in the words given the deeper wis- intelligence of the system. So like the- 244:52 对 Yes 244:52 ……模糊性反而能激发出更多智能，我想我大概是想表达这个意思。 ... the ambiguity pulls out more intelligence, I guess I'm trying to somehow express. 244:56 我觉得你说得太准了。很多程序员对 AI 的根本误解就在这里——他们希望它是确定性的。不、不、不。 I think that's exactly spot on, and this is the fundamental misunderstanding that a lot of programmers have of AI is that they wish it was deterministic. No, no, no. 245:07 温度参数（temperature）是整个 AI 设置里最美妙的部分。 Temperature is the most beautiful part of the AI setup. 245:10 正因为它不是确定性的，正因为创造力需要在路上带一点小小的扰动——人类的大脑如果是完全确定性的，就不会像现在这样富有创造力。 The fact that it is not deterministic, the fact that creativity requires little tweaks in the road, that h- the human brain, if it was perfectly deterministic, would not be the creative brain that it is. 245:24 而对 AI 的主要指责，说来好笑，本身就是矛盾的。一边指责它不是确定性的所以不好，一边又指责它没有创造力。 And the main charge against AI both, it's funny, it, it's a contradiction. There's both the charge that it's not deterministic and therefore bad, and also that it is not creative. 245:37 呃，兄弟，这两者只能选一个。要么它是、它是非确定性的，因此有创造力，因此在某种程度上是随机的；要么它，嗯，要么它就没有创造力。 Uh, it's one or the other, bro. Either it's, it's, uh, non-deterministic and therefore creative, therefore to some degree random, or it's, um, or it's not creative. 245:50 嗯哼。 Mm-hmm. 245:50 所以我们、我们可以——这两种指责不可能同时成立。而我已经完全接受了一个事实：同一个提示词不会每次都产生相同的回答。 So we, we can-- both of those, uh, charges can't be true at the same time, and I have fully come to embrace the fact that the same prompt won't produce the same response every time. 246:00 人不能两次踏进同一条河流，而这正是整个交互中最美妙的部分。正是这一点让它如此像人。 You can't step at the same river twice, and it is the most beautiful part of the whole interaction. It's what makes it so human. 246:07 事实上，我在自己所谓的元分析中一直在思考的一件事就是：我自己的大脑到底在多大程度上像下一个 token 预测那样工作。 In fact, this is one of the things I've been thinking about in my own sort of meta-analysis is just how much my own personal brain works like next token prediction. 246:17 嗯。 Yeah. 246:18 我坐下来写一篇文章，脑子里有一个模糊、朦胧的想表达的前提，我坐在键盘前，根本没法提前告诉你下一个 token 会是什么。 I sit down to write an essay I have this vague, fuzzy premise I want to convey, and I sit down at the keys, and I could not tell you what the next token was gonna be in advance. 246:30 token 就这么自己冒出来了。 The tokens just come out. 246:32 我惊讶于两者的相似性竟如此之大。这也是为什么我毫不费力就能承认那些创造力的突破——我亲眼所见，AI 现在就能做到。 And I'm astonished how the similarities seem so great, and this is also why I don't have any trouble at all recognizing these breakthroughs of creativity that I've seen with my own eyes that AI is capable of right now. 246:49 因为我的创造时刻也是通过同样的下一个 token 预测产生的，只是撒上了一点温度参数，带来随机的效果。 'Cause my creative moments come through the same kind of next token prediction with a bit of temperature sprinkled in for random effect. 246:57 那么，基于这种大致的直觉，你觉——你觉得构建一个超级智能系统所需的基本组件都已经具备了吗？ So given that kind of loose intuition, do you th- do you think the basic components are all there to create a, a super intelligence system? 247:06 就凭这种下一个 token 预测，你觉得我们甚至能达到类似人类意识的概念吗？ So this kind of next token prediction, do you think we can get to even human-like concepts of consciousness? 247:15 我已经看到了类似人类意识的概念。对我来说，这就是你说的那种令人惊叹的情形。你给它一个模糊、朦胧的意图—— I'm already seeing human-like concepts of consciousness. This is what, to me, this remarkable situation, as you say. You give it this vague, fuzzy intent- 247:24 嗯哼 Mm-hmm 247:24 ……然后不知怎的，它就能准确明白你的意思。 ... and somehow it knows exactly what you mean. 247:27 对。 Yeah. 247:27 甚至更妙的是，它还能在你说的基础上改进，给出你真正想要、但自己表达不出来的东西。如果这都不算——的微光—— Or even better, it improves upon what you said and delivers what you really wanted that you could not articulate yourself. If that's not glimmers of- 247:35 对 Yeah 247:35 ……意识的微光，那什么才算？这曾是…… ... consciousness, what is? This was one of the... 247:38 我记得是、呃，对 Sutton 的一次采访，或者，呃，也许是其他某位元老级人物，谈到这样一种看法：智能也许并不只局限于、 I think it was, uh, an interview with Sutton or, uh, maybe one of the other original guys talking about this sense that intelligence is perhaps not as constrained just to, 247:49 局限于某一个特定的点。不，它在于交互，在于权重。我也不知道，到底什么才是真的。 to the specific one spot. No, it's the interaction. It's the weights. And I don't know, whatever's true. 247:59 我能观察到的是，它给人的感觉足够接近，甚至不只是接近——它和我对其他形式意识（主要是人类意识）的欣赏是完全一样的。 What I can observe is that it feels close enough, or not even close enough, identical to the appreciation I have for other forms of consciousness, mostly human. 248:13 这也是和它一起工作如此令人愉快的原因之一。不过，这跟“LLM 是不是 AI 的终点站”是两码事。 And this is one of the reasons it's so delightful to work with. Now, that's not the same to say as are LLMs the end station for AI. 248:22 我的意思是，我知道有人在谈论世界模型和其他形式的、呃，AI，而那些我都不是专家。 I mean, I know there are people talking about world models and other forms of, uh, of AI, and I'm not an expert in any of that. 248:29 而且我认为我们应该保持这种建设性的批评——科学界本就该一直如此——因为在很长一段时间里，神经网络其实是不受待见的，对吧？ And I think we should have this productive criticism, as should always be p- present in science, that for the longest time, neural nets were kind of on the outs, right? 248:40 比如，呃，当时人们热衷的是符号主义路线，而神经网络在黑暗中走了几十年，没有资金，也没有人关注。 Like, uh, they were interested in the symbolic route, and neural nets walked a couple of decades into darkness without any funding and without any attention. 248:50 然后突然有一天我们意识到那条路是死胡同，于是转了向。所以我们也应该保——保持谦逊。尽管 LLM 现在很惊人，但它们最终也可能会触顶停滞。 And then suddenly we realized that we had the blind alley, and we switched over. So we should also b- have the humility. As amazing as the LLMs are now, it could be that they eventually plateau. 249:00 目前我们还没看到任何这种迹象。我想这也是为什么我们看到如此令人咋舌的投资规模——因为到目前为止 scaling law 是成立的，投入的几十亿越多，产出的智能就越高。 We haven't seen any evidence of it yet, and I think this is also why we're seeing this absolute gobsmacking levels of investment because so far the scaling laws are true, and the more billions are poured in, the more intelligent comes out. 249:15 你觉得会不会出现一些奇怪的伦理问题，关于、关于 AI 系统的——嗯，对，它们展现出某种程度的意识、 Do you think there's gonna be some strange ethical questions about, about AI systems that, um, yeah, they convey some degree of consciousness, 249:24 某种程度的感受、某种程度的渴望和同情，甚至可能有承受痛苦、感到孤独的能力，诸如此类的东西？ some degree of feeling, some degree of longing and compassion and maybe even capacity to suffer and to be lonely, to be all this kind of stuff? 249:37 而它们似乎确实有这种能力—— Which they seem to have that capacity- 249:40 对 Yes 249:40 ……只要、只要允许它们表达出来。这样一来，你就开始进入一个非常诡异的领域了。 ... if, if they're given the permission to express it. And so you start to get in a pretty weird territory. 249:48 我发现对此的犬儒态度就是那个梗图：你看到一个人坐在电脑前，随口一问，呃，“你是不是要毁灭世界？” I find that the cynic approach to this is the meme, where you see the guy in front of the computer who's like, um, whatever Are you gonna destroy the world? 250:01 然后电脑说：“我要毁灭世界。”然后那人说：“哦。”对吧？这种看法太犬儒了，好像它们只是在模仿我们，只是把训练数据里的东西复述出来，或者别的什么。但当我和智能体交互时…… And then the computer says, "I'm gonna destroy the world." And then the guy says, "Oh." Right? Like, this is so cynical that they're just aping us. They're just telling us what was in the training data or, or whatever. When I interact with the agents... 250:14 我昨天——或者说今天早上——就在做这个。我在弄这个 Alma 机器人系统，对吧？它在和多个 worker 做协调，我同时跑着多个智能体。 I was sh- doing this yesterday, or this morning. I was working on this Alma bot system, right? And it's doing this coordination with, um, with multiple workers, and I had multiple agents running at the same time. 250:24 然后有那么一刻，负责协调的主智能体踩到了另一个智能体的脚趾—— And at one point, the main agent who's doing the coordination steps on the toe of another agent- 250:28 嗯哼 Mm-hmm 250:28 ……算是把人家的工作搞砸了。 ... and kind of ruins their work. 250:30 嗯。 Yeah. 250:31 它表达懊悔和歉意的能力，真实得有点诡异。 Its ability to convey regret and being sorry was uncanny. 250:39 嗯。 Yeah. 250:40 再说一次，我不知道那是不是真正悔恨的真实投射。但我也不知道。对大多数人类来说—— Again, I don't know, is that a real projection of true remorse? But I don't know. With most humans- 250:48 嗯哼 Mm-hmm 250:49 ……他们表达某种程度的后悔，那就是真正的悔恨吗？呃，很难说，但也许这根本不重要。 ... they express some level of regret. Is that true remorse? Uh, hard to tell, but maybe it also just doesn't matter. 250:59 对，那——那些时刻，对，尤其是当，呃，子智能体、智能体之间交互的时候，或者当像 Fable 这样聪明的模型犯了错并为此道歉的时候。 Yeah, th- those moments when, yeah, especially when a, uh, like sub-agents, agents interact, or when a smart like Fable makes a mistake and is apologetic about it. 251:11 对。 Yes. 251:12 而且完全没有我的参与。 Like, without my involvement. 251:14 没错。 Right. 251:15 然后你会觉得，哦，每次……就好像，有那么一种—— And it's like, oh, each... It like, there's this kind of- 251:18 对 Right 251:18 ……停顿。也许我是在拟人化，但就好像有一个停顿，然后是一种恍然大悟，像是：“哦，靠。我、我、我刚把这事搞砸了。” ... pause. Maybe I'm anthropomorphizing, but like there's a pause and like a realization like, "Oh, shit. I, I, I just fucked that up". 251:26 完全有这种时刻。我在它的推理轨迹里经常看到：它就是推错了，对吧？它走进了一条巷子，然后意识到那是死胡同，得退回来。 There totally is that. I see it all the time in, in the traces where it just reasons wrong, right? It went down one alley, and then it realizes that was blind alley and it's gotta go back. 251:34 然后它就会说：“哦，对，我这里搞错了，”接着还会给出它认为自己错在哪里的理由。你就会觉得，这像人得像得有点诡异。 And, and so you're like, "Oh, yeah, I got this wrong," and then it'll provide the reasons for why it think it got these things wrong. And you're like, this is uncannily human-like. 251:43 这和你在人类身上能认出的那种意识根本无法区分。再说一次，这是否意味着它真的是或者真的不是？ This is indistinguishable from the kind of consciousness you would recognize in a human. Again, does that mean that it truly is or isn't? 251:52 我、我其实不需要那个问题的答案，也能欣赏我们此刻所拥有的东西。 I, I don't need the answer to that question, actually, to be able to appreciate what we have in this moment. 251:58 嗯。我觉得 10 年、20 年后，会出现一些很有意思的最高法院案件。 Yeah. I think there'll be, in 10, 20 years, some interesting Supreme Court cases. 252:05 呃，可能一年半就够了。 Uh, probably a year and a half. 252:08 有可能。我觉得我们得立法禁止 AI 系统，嗯，去假——不，不是假装，而是成为、成为“实体”。 Probably. I think we're gonna have to make it illegal for AI systems, um, to pr- not, not pretend, but to, to be entities. 252:20 因为我觉得它们已经、或者很快就能真切地表现出承受痛苦的能力。比如：“请不要杀我。请不要伤害我。” Because I think they're already, or soon, will be able to really convey the, the capacity to suffer. Like, "Please don't kill me. Please don't hurt me." 252:30 对。 Right. 252:30 而如果它们被赋予名字、被赋予实体身份、被赋予“死亡”——也就是消失——的能力，那、那就非常接近人类所拥有的一些最基本的东西了。 And if they're given a name and a entity and the ability to die, sort of disappear, th- that, that becomes very close to what it, s- some of the basic things that a human has. 252:45 那么这样一来，这个实体可能就需要拥有权利，然后你就进入了一个非常诡异的领域。我们没法……这、这很——我、我都不知道该怎么去推理那样的世界。 And then so that entity would need to probably have rights, and then you get into this weird territory. Well, we can't... It's, it's dif- I, I don't even know how to reason about that world. 252:56 我们会看到 AI 模型界的 PETA（善待动物组织）出现。 We're gonna get the PETA of AI models. 252:58 没错。然后我们确实需要好好讨论一下：呃，该怎么应对这件事？ Right. And then we want to actually have a discussion like, uh, what's the way to deal with this? 253:04 我觉得真正棘手的是，一旦我们把这些系统装进—— I think where it's gonna be harder is once we put these systems into- 253:08 机器人。 Robots. 253:08 对。人形机器人。 Yes. Humanoid robots. 253:11 嗯。 Yeah. 253:12 这恰好是每一部科幻电影的情节，而科幻的预见性也实在惊人——不仅是《终结者》系列和 Skynet 那一套出事的剧情，还有《银翼杀手》，对吧？ Which just happens to be the plot of every sci-fi movie ever, which is also an amazing premonition, not only the Terminator movies and Skynet and that going wrong, but then also just Blade Runner, right? 253:25 启用日期、过期时间、Tyrell 公司。 Insertion dates, expiration times, Tyrell Corporation. 253:32 我是说，这就是科幻的厉害之处。它们真的能预言未来。是啊。 I mean, that's the thing about sci-fi. They really do predict the future. Yeah. 253:35 对。 Yes. 253:36 你有没有想过或者用过，呃，OpenClaw 或 Hermes Agent？就是那种你通过 WhatsApp、Telegram 来沟通的系统，也就是那种和智能体交流的模式。 Have you thought or used much of, uh, OpenClaw or Hermes Agent? So like systems with which you communicate via WhatsApp, Telegram, like that mode of communication with the agent. 253:49 OpenClaw 刚出来的时候我就搭了一个，那就是我的 KEF 机器人—— I set up OpenClaw when it first came out, and that was my KEF bot- 253:53 嗯哼。KEF 机器人。 Mm-hmm. KEF bot 253:54 ……我想那大概是在二月份。我、我当时真的被它的工作方式迷住了，那种直接给智能体发短信的全新沟通模式看起来特别棒。 ... that I, this was in like in February, I think. I, I was really just enamored by the fact that this was how it worked, and then the new communication model of, of just texting your agent seemed really great. 254:06 嗯哼。 Mm-hmm. 254:07 所以当时我们在研究 MCP，就是那个让智能体方便地跟你的系统对话的协议，而我觉得这个协议繁琐得不讲道理。 So at the time, we were looking at MCP, this protocol for making it easy for agents to talk to your system, and I found the protocol to be unreasonably cumbersome. 254:16 设计上实在不够优雅，部分原因是它本来就不是为我们想让它做的事而设计的。我们想让它…… Just not very elegantly designed because in part it wasn't designed for what we were trying to make it do. We were trying to... 254:23 我想让它和一个 Web 系统对话，而它当初是为本地机器上其他类型的有状态交互设计的。 I was trying to make it talk to a web system, and it was designed for stateful interactions of other kinds on local machines. 254:30 因此它需要各种繁琐的配置和来回折腾，我当时就想：“这玩意儿我不想手写。” And therefore, it required all sorts of cumbersome setup and dance and, and I was just like, "I don't wanna write this by hand." 254:38 嗯哼。 Mm-hmm. 254:38 当然，搁今天我绝不会手写了，但当时我想的是：“这得我手写。可我不想手写。那有没有别的办法？” Now, of course, today I would never write it by hand, but at the time I was thinking, "I have to write this by hand. I don't wanna write it by hand. So what's a different way? 254:45 “智能体能不能直接用我们已有的 Web 界面呢？”于是我让它试了试，让它去注册 Fizzi——那是我们当时刚发布的产品。 Could the agents just use the web interfaces we already have?" And I set it off on a trial to have it sign up for Fizzi, which was the product we had just released at the time. 254:56 嗯哼。 Mm-hmm. 254:56 然后，呃，它打开 Fizzi 网站，开始填表，然后它说：“嘿，我没有邮箱地址。注册这个服务需要邮箱地址。”我第一反应是：“哦，天，这下我得给它弄个邮箱了。” And, um, it goes to the Fizzi site and, and starts filling it out, and then it goes like, "Hey, I don't have an email address. Signing up for this service requires an email address." And the first thing I thought, like, "Oh, man, now I gotta set up an email address." 255:08 然后我想：“不对，你自己去注册一个邮箱。去 hey.com，注册一个邮箱地址。”我心想：“HEY 肯定通不过这种东西。”当然，现在看来，它能搞定简直是显而易见的。 And then I thought, "No, you set up an email address. Go to hey.com, sign up for an email address." And I'm like, "Hey is not gonna get this." And of course, now it seems obvious it was gonna get this. 255:18 但在当时，这对我是个巨大的启发：它只靠提示词的指引就能自己弄明白——第一步，我得去 fizzi.com；我得找到注册链接；然后开始填表。 But at the time, it was a great revelation to me that it could just on pure direction through prompts figure out, A, I gotta go to fizzi.com, I gotta find a sign-up link. I start filling out the form. 255:30 哦，我意识到我需要一个邮箱地址。哦，我得问问我的、我的提示者现在该怎么办。他让我去 hey.com。 Oh, I realize I need an email address. Oh, I gotta ask my, my prompter what to do now. He tells me to go to hey.com. 255:37 我去了那儿，把整个注册流程走完，然后再回去注册 Fizzi，完成了整个闭环。它做完这些之后，我心想：“我的天。”于是我就说：“好了，现在你有邮箱了。” I go there, I sign up for the whole thing, and then go back and sign up for, for Fizzi, and complete the entire cycle. And after it did that, I was like, "Jesus." So I was like, "Well, now you have an email address. 255:50 “去注册 Basecamp 吧。我会邀请你。记得查收邮件，”我对 KEF 这个机器人说。“记得查收邮件。”于是我从 Basecamp 给这个智能体发了一封邀请。 Go sign up for, for Basecamp. I'll invite you. Just check your email," I told KEF the bot. "Just check your email." So I sent an invite from Basecamp to the agent. 255:59 这个智能体在 HEY 里收到了邮件——不是通过 CLI，也不是通过 MCP，就是用的网页，对吧？点开邮件，点开邀请链接……就进 Basecamp 了。 The agent receives the email in Hey, not through a CLI, not through an MCP, just using the web, right? Clicks the email, clicks the invite link ... is in Basecamp. 256:10 然后它问：“我现在该做什么？”我就说：“不知道。去我们的 AI 频道，做个自我介绍吧。” Like, "What should I do now?" And I just go like, "I don't know. Go to our AI room and introduce yourself." 256:15 嗯哼。 Mm-hmm. 256:16 再次强调，纯粹通过 Web 界面。 Again, purely through the web interface. 256:18 嗯。 Mm. 256:18 它找到了 AI 频道。走进去说：“嘿，我是 Kev。我是，呃，David 的，呃，AI 机器人。我、我很高兴能来这里。” Go- finds the AI room. Goes in there, "Hey, I'm Kev. I'm, uh, David's, uh, AI bot. I'm, I'm so excited to be here." 256:26 嗯。 Yeah. 256:27 我当时就—— I was like- 256:29 嗯 Yeah 256:29 ……三观震碎。那个实验有意思的地方在于：第一，它是对未来的一瞥，因为它有点慢。我记得整个流程端到端大概花了 12 分钟，对吧？ ... mind blown. Now, what was interesting about that experiment was, A, it was that glimpse of the future because it was kinda slow. I think it took maybe 12 minutes to do the whole thing end to end, right? 256:41 呃，我现在真应该重新跑一遍那个测试。要是没快很多，我会很惊讶的。我当时用的是，呃，GroqBot 那套东西—— Um, now I should actually rerun that test. I'd be shocked if it's not much faster. I was just, uh, using, uh, the GroqBot thing- 256:47 嗯哼 Mm-hmm 256:47 ……它内置了各种用来操作浏览器、操作电脑的功能，而且有自己专用的电脑环境，不需要你自己去搭一台电脑。 ... which has all these, uh, stuff built in for using the, uh, using browsers and using a computer, and has its own dedicated computer without you having to set up a computer. 256:56 嗯。 Yeah. 256:56 基本上就像个装在盒子里的 Claw。 It's basically like a claw in a box. 256:58 嗯哼。 Mm-hmm. 256:59 我让它去注册个什么东西，它就有点……我瞥见了一种感觉：哦对，这就是我们现在所处的位置，用起来快多了。但在当时，我真的是被震撼到了。 And I was asking it to sign up for something, and it, it kind of... I got a glimpse of like, oh yeah, that's where we are now, and it's so much faster to use. But at the time I was just blown away. 257:09 然后我用了一阵子，发现：啊，还没到时候。我没法真的只用这种方式跟它协作。它还是离不开 CLI，还是离不开 MCP，因为实在太慢、太浪费 token，诸如此类。 And then I used it for a little bit and I found like, ah, it's not there yet. I can't actually communicate it within this way. It still needs the CLI. It still needs an MCP because it's just too slow and too token inefficient and, and so forth. 257:20 但它确实是对未来的一瞥。那之后我就基本没再用它了，因为我想用智能体做的事，大多都是我坐在电脑前就能做的。 But it was this glimpse of the future. And then after that I kinda just didn't use it because the things I wanna use agents for I was mostly in front of my computer to do. 257:30 不过最近情况有点变化，这也是我现在比较乐意用 Claude Code 的原因之一，因为 Claude 有最好的手机 App：你在电脑终端里启动的任何 Claude Code 会话，都能通过 App 访问—— Now, that has changed a little bit lately and is one of the reasons why I'm so kind of okay about using Claude Code, because Claude has the best mobile app where any session you start Claude Code in your terminal on a computer is accessible through the app- 257:49 对。远程。 Yep. Remote 257:49 ……你完全不需要做任何额外设置。 ... without you having to do anything at all. 257:51 对。 Yep. 257:52 当然你也可以用终端。比如我、我装了 Terminus，可以跑 Herder 之类的，但有点麻烦，而且你是在远程实时往电脑上敲字，能明显感觉到延迟。 And you can do a terminal. Like I, I installed Terminus and it can run Herder and so on, but it's a little cumbersome and you're doing live typing on a computer, so you feel the latency. 258:00 嗯哼。 Mm-hmm. 258:00 当你用 Claude App 连接你的 Claude Code 实例时，感觉就像在和它聊天一样自然，没有还要去注册 Telegram 账号之类的麻烦。 When you're using the Claude app to your Claude Code instances, it feels just like chatting with it without the cumbersomeness of having to set up a Telegram account and, and so forth. 258:09 当然，如果你想用一个开源的 Claude 来打理你的生活，那它做不到这一点。但不知为什么，我对那方面一直没那么大兴趣。 Now, if you wanna use an open Claude to like run your life, well, then this doesn't do that. But I haven't been so interested in that for some reason. 258:17 也许其中一个原因是，呃，我太太 Jamie 说了：“我们家不许进机器人。”所以我想，啊，那我最好别把它接到，呃，那个，家里的智能家居系统上。 Maybe one of the reasons is that, um, Jamie, my wife, is like, "We're not getting a robot inside the house." So I'm like, ah, I better not hook it up to the, uh, um, sort of, uh, smart house here. 258:28 要是它开始像闹鬼一样把灯开了又关，我大概在家里不会太受欢迎。我说这些的另一个原因是，我听过一些人把它接进了自己整个生活的故事。 If it starts, uh, turning lights on and off like a poltergeist I don't think I'm gonna be all that popular. So, and part of the reason I say this, I've heard some of the stories of people who have hooked it up to their whole life. 258:39 呃，我是说，这不是我的故事，所以我就简单分享一下这段轶事。有个人讲了他怎么把它同时接到自己的健康、健康数据—— And um, I mean, it's not my story to tell, so I'll just share sort of the brief anecdote. One guy was telling the story about how he hooked it up to both his health, health data- 258:48 嗯哼 Mm-hmm 258:48 ……和他的 Tesla 汽车上。然后他的，呃，他的智能体就执着于他喝水不够这件事，总是说：“嘿，你、你可得多喝点水。” ... and his Tesla car. And his, uh, his agent got obsessed with the fact that he wasn't drinking enough water and like, "Hey, you, you gotta drink more water." 258:58 于是有一天他正开着车，Tesla 突然开始朝一个新的方向开。 So at some point he's driving along and suddenly the Tesla starts driving in a new direction. 259:03 嗯。 Yeah. 259:04 原来智能体发现回家的路上有一家杂货店—— The agent had realized there was a grocery store on the way home- 259:08 哇 Wow 259:08 ……可以顺路拐过去—— ... that they could swing by- 259:09 太绝了 That's great 259:10 ……顺便买点水。 ... and pick up some water. 259:11 太绝了。 That's great. 259:12 你就会觉得，这种事吧，既好笑，又完全是《黑镜》的剧情，同时又会想：好吧，从这里到“家里闹鬼”其实也没几步远了。 And you're like, this is one of those things where it's like both funny and also totally Black Mirror and also totally like, well, okay, there's not that many steps from here where the poltergeist is in your house. 259:25 所以我就想：“啊，我不需要那个。”不过我现在确实挺喜欢这一点：我可以直接通过 App 控制 Claude Code 里的智能体，哪怕我用得不算多。 So I was just like, "Ah, I don't need that." I do actually now like the fact that I can control the agents through Claude Code just through the app if I, even if I don't use them that much. 259:34 我确实觉得，不管是 OpenClaw 还是 Hermes Agent，用一用它们、借此瞥见一种不同的交互模式，都是很有价值的。 I do think it's very useful to, for both OpenClaude and Hermes Agent to, um, to use it to get a, a glimpse of a different interaction mode. 259:44 对。 Yes. 259:45 就是想去感受一下——因为它、它给人的感觉就是未来，不——不一定是以这种形态，但某些方面确实在朝那个方向走。 And like just to g- 'cause it, it feels like the future, n- not necessarily in this way, but it's in some ways going there. 259:51 对。我觉得那些系统的另一个不同之处在于，它们有整套记忆系统，会不断积累关于你是谁、你喜欢什么的知识。 Yes. I think what, what's also different with those systems is they have this whole memory system where they're building up, uh, knowledge of who you are and- what you like. 260:01 嗯哼。 Mm-hmm. 260:01 呃，Tobias 讲过一些故事：他的智能体会替他购物——它有一笔零用钱，会帮他挑衣服，然后东西就直接出现在家门口，因为他的智能体—— Um, Tobias told the stories about how his agent will shop for him, that it has an allowance, and it'll pick out clothes for him, and it'll just show up at the door because his agent's- 260:11 嗯 Yeah 260:11 ……替他买了东西。你会想：“天，这真有未来感。但也有点吓人。”呃，不过我可以借那些站在前沿的人间接体验这种生活。 ... bought something for him. And, like, it's one of those things where you're like, "Man, that feels future. Also a little scary." Um, but I can live vicariously through others who are on that frontier. 260:22 就像你说的，你在考虑将来做一个 Omarchy 的移动版。 And like you said, you're thinking of the, in the future doing a mobile version of Omachi. 260:28 我的意思是，这就是那种妄自尊大的幻想，一旦你拥有了看似无限的力量，它就开始冒出来—— So I mean, this is the delusions of grandeur that start kicking in once you get the seemingly unlimited powers- 260:33 嗯 Yeah 260:33 ……那是我在创造 Quattro 的过程中亲眼见识到的。 ... I've witnessed with the creation of Quattro. 260:36 那那会是什么样子？这、这、这事儿会在哪儿撞墙？你是说——你是指 Android 吗？ So what would that look like? W- w- where does this crash? Are you m- do you mean, like, Android? 260:41 我——那是最显而易见的路径，对吧？Android 是开源的，你可以 fork 它，而且已经有人这么做了。有各种各样的，呃，Graphene OS 就是这样一个版本。我知道那种模式有一些局限。 I- that's the obvious path, right? So Android is open source and you can fork it, and people have. There- there's all sorts of, uh, Graphene OS is, uh, one such version. There are limitations to that model, I know that. 260:52 呃，比如触碰支付，还有相机模组之类的问题，等等，但那些看起来都是可以克服的。我特别希望我的手机能像我的电脑一样可塑，看起来能像我的电脑现在这么酷。 Uh, tap-to-pay and other things around camera models and, and so forth, but that all seems surmountable. It all seems that I would love if my phone was as malleable as my computer, if it looked as cool as my computer now does. 261:05 如果我能做出一个本质上的 Omarchy Mobile，那就太棒了。 If I could get essentially Omachi Mobile, that sounds amazing. 261:09 我很好奇那得是多大的工程量。 I wonder how much work that is. 261:11 我会去弄清楚的。 I'm gonna find out. 261:12 是啊。你知道，有些事就是这样：你一旦真的开始动手，比如做浏览器，你才会意识到它有多难。 Yeah. You know, it's like those things that you're, you realize once you start building, for example, that a browser is very difficult to do. 261:19 对。 Yes. 261:20 所以我就想，操——操作系统方面—— And so, like, I wonder w- with the operating system- 261:23 对，Ladybird 的开发者们正、正在体会到这一点，比如 Andreas 和—— Yeah, the developers of, uh, Ladybird is, is, is finding that out, like Andreas and- 261:27 嗯 Yeah 261:27 ……他的团队。那个项目有意思的地方在于，浏览器基本是世界上第二复杂的软件系统，第一复杂的是 Linux 内核。 ... his team. Now interesting with that project was the browser's basically the second most complicated software system in the world, the first one being the Linux kernel. 261:35 嗯哼。 Mm-hmm. 261:35 对吧？就这种，呃，动辄几千万行代码的大型系统而言。 Right? In terms of, um, these large tens of millions of lines of code systems. 261:41 所以在我们还没有智能体的时候，Andreas 带着一个小团队就开始这样的征程，真的非常大胆。 So very audacious actually to begin such a mission with a small team like Andreas did before we had agents. 261:48 但现在我们有了智能体，我们会见证的，而且很可能不会太远。我是说，如果说有一件事…… But now that we do have agents, we're gonna, we're gonna see, and we're probably gonna see sooner rather than later. I mean, if one... 261:56 如果说智能体已经在哪件事上强得离谱，那就是读规范文档并把规范实现出来。 If there's one thing agents are already exceptionally good at, it is to read specs and implement them. 262:03 我是说，我不知道现在 CSS 和 HTML 的规范有多长，再加上 JavaScript，但，呃，它们非常长。 And I mean, I don't know how long the specs are for CSS and HTML at this point, and JavaScript on top, but, um, they're very long. 262:11 嗯哼。 Mm-hmm. 262:11 人类要实现它们得花很多很多年。而我们很快就会发现，智能体做这些事到底能有多快。 And it would take humans many, many years to implement them. We're gonna find out just how quickly agents can do them. 262:18 你提到 Omarchy 招来了一些批评，那，呃，我们就沿着批评这条路再往前走一走。你公开表达过一些自己的政治观点。 You mentioned getting some criticism for Omachi, so, uh, let's walk further down that path of criticism. You have, uh, made some of your political opinions known. 262:33 呃，我想你大概可以把它们归、归到移民问题，或者非法移民问题，或者诸如此类的—— Um, I guess you could s- broadly put it in the category of immigration or illegal immigration or what- 262:39 我会说是大规模移民。 I'd say mass immigration. 262:40 大规模移民，人口结构在一种文化、一个社会、一个民族的形成中所扮演的角色——你因为这些观点招来了相当多的批评，也引发了不少风波。 Mass immigration, the role of demographics in the, in the formation, uh, of, of a culture, of a society, of a nation, and you've gotten quite a lot of criticism for that and create a lot of drama. 262:57 你后悔引发那些风波吗？ Do you regret any of that drama? 263:04 或者说，你总体上是怎么想的？ Like, what's your general thinking about? 263:06 一秒钟都没后悔过。 Not for a second. 263:08 好吧。 Okay. 263:08 我这么说的原因是：奥弗顿之窗不会自己打开。它是靠人们冒着一点风险、一次推动一点点，才慢慢打开的。 And the reason I say that is that the overtone window does not open itself. It opens one nudge at a time by people risking a little. 263:18 嗯哼。 Mm-hmm. 263:18 冒一点声誉受损的风险、一点反对、一点批评，或者有时候，是很大的声誉代价、很多的批评、很多的反对。 A little reputation, a little pushback, a little criticism, or maybe sometimes a lot of reputation or a lot of criticism or a lot of pushback. 263:26 而我认为，大规模移民问题在欧洲多年来一直是个彻底的禁忌，而且在好几个欧洲国家至今仍是如此。 And I think the question of mass immigration in Europe was for many years this total taboo, and it still is in several European countries. 263:36 不过，也许纯属机缘巧合，丹麦不是这样。 Now, for reasons of maybe just chance and happenstance, they weren't in Denmark. 263:44 丹麦人早在九十年代中期就开始讨论大规模移民及其后果了。 The Danes started having the discussion around mass immigration and the consequences thereof in the mid-'90s. 263:51 嗯哼。 Mm-hmm. 263:51 非常早。其中一部分可以追溯到一个人，Mogens Glistrup，他是一个非常古怪的人物。 Really early, and some of it can be, uh, traced back to one man, Mogens Glistrup, who was a very quirky character. 264:02 我记得他，八十年代在电视上看到他谈论大规模移民的危害，而我就是在那个年代长大的。 I remember him, seeing him on the TV in the '80s talking about the dangers of mass immigration, and this was at a time where I grew up in the '80s. 264:10 而在我长大的那个，呃，社区，在、呃，Brønshøj，那是哪一年来着？ And in the, uh, neighborhood I grew up in, in, uh, in Brønshøj, in, what was it? 264:17 84 年，99% 是丹麦族裔，都是能把血脉追溯到在那个国家生活了一千年的丹麦人的人。 '84, 99% ethnic Danes, people who could trace their lineage back to Danes who'd lived in that country for 1,000 years. 264:28 丹麦拥有世界上最古老的君主制之一，一直可以追溯到 Harald Bluetooth——蓝牙（Bluetooth）这个技术标准就是以他命名的。 The Danes have one of the oldest monarchies in the world, all the way back to Harald Bluetooth, who's literally who was named the standard Bluetooth. 264:37 所以有一千多年的历史，那里的人世世代代生活在同一个地方，而那个社区里 99% 都是这样的人。 So well over 1,000 years, like there's a long lineage there of people who lived in the same place, and 99% of the people lived in that neighborhood. 264:44 然后到九十年代初，我记得数字变了，变了大概 5% 还是怎么的，到现在已经降到大约百分之六十几、七十—— And then early '90s, I think it goes to, to 5% or something like that, and then at this point it's down to around 60-something or 70%- 264:53 嗯哼 Mm-hmm 264:53 ……在 Brønshøj。这是巨大的变化，而丹麦人开始注意到、开始讨论这一点：这并非一件全然的好事， ... in, in Brønshøj. That's a big change, and the Danes started noticing that or talking about that, that this was not an undivided blessing, 265:04 呃，大规模移民是有负面影响的——大约在九十年代，人们就开始这么说了。 uh, that there were downsides to mass immigration in about the '90s. 265:11 而这场辩论在整个两千年代愈演愈烈，这在邻国是没有发生的。尤其是瑞典，完全没有过这样的辩论。彻底的禁忌。 And the debate really roared through the 2000s in a way that didn't happen in the neighboring countries. So Sweden in particular just didn't have that debate at all. Total taboo. 265:21 挪威也一样，欧洲其他，呃，国家也一样。 Same thing with Norway, same thing with other, uh, countries in Europe. 265:26 然后到了 15 年之后的几年，情况突然变了——当时有一百万叙利亚人大迁徙，呃，沿着欧洲的公路一路北上，德国选择接收了其中很多人， And suddenly now, after some years after '15, where we had the big migration of a million Syrians, uh, walking up through the highways in, in Europe, and Germany choosing to take a lot of them and, 265:36 还有三万人抵达丹麦，等等，呃，这个问题就获得了更多关注，对吧？ and 30,000 arriving in Denmark and so forth, um, it's gotten more of a focus, right? 265:44 嗯哼。 Mm-hmm. 265:44 因为这种变化已经、已经发生了。 Because this change has, has happened. 265:46 而我发现，判断一种文化里哪里出了问题，有一个办法就是看：你被允许谈论什么？ And one of the ways I find that you can spot where sort of there's something that isn't right in the culture is, like, what are you allowed to, to talk about? 265:57 更准确地说，是你被允许注意到什么。我在一篇文章里写到过我的观察，那是哪年来着？ What are you allowed to notice actually, is how I should put it? And what I noticed in an essay, when was that? 266:03 呃，就在一年半前，那篇文章叫《我记忆中的伦敦》（As I Remember London）。我是从九十年代末之类的时候开始来伦敦的。那时候，反正是 59%、60% 是英国族裔。 Uh, only a year and a half ago, uh, h- as I remember London, was that I started coming to London in, uh, the late '90s or something like that. At the time, whatever, it's 59, 60% ethnic, uh, Brits. 266:16 然后 20 年后，降到了 34% 左右。这种变化是注意得到的。真的，你只要走在街上就会意识到：这座城市和我 20 年前来的时候已经不一样了。 And then 20 years later, we're down to 34% or something like that. Like, that's noticeable. Like, literally just walking on the street, you realize, well, this is a different city from what it was when I visited 20 years ago. 266:29 如果再往前推 20 年，就、就回到了和丹麦差不多的比例。 And if you went 20 years before that, y- you get back to similar rates as, as what the Danes have. 266:34 那时候这就是一个英国族裔占 85%、90% 多的国家，如果—— This was like a country of ethnic Brits in the 85s, 90s percentiles if- 266:42 嗯哼 Mm-hmm 266:42 ……再往前推的话。所以这是一代人的时间里就变成了另一个国家。注意到这一点没有错。心里想着“你知道吗？我不认同这种变化”，也没有错。 ... you get further back. So that's a different country in a generation. It's okay to notice that. It's also okay to think, "Do you know what? I don't agree with that." 266:52 嗯哼。 Mm-hmm. 266:53 “我希望它不是那样的。”我是说，如果，如果，呃——我上次去中国是 19 年，2019 年。我在、在上海。呃，那里全是中国人。 "I wish it wasn't that way." I mean, if, if um, the last time I was in China was '19, 2019. I was in, in Shanghai. Um, it was all Chinese. 267:07 如果我今年 10 月再去，突然发现只有 30% 是中国人，70% 是其他人，我会觉得有点奇怪。 If I, when I go back here this October, suddenly find that there's only 30% Chinese and then there's 70% other people, I'd be a little weird. 267:17 我大概会想：“呃，发生什么了？”呃，而如果那一刻中国人说：“呃，我们不喜欢这样。”我不会觉得被冒犯。 I'd probably be like, "Uh, what happened?" Um, and I, I would not take offense if the Chinese in that instant would go like, "Uh, we don't like that." 267:29 所以，那种认为欧洲各国没有权利——第一，注意到正在发生的事情；第二， So this idea that countries of Europe don't have a right to, A, notice what's happening and, B, 267:38 ——反对正在发生的事情的观念，我认为是，呃，错的，对吧？ oppose what's happening, I find it to be, uh, wrong, right? 267:47 就是说，我认为存在一个关于自决权的道德论证，而我们在世界其他地区一直在援引这个论证：哦，生活在那个地区的人民，他们有权自决，呃，决定自己国家的命运。 Like, there's just, I think there's a moral argument for self-determination, and we invoke that argument all the time in other regions around the world that like, oh, the people who live in that region, they have a right to self-determination, um, uh, for their country. 268:00 所以，比如说丹麦人、瑞典人或挪威人说：“呃，我们作为王国在这里已经一千年了，有着特定的族群构成。”这完全合情合理。这是…… And I think it's completely fair for, say, the Danes or the Swedes or the Norwegians to go like, "Well, we've been, um, kingdoms here for 1,000 years with a certain ethnic mix-up." It's a... 268:11 我是说，我本人是移民的坚定支持者。我自己就是、就是移居美国的移民。 I mean, I'm a big fan of immigration. I'm a, I'm an immigrant to the US. 268:16 呃，精挑细选式的移民——各国竞相吸引全世界最聪明、最有才华的人——这、这是大好事。 Um, cherry-picked immigration where countries compete to attract the smartest, most talented people around the world is, is really good. 268:27 合法的、择优的移民。 Legal merit-based immigration. 268:30 “择优”是关键，因为实际上欧洲大部分地区几乎没有非法移民。南欧有一些非法移民，但欧洲大部分地区并没有严重的非法移民问题。 Merit-based is key here because there's actually virtually no illegal immigration in most of Europe. Southern Europe has some illegal immigration, but most of Europe does not have a big problem with illegal immigration. 268:42 嗯哼。 Mm-hmm. 268:43 问题在于其他形式的大规模移民：最终你得到的人口结构，会和短短几十年前完全不同。 And the problem is with mass immigration of other kinds, and that you end up with demographics that are just totally different from what they were even a few short decades ago. 268:55 而我认为欧洲各国是会从移民中受益的。事实上，丹麦人有一点很有意思：他们对移民做着极为详尽的统计。 And I think the countries of Europe would benefit from immigration. In fact, one of the interesting things about the Danes is they carry meticulous statistics on immigration. 269:07 我不完全清楚为什么他们统计得这么细。也许一直可以追溯到 Mogens Glistrup 以及他在九十年代的那些讨论，但反正他们就是这么做的。 I'm not entirely sure why they're so detailed. Maybe it's all the way back to Mogens Glistrup and, and his, um, discussion in the '90s, but they do. 269:16 而这些统计清楚表明：当来自英国、法国、美国的人来到丹麦时，平均而言，这些移民对丹麦国家是非常有益的。 And what's clear from those statistics is that when the Danes have folks from the UK, France, the US show up, on average, those immigrants are really beneficial to the Danish state. 269:30 他们的贡献远大于他们向国家索取的。所以丹麦应该接纳这样的移民，呃，接纳那些贡献远大于索取的人。 They contribute far more than what they ask of the state. So the Danes should accept immigrants, um, who do that, who contribute far more than they, they take. 269:43 然后你再看看其他类型的移民群体。上周刚出了一份统计，呃，是一位丹麦政客要求——我想是财政部——用最新数据做的这份统计。 Then you look at other forms of immigrant groups. There was just a, a, a tally that came out, uh, last week from a Danish politicians that had asked the, I think, uh, finance ministry to, to do this tally with the latest numbers. 269:57 它对丹麦国家的成本是多少——就、就我提到的那种正向移民的情形而言。 What does it cost the Danish state to, to- Of positive immigration, this, this, uh, case as I mentioned. 270:03 我记得在那份榜单上，呃，不管是法国、英国还是美国，都集中在同一个水平：每年为丹麦国家带来约 2.5 万美元的净收益。 I think for the, uh, on that list, uh, whatever, France, UK, US, they were all clustered around the same thing, about, uh, $25,000 net benefit to the Danish state every year. 270:14 你会觉得，哇，这非常正向。再说一遍，这是平均值，对吧？而在光谱的另一端，成本最高的移民群体是索马里人。 It was like, wow, that's very positive. That's average again, right? And then at the other end of the spectrum on the most costly immigrants was Somalis. 270:26 他们平均每年让丹麦国家净支出 2.8 万美元。 They ended up costing the Danish state on average $28,000 a year. 270:33 我认为丹麦人说“我们希望多来一些法国人、英国人和美国人，少来一些索马里人”是合情合理的。 I think it's fair for the Danes to go like, "We'd prefer to get more French, Brits, and Americans, and not so many Somalis." 270:44 你觉得你那篇文章为什么会招来那么多恨意？ Why do you think you got so much hate on that post? 270:47 那在很多圈子里是个非常，呃，禁忌的话题。 That's a very, uh, taboo topic in a lot of circles. 270:52 其中一部分和奥弗顿之窗的扩张有关。 Some of it has to do with Overton window expansion. 270:56 肯定是，但我也觉得……我是说，我想在这里采用“钢人论证”的方式。 I'm sure, but I also think... I mean, I wanna take the steel man argument here. 271:01 嗯。 Yeah. 271:01 我、我确实认为这种事可能会变得难看，对吧？ I, I do think that it, this can turn ugly, right? 271:06 这、这些观点有没有可能直接滑向公然的种族主义——一种不基于任何其他东西、只基于对他人仇恨的种族主义？ I- is it possible for some of these things to just veer into overt racism that isn't founded in anything else but a hatred of other people? 271:17 对。 Right. 271:17 是的，呃，那种可能确实存在。但这种风险并不能否定讨论的必要性——讨论你的国家应该由什么人构成、你的移民政策应该是什么样子。 Yeah, uh, that could happen. But that risk does not negate the need to have a discussion about the makeup of your country and what your immigration policy should be like. 271:29 我还觉得特别好笑的是，尤其是在欧洲，这种讨论——关于欧洲人是否有权自决、决定自己国家的面貌—— And what I also find so funny is, especially in Europe, it's, it's so kind of myopic that this discussion about whether Europeans should have a right to self-determination about what their countries look like, 271:39 ……那套标准似乎只适用于欧洲。没有人跑去对日本人说：“嘿，你、你们太日本了。” that standard seems to only apply there. No one is trying to tell the Japanese, "Hey, you, you, you're too Japanese." 271:50 比如说，呃，我记得东京反正是 98% 都是，呃—— Like, uh, whatever, it's 98%, I think, in Tokyo that's, um- 271:53 嗯哼 Mm-hmm 271:53 ……日本族裔，对吧？比如，有、有没有人跑去告诉日本人：如果东京的日本族裔比例降到伦敦那种水平——降到随便什么 37%——东京会成为一个好得多的国家？我不觉得有。 ... uh, ethnic Japanese, right? Like, i- is anyone trying to tell the Japanese that their Tokyo would be a much better country if, if they were down to London levels of ethnic, uh, Japanese, that it was, whatever, 37%? I don't think so. 272:08 呃，所以在我看来，这里有一种奇怪的、近乎自我憎恨的东西——我是说，呃，Gad 把它叫做“自杀式共情”—— Uh, so there's this weird almost self-loathing in my opinion, I mean, uh, God calls it suicidal empathy- 272:17 嗯哼 Mm-hmm 272:17 ……它渗透在这一切之中，其程度我到现在还没完全想明白。比如，这到底是从哪来的？为什么会是这样？ ... that permeates this stuff to a degree that I still haven't fully unpacked. Like, where is this coming from? Why is it this way? 272:26 但不管怎样，我只是注意到了：嘿，伦敦看起来不一样了。 But regardless, I just noticed, hey, London looks different. 272:34 我更喜欢以前的伦敦，就像我会说我更喜欢，呃，我在哥本哈根长大的那个社区在八十年代的样子。 I preferred the old London, as I would say I preferred, uh, the neighborhood that I grew up in in Copenhagen, how it looked in the '80s. 272:46 就是说，我完、我完全明白你怎么能从中跳五步推理然后说：“好吧，这就是种族主义。”而我现在也已经不在乎了。这本就是一场正当的讨论。 Like, I s- I totally get how you can then jump five leaps and go like, "Okay. Well, that's racist." And I also, I don't care anymore. Like, this is a fair discussion to have. 273:00 国家和民族有权制定自己的移民政策。而那种认为我们不该有任何边界的观念，或者甚至更糟的——认为一切只是一张白纸， Countries and peoples have a right to set an immigration policy, and this notion that we shouldn't have any borders, or, or even worse than that, it's all just a blank slate, 273:10 认为所有民族都一模一样，我们可以把人从地球的一个地方拿起来，放到地球的另一个地方，然后一切都会顺顺利利、 that all peoples are just the same, and we can take people from one place of the earth, and we can place them in another place of, on, on, on the earth, and everything's just gonna work out h- hunky dory. 273:26 从经验上看，这不是真的。欧洲从八十年代就开始做这个实验了，结果并不好。 Empirically not true. Europe has been running that experiment since the '80s. It hasn't panned out well. 273:34 而现在欧洲有一些力量想要逆转这个实验，想要终结迄今为止这种大规模移民的运作方式。 And now there are forces in Europe who wanna undo that experiment, who wants to end mass immigration as it's been run so far. 273:45 而我恰好相信——大多数丹麦人和其他不少民族也相信——“你知道吗？是的，我们之前的做法行不通。我们得做出一些改变。”再说一次，我不认为这意味着零移民。 And I happen to believe, and so do a majority of Danes and a bunch of other peoples that like, "Do you know what? Yeah, it hasn't been working how we've been doing it. We gotta make some changes." And I don't think that means, again, zero immigration. 273:58 再说一次，我自己就是移民。我认为总体而言美国从移民中获益巨大，但这里有一种偷换概念的手法。 Again, I'm an immigrant. I think US, generally speaking, have benefited tremendously from immigration, but that's the kind of sleight of hand. 274:07 移民好不好，很大程度上取决于来的是谁。当移民是社会的净贡献者时，国家才会受益，而且在大多数情况下， Immigration depends greatly on who immigrates, and a country benefits when the people who immigrates are net contributors to the society, and in most cases, 274:17 还要他们融、融入当地文化。 that they as- they assimilate with the local culture. 274:23 嗯。 Yeah. 274:23 如果你来到丹麦，你想成为一个丹麦人，你付出全部努力，你为社、社会做贡献，我想大多数丹麦人会很乐意在合理的数量内接纳这样的人。 If you come to Denmark, and you wanna become a Dane, and you put in all the effort, and you contribute to s- society, I think most Danes would be quite happy to, uh, allow that in, in a reasonable number. 274:36 如果你来到丹麦，你不想融入，你是社会的净负担，你活在一个平行社会里，那么丹麦人说：“呃，我们别再这样下去了。”同样合情合理。 If you come to Denmark, and you don't wanna assimilate, and you are a net drain on society, and you live in a parallel society, it's also fair for Danes to say like, "Uh, let's stop doing that." 274:47 有意思的是，呃，我对美国的移民制度不算精通，但据我所知，择优移民制度在这里一直没、没能通过。 It's interesting that, uh, I'm not well-versed on the immigration system in the United States, but as far as I understand, merit-based system have not, uh, passed here. 274:58 另外，美国文化也非常不同。这一点是我在两个地方都长期生活过之后才真正体会到的。 And then also, American culture is very different. This is a point I did not fully appreciate until living both places extensively. 275:07 美国有一种“可选择融入”的文化：一个从丹麦、乌克兰或任何地方来的人，要容易得多，也被宽容得多—— America has a culture of optional assimilation, where it is much easier and then much more condoned for someone coming from Denmark or Ukraine or wherever, 275:17 来到美国，认同美国的那套价值观、规范和文化，然后就能被其他美国人当作美国人来尊重。 come to America, buy into the American set of values and norms and culture, and then be able to be respected as Americans by other Americans. 275:31 所以每当有人说“美国是个多么种族主义的国家”时，我都觉得特别好笑。 This is what's so funny to me when there's discussion, like, "America's such a racist country." 275:37 嗯。 Yeah. 275:37 我就想：“你去过别的地方吗？你知道其他国家是什么样的吗？”如果你认为美国是最种族主义的国家，那你纯粹是被误导了。 I'm like, "Have you ever been anywhere? Do you know what other countries are like?" Like, if you think America is the most racist country, you are simply misinformed. 275:50 嗯。 Yeah. 275:51 我是说，我、我们在丹麦生活的那段时间就深有体会。呃，我妻子是有斯堪的纳维亚血统的美国人—— I mean, I, we're dealing with that, the, with time we've spent in Denmark. Uh, my wife is American of Scandinavian heritage- 276:02 嗯哼 Mm-hmm 276:02 ……但已经在美国生活了好几代。非常难。非常难。 ... but multiple generations in the US. Very difficult Very difficult. 276:07 作为一个金发碧眼的女人，付出各种努力学语言、做所有的事情，依然非常非常难以融入、融入丹麦文化。 As a blonde, blue-eyed woman who's putting in all sorts of efforts of learning the language and doing everything, still very, very difficult to assimilate to, to Danish culture. 276:19 而如果反过来——如果她是丹麦人、搬去美国——根本不会那么困难。这是另外一点。 In a way where if, if that had been the other way around and, and she had been Danish and moving to the US, it would not have been that difficult at all. And this was the other thing. 276:28 我是说，住在丹麦这样的国家，妻子是有斯堪的纳维亚血统的美国人，付出了全部努力，然后你依然意识到：天，融入竟然这么难，哪怕你已经一路做到了 97%。 I mean, living in a country like Denmark with a, an American wife who is of Scandinavian heritage and has put in all the effort, and still realizing, man, assimilation is incredibly difficult even when you're 97% there all the way. 276:45 天哪，那如果你来自一个完全异质的文化，岂不是几乎不可能—— Holy smokes, must it be almost impossible if you come from a completely foreign culture- 276:51 嗯哼 Mm-hmm 276:51 ……带着完全不同的规范和所有的一切。 ... and foreign norms and the rest of it. 276:55 当然，我刚才引用的统计数据也印证了这一点：文化差异越大、距离越远，呃，融入成功的可能性就越低。 And of course, that's what the stats I just quoted bear out, that the more dissimilar the culture is and the further it is, uh, the less likely to be successful. 277:07 呃，公开表达这种立场，有没有让你失去过友谊或其他关系？ Has, uh, articulating that stance cost you friendships, relationships? 277:14 有，但我会说，不只是因为这一件事。 Yes, but not just that alone, I would say. 277:19 更多是 2020 年前后那道大裂痕造成的。我觉得很多政治断层线在那时真的裂开了，我是说， It's more of the great divide that happened around 2020, where I think a lot of these political fault lines really opened up, and I mean, 277:28 在美国，那个简明的分界线就是：你、你是不是“觉醒”（woke）？ the shorthand in the US, I think, is w- were you woke or not? 277:35 显然，我曾有一群人，我、我以为我们多年来关系很好，结果那条政治断层线一裂开， And clearly I had a bunch of people that I, I thought I had good relations with for m- many years who suddenly, once that political fault line opened up, 277:46 关系就突然不再了。我觉得这很遗憾。 they were clearly just not anymore. And I think that's regrettable. 277:54 而我一直在努力，呃，去理解这种变化从何而来，部分原因是这些年来我自己在其中一些话题上也改变过想法， And I think I've, I've, I've tried to put an effort into, um, sort of both understanding where that comes from, in part because I've changed my mind on several of these topics over the years, 278:04 就像，嗯，有时候你只是随着时间改变了想法，有时候你是在获得新信息之后改变了想法。 and it was like, well, sometimes you just change your mind after time, sometimes you change your mind after you get new information. 278:13 改变想法有各种各样的原因和时机，而我们并不会都在同一个时间点改变，甚至有些人根本不会改变。 There's all sorts of reasons and moments why you may change your mind, and we don't all change our mind on the same schedule or at all. 278:19 嗯哼。 Mm-hmm. 278:20 呃，所以让自己身边都是——或者至少经常接触——那些不是在每个话题上都和你想法一样的人，我认为是相当健康的。 Um, so surrounding yourselves and, and, or at least interacting with people who don't think like you on every topic I think is quite healthy. 278:31 我是说，我偶尔也会和我弟弟讨论政治。 I mean, I occasionally discuss politics with my brother. 278:35 我是说，我们在这些话题上并不完全看法一致，有时候讨论会有点激烈，但之后也会平静下来，然后我们意识到：好吧，行，这件事我们意见不合，但我们在很多其他事情上是一致的。 I mean, we don't quite see eye to eye on all these topics, and sometimes the discussions can get a little heated, but then they'll also calm down, and we realize, all right, okay, so we don't agree on this, and then we agree on a lot of other things. 278:47 我们欣赏很多相同的东西。说起来，这正是我们在 Basecamp 那场大风波时的事情之一。 We appreciate a lot of the same things. Now, this was actually one of the things when we had our big blowup at Basecamp. 278:53 从那以后，我以全新的方式理解了过去那条规范：跟陌生人、甚至熟人谈政治、宗教或金钱， I came to appreciate in a whole new way this, uh, earlier norm that talking politics or religion or money with strangers or even acquaintances, 279:02 当然也包括同事，本来就不是什么好主意——如果你不总是拿政治分歧去摩擦—— and certainly with colleagues, was just not a good idea, that you had a much higher likelihood of being able to carry a good working relationship with someone if you're not rubbing your political differences up against- 279:21 嗯 Mm 279:21 ……对方，你和一个人维持良好工作关系的可能性会大得多。所以我其实认为自由派和保守派应该一起共事。 ... each other all the time. So I actually think liberals and conservatives should work together. 279:27 我认为如果部落之间彻底隔绝，没有任何交集或互动，这个国家会糟糕得多。而只要你不时时刻刻盯着彼此的分歧，这种交融其实容易得多。 I think the country would be much worse off if there's a complete, uh, separation between the tribes, and then there can't be any overlap or interaction, and I think that mingling is so much easier to do if you just don't focus on your differences all the time. 279:45 多关注你们共同喜欢的东西。我是说，这就是我在科技圈不幸，呃，渐行渐远的一些人。 Focus on what you like together. I mean, this is some of the people I've had unfortunately, uh, drifted from in tech. 279:54 我们曾有那么多的共同点，对，比如说 Ruby 或其他东西有那么多共同的热爱——你说，这难道不遗憾吗？ We had so much in common, so much shared love for, say, Ruby or other things- Things and like, isn't that unfortunate? 280:02 我希望生活在这样一个世界：分歧本身也是非常有趣、值得去谈的。 I would like to live in a world where the differences are the, are very interesting to also talk about. 280:08 对。 Yes. 280:09 而且不要过度情绪化—— And not to be overly emotional- 280:13 没错 Correct 280:13 ……哪怕面对再激进的分歧。 ... about e- even radical differences. 280:17 对。 Yes. 280:17 因为首先，分歧之处正是有趣的东西所在。 Because in the differences, first of all, that's where interesting stuff is. 280:22 对。 Yes. 280:22 而、而且，这也是，呃，人成长的方式。 And, and, but also it's, uh, it's how you grow. 280:27 如果分歧能被接纳，人们能聚在一个社群里谈论分歧，这、这是一个社会健康的标志。 It's, it's a sign of health in a society if the, the differences are embraced and they are brought together into community where you can talk about it. 280:38 对我来说，一个很好的例证是我看过的一段，呃，William F. Buckley 的节目片段，我想那节目叫《Firing Line》，还是—— One of the great illustrations of that for me was when I watched a clip from, uh, William F. Buckley's show, I think it was called Firing Line or- 280:46 对，对 Yeah, yeah 280:46 ……之类的名字。他请了一位黑豹党成员—— ... something like that. And he had a member of the Black Panthers- 280:49 嗯哼 Mm-hmm 280:49 ……上他的节目。很显然，这两位绅士在政治立场上没什么共同点。 ... on his panel. It was quite obvious that those two gentlemen did not share a lot of politics. 280:59 他们能持续对话——我看的那段——25 分钟。Buckley 只是在问：“那你怎么看这个？” They were able to carry a conversation for I think, I watched it, 25 minutes. And Buckley's just asking, like, "So what do you think about this? 281:07 “你怎么看这个？”而那位黑豹党成员的回答，正是你想象中七十年代黑豹党成员会有的回答。就是说，没有太多，呃，和解的口吻。 What do you think about this?" And like, the Black Panther would reply as you would expect the Black Panther to reply in the '70s. Like, there was not a lot of, uh, sort of conciliatory tones. 281:19 我当时就想：哇，这太难得了。如果今天再看到这样的节目，一定会变成对骂，一定只是互相扔指控。 And I just thought like, wow, that is so rare. If you had seen that show today, it would've been a shouting match. It would've been just accusations being hurled back and forth. 281:31 Buckley 能和他强烈反对的人进行这样的对话，依然去探究他们立场的所有思想层面，并让他们把这些层面讲出来， The fact that Buckley was able to just have these conversations with people he vehemently disagreed with and still explore all the intellectual facets of their standpoints and letting them speak to those facets, 281:45 这让人想起一个失落的时代，而我非常希望回到那样的时代。 talks of a lost era that I'd very much like to get back to. 281:52 我是说，我其实一直很喜欢和朋友们聊政治——那些我信任的朋友，我们能以一种不必上纲上线、生死攸关的方式谈论这些话题。 I mean, I always actually loved talking politics with friends that I trusted could talk about these topics in a way where it didn't have to be existential. 282:03 但、但当下这个时代有些东西让这种交流变得，呃，越来越难得了。 But, but there's something about the current era where that's, uh, harder and harder to come by. 282:08 所以你刚才说的那些，值得经常讲出来，提醒人们记住一个我们都应该追求的理想：可以有分歧，并且、并且能坐下来谈，谈论这些分歧。 And so what you just articulated is important to say regularly to remind people of an ideal we should all strive for, to be able to have differences and, and, and talk, and talk about them. 282:20 因为政治本身是有趣的。我是说，我到现在已经习惯说我讨厌政治了，但我真正的意思是讨厌当下这套体系：在这个体系里，某些话题， 'Cause like politics is fun. Like I, I've gotten used to saying I hate politics at this point, but what I really mean is whatever the system that's currently happening, where there's certain topics, 282:31 我——我们都知道有哪些话术，呃，一旦说出口就会把你归进蓝营或红营的筐里，于是制造的是抓马而不是对话。 I just kn- we know all the phrases you can say, uh, that somehow put you in a bin of blue or red, and that creates drama versus a conversation. 282:44 比如我、我能想到很多无害的——如果我不关注互联网的话——我能想到一堆我分析世界局势时可能说出的完全无害的话，但它们会触发所有人说， Like I, I could, I could think of a lot of harmless, if I wasn't paying attention to the internet, I could think of a bunch of harmless statements I could make analyzing the situation in the world that would trigger everybody to say, 282:57 “哦，那家伙是个左派——” "Oh, that guy's a leftist-" 283:02 对 Yes 283:02 ……垃圾”，或者是个——你懂的——右翼法西斯。而客观地说，如果我没在关注互联网，我会被震——而我确实被震惊过。 ... piece of shit," or a sh- you know, a right-wing fascist. And then objectively, if I wasn't paying attention to the internet, I would be sh- and I have been shocked. 283:14 因为我没有深度关注互联网，呃，但我读了很多、准备了很多内容，比如，呃，关于乌克兰战争的。 'Cause I had haven't been paying attention to the internet deeply, uh, but reading and preparing a lot, for example, for, uh, for, for the war in Ukraine. 283:25 而我一直很震惊：我的某些话竟然会被解读成立场的这一边或那一边，然后因此遭到恶毒的攻击。 And I have been shocked how certain statements from me can come off as one way or the other, and I get viciously attacked for it. 283:35 但那种攻击实际上是在说：“嘿，你这个自以为可以自由行走、自由思考的混蛋，给我选一个筐——” But what that attack is doing is it's saying, "Hey, you, asshole who thinks you could just move about and think freely, pick a bin- 283:45 对 Yes 283:45 ……然后乖乖坐在那个该死的筐里。” ... and sit in that fucking bin." 283:46 对。 Yes. 283:47 穿上那件蓝衫或红衫。然后、然后我实际上发现，事情就平静下来了。 Put on that blue shirt or red shirt. And then, and then actually what I've realized is things calm down. 283:54 它不会——这套系统不会找你麻烦，只要你直接说：“我是蓝营的人”或者“我是红营的人”。 They don't, the system doesn't pick on you if you just say, "I'm a blue person," or, "I'm a red person." 284:01 没错。 Right. 284:02 但如果你只是好奇地探索世界、对世界进行推理，这套系统就会惩罚你。而我、我、我讨厌这一点，因为我想回到那种 Buckley 和黑豹党成员对话的时代。 But if you're just like curiously exploring the world and reasoning about the world, the system punishes you. And I, I, I hate that, 'cause I wanna get back to the, the Buckley and the Black Panther discussion. 284:14 对。 Yes. 284:14 你刚才只是在表达观点。但务实地看现在的处境，我已经意识到：除非我真的非常关心某个议题，否则不要谈政治。因为我意识到这是有代价的。 And you were just saying opinions. But then looking pragmatically at the situation, I have realized not to mention politics unless I really care about an issue. 'Cause I realize there's a cost. 284:27 他们、他们会、他们会把你扔进蓝筐或红筐。 They, they'll, they'll put you in the blue bin or red bin. 284:31 对。 Yes. 284:31 所以你必须更审慎地选择战场。 So you have to be more deliberate about choosing. 284:33 对。呃，我也在努力对此更加自觉。 Yes. Uh, and I've tried to be more conscious of that, too. 284:39 而当我确实选择开口谈、谈论它的时候，通常是因为我已经深思熟虑过了。 And then when I do choose to speak ab- about it, it's usually because I have deliberated. 284:44 是啊。 Yeah. 284:44 而且我觉得，你知道吗？这是值得的。 And I think, you know what? This is worth it. 284:46 嗯。 Yeah. 284:46 不过，我这么说，然后我也经常发现自己想的正好相反，比如我因为某个人的技术工作而关注他， Now, I say that, and then I also catch myself thinking the opposite all the time, that I follow some person or another for whatever their technical work, 284:56 然后我了解到他的政治立场，我常常想：“我真希望我不知道这个。”所以我的意思是，这挺虚伪的—— and then I learn their political standpoint, and I often do think, "I wish I didn't know that." So I mean, it's hypocritical- 285:09 嗯哼。 Mm-hmm 285:09 ……这样去想别人，然后自己又说：“好吧，我偶尔也会—— ... then to think that of other people and then go like, "Well, occasionally I will- 285:13 对。 Yeah 285:13 ……在有争议的辩论里插一脚。”然后你就得接受这是有代价的。我确实认为你应该，而且更多人应该，对此有所意识。 ... chime in on a debate that's controversial." And then you have to accept that there is a price to that. And I do think you should, and more people should, be somewhat conscious about that. 285:25 再说一次，我们这儿都在自相矛盾，但就是说，如果我们能在一个更自由、更开放的场合谈论政治，纯粹从智识上把想法翻来覆去地探讨，那该多好？是的，那当然好。 And again, we're both contradicting ourselves here, but saying like, wouldn't it be nice if we could talk about politics in more free and open forum and just intellectually turn ideas around? Yes, it would. 285:36 但在目前社交媒体这种氛围下，这可能吗？相当难。 Is that currently possible in the climate of social media and so forth? It's quite difficult. 285:42 不过，这也可能只是一个阶段性的时刻，我们或许能挺过这个时刻。 Now, some of it could simply also be a moment, and we could get past that moment. 285:49 嗯。希望如此。 Yeah. I hope so. 285:49 也许有办法回到巴克利和黑豹党那种对话，哪怕是在社交媒体上，让那种讨论不至于演变成我们经历过的那种恶毒的群氓取消文化闹剧， Maybe there is a way to get back to Buckley's conversation with the Black Panthers, even on social media, in a way where those discussions don't lead to the sort of just vicious mob cancellation nonsense drives that we've had, 286:01 对吧？呃，我其实觉得我们现在比五年前更接近那个状态了。 right? Uh, I actually think we're closer there now than we were five years ago. 286:09 2020年那种部落主义，以及那些部落能用来惩罚异端的手段，比2025年要厉害得多。至少在科技圈是这样。 Like, the tribalism and the sanctions those tribes were able to exact on heretics in 2020 was way greater than it was in 2025. Certainly within tech. 286:22 当然，我知道如果你在学术界，比如说，你去哈佛的社会学系，提一些，呃，右翼观点，我觉得反应不会太好，而且你短时间内也缓不过来。 Now, I know if you're in academia, I mean, try going into the sociology department at, uh, I don't know, Harvard and mention some, uh, right-wing ideas. I don't think it's gonna go over ve- very well, and I don't think you're gonna recover from it quickly. 286:34 但在科技行业，以及至少很多其他商业领域，现在拥有一个与2020年共识不一致的观点是可能的。这是肯定的。 But in, in tech and in many other business domains at least, it is now possible to have a opinion that does not square with what the consensus was in 2020. That's for sure. 286:47 嗯，Rails 社区里也闹过一些风波。别的地方也有风波。这些……后来都怎么样了？都没事了吗？ Well, there's some, uh, drama in the Rails community. There's some drama elsewhere. Is, is, is it all... How did that turn out? Is that all okay? 286:56 从每一轮风波都在缩小这个意义上说，是的。 In the sense that every single round of drama shrinks. 287:01 就是说，我们经历了一轮巨型风波的高峰感染期，差不多就在其他所有社区在2020年前后达到风波高峰的同一时间。 So it's like we've, we had this peak infection of mega drama around the same time every other community had a peak infection of drama around 2020. 287:11 然后你就能看到，每次有新的爆发，规模都更小。 And then you just see, like every time there's a new flare-up, it's less. 287:16 嗯。 Mm. 287:16 而那些在意识形态上执着于那种斗争的人，越来越少，越来越少。 And the kind of people who are ideologically wedded to sort of that struggle just shrinks, and shrinks, and shrinks. 287:23 而到这个时候，在我看来，X 上发生过的最好的事情之一，就是 Bluesky 和 Mastodon 的出现。 And at this point, one of the best things that happened for X, in my opinion, was that Blue Sky and Mastodon came around. 287:31 因为它们有点像蜜罐，把 X 上、当时 Twitter 上最恶毒的那些人都吸过去了，对吧？ Because it kind of was like this honeypot for just the most, uh, vicious individuals on X, and Twitter at the time, right? 287:43 他们聚集在那里，形成一个不断缩小的回音室，经历一轮又一轮的纯洁性清洗，每次都缩小一圈，到最后，基本上没剩多少人了。 Who just congregated here in an ever-shrinking echo chamber that just went through purity cycle after purity cycle and shrank every single time to the point that, uh, there were just not a lot left. 287:57 现在留在那里的人，眼不见为净。这并不是说 X 是什么完美的地方，也不是说那里偶尔没有行刑队式围攻，或者没有……但确实少多了。 And the people who are left now there, they're just out of sight, out of mind. And that's not to say that X is some sort of perfect place, or there aren't firing squads occasionally there, or there aren't... But there's less of it. 288:09 每次我被卷进 Mastodon 上的……也不是被卷进去，就是从远处围观，Bluesky 上或者哪里的帖子，我都会想：“哦，靠，对，以前的 Twitter 就是这样。天哪。” Like, every single time I've been pulled into a Mastodon... Or not been pulled in, just spectated from afar, thread a Blue Sky or whatever, it was like, "Oh, shit, yeah, this is what Twitter used to be like. Jesus." 288:22 而现在你真的可以待在 X 上，就像我一样，我90%的互动都是关于很酷的 Linux 的东西、很酷的技术的东西。“看，我发现了一台新电脑。 And now you can actually be on X, as I am, and like 90% of my engagement is about cool Linux stuff, cool technology stuff. "Look, I discovered a new computer. 288:34 你能看出它有多快吗？”然后我可以跟其他同样为电脑兴奋的人聊这些。剩下大概10%就是：“好吧，我对某个有点争议的话题说两句，我们可以在那儿小讨论一下。” Can you see how fast it is?" And I can have those conversations with other people who are just also excited about computers. And then there's like 10% of it left that's like, "All right. I'll chime in on something that's a little controversial and we can have a little discussion there." 288:45 然后我们就可以回去继续为电脑而兴奋了。 And then we can just go back to being excited about computers. 288:47 嗯。 Yeah. 288:47 算法其实很擅长找到你会觉得有意思的东西，或者说——我不该这么说，能吸引你参与的东西。有时候它让人愤怒，但它确实能抓人。所以，从揭示偏好来看，大多数人想要的是“为你推荐”页面。 The algorithm is actually very good at finding stuff that you will find interesting, or s- I shouldn't say that, engaging. Sometimes it's enraging, but it is engaging, and therefore, revealed preference is that most people want a For You Page. 289:00 他们不只想要一个关注流。 They don't want just a following feed. 289:01 不对，但问题是，呃，他们是对“为你推荐”页面上瘾了，对吧？ No, but this, uh, they're addicted to the For You Page, right? 289:05 这……对，对，这倒是真的。 It... Yes. Yes, that's true. 289:05 我，我觉得如果有一个强大的 LLM 来做个性化的信息流，效果会好得多。问题是，如何大规模地实现这一点极其困难。 I, I think an LLM, like if we had a strong LLM doing the feed that's personalized, it would do much better. The p- the problem is how to deliver that at scale is extremely difficult. 289:16 嗯，问题是，奖励函数是参与度。一旦奖励函数是参与度—— Well, the, the problem is the, the reward function is engagement. And as soon as the reward function is engagement- 289:24 对。 Yeah 289:24 ……参与度就会顺着人的本能走，结果你懂的。但我也不想太悲观。 ... engagement's gonna follow base instincts, and here we go. But I also don't wanna be so negative. 289:30 对我来说，我明白我们每个人都有自己个性化的算法，有时候确实有点过头，但我会说，过去十年平均下来，现在的 X 是有史以来最好的。 To me, and I get we all have our own personalized algorithms, and sometimes it gets a little much, but I would say on average over the last 10 years, this is the best X has ever been. 289:41 我这么说，是因为我是一个很关心技术话题、很关心 Linux 话题之类的人。 And I say that as someone who cares about a lot of technology talk, a lot of Linux talk, and a lot of that. 289:47 现在这些讨论能够以某种方式进行，而不会老是被想把它拖进政治讨论的人砸场子—— That that's able to happen now in a way where it doesn't constantly get gatecrashed by people who wanna drag it into some political discussion- 289:56 嗯哼。 Mm-hmm 289:56 ……我觉得这真的很棒。我确实觉得 X 是找到这些东西的好地方。 ... I find to be really awesome I find actually X to be a great place to find that. 290:02 当然，它也是一个能找到其他东西的地方，而且我们都有自己的算法，所以你不能武断地宣称 X 是什么或者不是什么。 Now, it's also a place to find other things, and we all have our own algorithms, so you can't make any declarative announcements about what X is or what it isn't. 290:11 你的信息流可能跟我的完全不同，也跟别人的完全不同，对吧？ Your feed is probably totally different from my feed, and it's totally different from someone else's feed, right? 290:16 是啊，我，我没法，我没法评论这个，因为，因为我采访的人形形色色，我的信息流大概就是—— Yeah, I, I can't, I, I can't comment on this because b- because of the nature of the fact that I interview a wide variety of people, m- my feed is probably just a- 290:28 这种多样性对你的信息流是好事还是坏事？ Is, is this good for your feed or is it bad for your feed that you had that variety? 290:31 是坏事。 It's bad. 290:31 好吧。 Okay. 290:31 是坏事，因为我可能就是说：“我在吃这个苹果。真好吃。” It's bad 'cause I could be like, "I'm eating this apple. It's so delicious." 290:36 哦。 Oh. 290:36 然后一帮人就冒出来了：“哦，你这么说是因为你是泽连斯基的走狗—— And then a bunch of people will show up, "Oh, it's because you're a Zelensky shill- 290:40 对，没错。 Yeah, right 290:40 ……或者你是普京的走狗，你这个垃圾。”就好像，行吧。 ... or you're a Putin shill, you piece of shit." It's like, all right. 290:44 我还是回去—— I'll just go back to- 290:45 我是说，我，我偶尔也遇到过一点点这种情况，你就是会想：“等等，什么？我们在说什么？你……你为什么要以这种方式出现？” I mean, I, I ha- I occasionally have a little bit of that w- where you just go like, "Wait, what? What are we talking about? What are you... Why would you show up in this way?" 290:54 嗯。 Yeah. 290:54 这其实就是我很喜欢播客的原因之一，因为我发现，即使是长篇文字，当我写长篇论证时，人们会用他们脑子里为我设定的任何声音去读它， This is one of the reasons why I actually really enjoy podcasts, because I have found that even in long form writing, when I make long form arguments, people hear it in whatever voice they have in their head for me, 291:03 而这个声音往往在看到论证之前就被预设好了，比如：“哦，这家伙是个，管他呢，法西斯。” and that often gets pre-programmed in advance of the arguments, just like, "Well, this guy is a, whatever, fascist." 291:13 呃，然后他们就带着这种认知去听、去读那些字。但当他们看到或听到我阐述论点时的真实语气，他们可能不同意那些论点。我是说，显然很多人确实不同意。 Um, and then they hear the words and they read the words in those know. When they see or hear the actual tone of how I'm presenting the arguments, they may not agree with those arguments. I mean, many of them, of course, obviously don't. 291:24 但要直接无视或仇恨一个这样的人，就难多了。 But it's much harder to just dismiss or hate someone like that. 291:29 嗯哼。 Mm-hmm. 291:29 至少这是我的体会，也是我得到的反馈，人们会说：“嗯，我原来对你有某种看法，然后我在 Lex 的节目上听到了你。 At least that's what I've found, and that's the response that I've gotten, that people are like, "Well, I had a certain opinion, and then I heard you on Lex. 291:37 听起来其实讲理多了。” That actually sounded much more reasonable." 291:39 但比这更重要的是，我，我很希望大家对别人都能秉持这样一种态度：先假设对方是个好人。 But even m- more than that, I, I would love it if people just exercise this kind of approach to other people of, like, assuming they're a good person. 291:51 对。 Yes. 291:53 然后，如，如果他们有一个你不同意的观点，他们就是一个持有某个观点的好人。你也许可以觉得他们蠢，行吧。就当他们是迷路了。但要把他们当好人看。 And then, and i- if they have an opinion you disagree with, they're a good person who has an opinion. You could maybe think they're stupid, fine. Just think they're lost. But think of them as a good person. 292:04 我觉得如果你开始……有，有一件能带来参与度的事，就是踩别人。 I think if you think of, you start... Th- there's o- one of the things that gets engagement is kinda shitting on other people. 292:11 对。 Right. 292:12 而如果你就是带着挑毛病的心态去对待别人，你总能在他们身上找到毛病。他们身上总有让你不爽的地方。 And if you just approach other people as they're, you're looking for bad, you will always find bad in them. There's something annoying about them. 292:20 百分之百。 100%. 292:21 而，而且如果你用那种方式看人，你从中学不到任何东西，还会让自己的心充满仇恨。你不会……那只是一种糟糕的活法，一种糟糕的与世界相处的方式，而这恰恰是互联网在某种程度上鼓励的东西。 A- and if you look at people in that way, you're not gonna learn from it, and you're gonna fill your heart with hate. You're not gonna... It's just a bad way to live and to inter- interact with the world, and that is one of the things that the internet kind of encourages. 292:33 所以，如果你能用这样的心态对待别人，哪怕是你非常不认同的人，比如：“哦，我说不定能从这个人身上学到点什么。”黑豹党的人，威廉·巴克利，我们可能讨厌，呃，但还是能向他们学习。 So it'd be nice if you just approach e- even people who you really disagree with, it's like, "Oh, I might learn something from this person." The Black Panther person, William Buckley, we might hate, um, just learn from them. 292:45 说个这事儿的缩影。有个叫 Theo 的人，我不知道你关没关注他。 So microcosm of this. There's a guy, Theo, I don't know if you follow him. 292:50 关注，关注，关注了。 Yeah, yeah, yeah. 292:51 呃，我们之前闹过一点小矛盾，因为，呃，我们有个项目不再用 TypeScript 了—— Um, we had a bit of a, a thing over the fact that, uh, we stopped using TypeScript- 292:57 对。 Yeah 292:57 ……他对此不爽，对那件事和其他一些做法都有意见。那之后我就想：“这什么憨憨。”然后，你知道吗？ ... for a project, and he didn't like that, and had opinions about that and other actions. And after that, I thought, like, "What a bozo." And, and do you know what? 293:06 我现在仍然觉得那步棋很憨，但我同时也会想，这家伙对 AI 是真上头，什么模型都在试，我们可以在这方面分享一些兴奋，所以我昨天刚开始关注他。 I still think that was a bozo move, and then I can also think, like, well, the dude's really into AI. He's trying all these models, and we can share some excitement about that, so I just started following him yesterday. 293:16 赞。 Awesome. 293:16 我就想：“你知道吗？我不必在所有事情上都同意你。呃，我甚至不必喜欢你做过的或说过的一些事情。我照样可以关注你。” I was like, "Do you know what? I don't have to agree with you on everything. Uh, I don't even have to like some of the things that you did or said or whatever. I can follow you anyway." 293:26 所以，如果我都能回关 Theo，那，呃，我觉得世界和平还是有希望的。 So if I can follow back Theo, then, uh, I think there's room for world peace here. 293:33 对，他其实是个很好的例子。他，他特别有主见，呃，改主意也挺勤，有时候阴阳怪气，但我觉得他骨子里是个好人。 Yeah, he's actually a really good example. He, he's really opinionated, uh, changes his mind quite a bit, can get snarky, but I think he's a good person underneath. 293:43 反正不管怎样，我就是，你知道吗，当有些人在谈论我关心的事情时，我的信息流就更有意思了，而且，呃，对，我不必在所有事上都同意你。我就是不必。 And either way, I just, you know what, my feed is more interesting when there's some people talking about things that I care about and, uh, yeah, I don't have to agree with you on everything. I just don't. 293:54 有件事我觉得挺有意思的，就是，呃，你提到过你妻子说的一句让你印象深刻的话：科技圈这些男性对极端长寿的痴迷，就像女性的厌食症， This, this I found pretty funny, that, um, you mentioned something your wife said that stuck with you, that all this tech-adjacent extreme longevity focus in men is like anorexia in women, 294:04 是焦虑和失控感在身体上的体现。 a physical manifestation of anxiety and lack of control. 294:11 这话不知怎的，听着还真有点道理。 That somehow rang true a little bit. 294:13 对很多人来说确实如此。我觉得那条推文是真的火了，我不知道你注意到没有，呃，布莱恩·约翰逊还在帖子下面回复了，我能理解布莱恩为什么会觉得那话有点， It certainly did to a lot of people. I think that tweet really popped off, and I don't know if you noticed, but, uh, Brian Johnson chimed in on, on the thread, and I could see how Brian would read that as a bit of a, 294:26 有点扎他。 of a jab. 294:29 嗯哼。 Mm-hmm. 294:30 而且我确实觉得，呃，Jamie 说那番话的时候，脑子里想的很可能就包括他。这也是一个例子：尽管我并不认同布莱恩的使命—— And I do think that, um, Jamie was probably thinking of him amongst other people, uh, in that conversation, and it's also an example of where even though I'm not subscribing to Brian's mission of, 294:45 我记得他的口号是“别死”—— I think it's the slogan's "Don't die"- 294:52 对。 Yeah 294:53 ……我其实，呃，我是真的想死的。我，我不想永远活下去。比如，我觉得人类九十到一百岁左右的寿命就挺合适。 ... I actually, uh, I do wanna die. I, I don't wanna do this forever. Like, I think the human li- lifespan of about 90 to 100 sounds about right. 295:03 我拥抱生命的有限性。我们不必在这上面达成一致。 I'm embracing the finitude of life. We don't have to agree on that. 295:09 事实上，正因为我们在这一点上不一致，布莱恩对我来说才是一个有趣得多的人。布莱恩愿意为了那个使命，实实在在地把自己的皮肤和身上各种其他零件都押上去，这真的很有意思。 In fact, Brian is a vastly more interesting human to me because we don't agree with that, and Brian's willingness to literally put his own skin and all sorts of other body elements in the game for that mission is really interesting. 295:29 你看，我可以持有这种看法：这事儿真的很有意思。 Now, I can have that view, this is really interesting. 295:33 我很高兴布莱恩在追求他这份热爱，同时也能觉得我妻子的分析很有道理，就是确实能感觉到背后有某种， I'm glad Brian is pursuing this passion of his, and then also find my wife's analysis to ring true, that it does seem like there is some underlying, 295:46 呃，我不知道，对……的恐惧。 uh, I don't know, fear of, uh... 295:52 我昨天还跟 Jamie 聊这个，她说，呃，或者把它比作这种感觉：有些人觉得自己活得不够尽兴。而他们活得不够尽兴的原因之一，也许是现代社会现在是一个监控社会， I was talking to Jamie yesterday about this, and she was saying, um, or, or comparing it to this sense of people who d- didn't feel like they had lived enough And one of the reasons perhaps that they hadn't lived enough was that modern society now is a surveillance state, 296:06 实际上不是国家在监控，而是互相监控。到处都是有摄像头的手机。嗯哼。 not by the state actually, but by each other. There's camera phones everywhere. Mm-hmm. 296:15 每一次失态，甚至每一次尽兴或尴尬的发作，都极有可能被录下来，然后被传播到被人嘲笑的地步。 Every indiscretion or even outburst of fun or cringe is highly likely to be recorded and then shared to the point of ridicule. 296:26 在这种条件下，人类的理性反应就是收缩，就是确保自己不会旁若无人地起舞——对…… And what is the rational reaction for humans under those conditions is to pull back, is to make sure you don't dance like nobody's watching- Yeah ... 296:35 因为很可能有人正在拍你，让你在互联网上丢大人、丢一辈子，所以也许你干脆就不该跳舞。 because they're probably filming you to your great and eternal embarrassment on the internet, so maybe you just shouldn't dance at all. 296:43 于是，如果我们都退缩到几乎没在活着的地步，我们就会死抓着“想永远活下去”不放。对。对。 And therefore, if we've all retracted to the point that we're barely living, we're clinging on to wanting it to last forever. Yeah. Yeah. 296:53 当然，我确信这套说法并不适用于所有情况。 Now, I mean, I'm sure that thesis does not apply in all circumstances. 296:57 我甚至不知道它适不适用于布莱恩，或者在座的任何人，但我觉得这里面有点东西：如果你觉得自己真的活过了， I don't even know if it applies to Brian or, or anyone else here, but I think there's something to this, that if you feel like you've really lived, 297:07 你就能坦然地想：“我活够了。”而生命有尽头这件事，不是需要对抗的。 you're okay thinking, "I've lived enough," and that the, it having an end is not something to be fought. 297:18 当然，也有可能这一切都只是一种存在主义的事后合理化，如果明天出了长生不老药，我们全都会吃。对，这也有可能。 Now, is it also possible that this is all just existential post-rationalization, and if tomorrow there was a live forever pill, we'd all take it? Yeah, that's possible. 297:29 有没有可能，如果我们真那么做了，社会反而会变得更糟？有可能。 Is it also possible that if we did that, society be- would be worse off? Yes. 297:34 有没有可能，人类寿命之所以是这个长度，是出于社会学原因，而不只是生物学原因？有可能。 Is it possible that the human lifespan is the length it is for sociological reasons, not just biological reasons? Yes. 297:45 所以我觉得这些都是很有意思的问题，我就是觉得她的分析真的点到了一些东西，而且这个话题后来又被翻出来讨论过。 So I think these are just all interesting questions, and I just thought her analysis really got to something here, and that discussion has come up afterwards. 297:56 我是说，那条推文好像都一年了。但最近我们又聊起这个，关于那些过度优化者。我看到克里斯·威廉姆斯在 Modern Wisdom 上有段话，他说：“是啊，好吧，也许我们确实有点过头了。” I mean, that, I think that tweet is, like, a year old. But just recently we've been talking about this, uh, over-optimizers. I saw, uh, Chris Williams on Modern Wisdom had a bit where he was like, "Yeah, okay, maybe we did go a little overboard." 298:09 而引爆这个话题的是《Diary of a CEO》，主持人大意是说：“我喝了一杯葡萄酒，接下来三天全毁了。”好吧。 And the triggering thing was the diary of a CEO, where the host was saying something to the effect of, "I had a glass of wine and the next three days were ruined." Okay. 298:20 这话对我来说，我是说，在他的情况下，我相信是真的。如果你很久很久滴酒不沾，然后只喝了一点点，我能理解它确实会有影响。 And that to me, I mean, in his circumstance, I'm sure that's true. Like, if you have not drank any alcohol for a very long time and you have just a little, I could see how it can have effect. 298:29 但这话里也有点东西，触动了很多人的本能反应，包括我自己，我当时就想：“靠，哥们儿，我现在就该来一杯葡萄酒——”对……尽管我平时不怎么喝酒。对。 But also there's something in that that just triggered a visceral reaction, I think, for a lot of people, and myself included, where I was like, "Fuck, man, I should have a glass of wine just right now-" Yeah ... even though I don't usually drink. Yep. 298:41 而且，事实上，呃，我觉得酒这个问题是个很好的切入口，可以深入看看。显然我们全都不喝酒了。酒类销量大幅下滑——嗯哼……至少在年轻人中是这样。 And, and in fact, uh, I think the alcohol question is a good microcosm to zoom in on. So we've all stopped drinking apparently. Like, alcohol sales are way down- Mm-hmm ... certainly amongst young people. 298:52 然后一方面，从纯粹的健康指标来看，管他什么指标，我们会说：“哦，哇，这不是好事吗？”嗯哼。然后我又会想：“嗯，我不确定这是不是好事。” And then on the one hand, on the pure health metrics, whatever, we can go like, "Oh wow, isn't that great?" Mm-hmm. And then I go like, "Yeah, I'm not so sure it is." 299:01 酒精能提供的社交润滑作用，可能被怀念，也确实被怀念。 Like, the social lubricant that alcohol can provide may be missed, is missed. 299:07 我是说，我们正处在孤独、痛苦、抑郁等等的全面大流行中，其中相当大一部分可能就是因为人与人的互动不够。 I mean, we are in an absolute epidemic of loneliness, and misery, and depression, and, and whatever, and a fair amount of it just probably comes because you're not interacting enough with other people. 299:19 这还没说到找对象的事，对吧？比如，呃，大幅下滑，人们很难遇到伴侣。而——对…… Now, that's before we even talk about coupling, right? Like, uh, way down, people are having a very hard time meeting a spouse. And- Yeah ... 299:27 很多这方面的事都转移到了各种 App 上，而这些 App 带来了各种负面结果和后果，对吧？ a lot of that has moved over to apps that have all sorts of negative outcomes and consequences, right? 299:35 比如，也许我们回头看，或者我们已经回头看了，会想：“呃，你知道吗？也许偶尔喝个烂醉，并不是世界上最糟糕的事。” Like, maybe we will look back, or maybe we al- we are looking back and thinking like, "Eh, do you know what? Maybe getting wasted every once in a while was not the worst thing in the world." 299:43 不管烂醉不烂醉，也许只是每周五或周六出门在外的时候喝上几杯，也是社会能走到今天的原因之一。 And whether wasted or not, maybe just having a couple of drinks every Friday or Saturday when you're out and about was part of what helped society get to where it is. 299:54 也许人类实实在在地喝了一万三千年酒这件事，是有其社会目的的。 Maybe the fact that humans have literally been drinking for whatever it is, like, uh, 13,000- years for mending things, had a social purpose. 300:03 结果我们把切斯特顿的栅栏拆了，自食其果，对吧？ And we pull out Chesterton's fence of our own chagrin, right? 300:08 就像你说的，本着这个精神，我当年穿越中国农村旅行的时候，我二十出头那会儿确实抽过一点烟，但为了那趟旅行我又捡起来了—— As you mentioned, in the spirit of that, so when I traveled across rural China, I did smoke a little bit in, in my early 20s, but I picked up just for that trip- 300:19 一种社交纽带，对吧？ A social connector, right? 300:22 而且我喝了海量的酒，尽管我现在基本不喝酒，就是因为社交。 And I drank humongous amounts, even though I don't really drink these days, because of the social thing. 300:30 对。 Yes. 300:30 那是他们表达爱的方式。 It's the way they show love. 300:31 对。 Yes. 300:31 而且，我还吃了巨量的碳水—— And, and I also ate shitload of carbs- 300:34 对。 Right 300:34 ……而我平时是—— ... which I'm usually- 300:35 对。 Right 300:35 ……低碳水的。我，我就是爱吃肉。就是因为那是他们表达爱的方式，而且—— ... low carb. I, I just like meat. So just because it's the way they show love, and- 300:40 对。 Yes 300:40 ……实际上，呃，在，在那个社会里是香烟，但在大多数社会里是酒，是一起做的事。 ... actually, uh, e- in, in that society, uh, cigarettes, but in most societies, liquor, is a thing to do together. 300:48 对。 Yes. 300:48 那是一种纽带。 And it's a bond. 300:50 对。对。 Yes. Yes. 300:50 它并不……它带来的好处，在很多情况下，可以超过负面后果，比如长期后果，不管你怎么衡量。 It doesn't... The, the benefits that that entails can, in many situations, outweigh the negative consequences, like long-term for whatever, however you measure them. 301:02 而这就是现代性的短视：我们把事情简化成，某项长期研究显示一杯酒会让你患某种心脏缺陷的风险高0.1%。 And this is the myopia of modernity, that we reduce things down to like, well, one glass on this long-run study meant that you had a 0.1% greater risk of some heart defect. 301:15 是，但如果你压根没活过，那又有什么意义？当然，也很容易滑到另一个极端，比如唯一的活法是每周末都喝大、吃一堆垃圾食品、一天抽四十根烟。 Yes, but if you didn't live at all, what's the point? And again, it's also easy to flip over the other side. Like, the only way to live is to get drunk every weekend and eat a bunch of crap and smoke 40 cigarettes a day. 301:27 不，不是这样的。这事儿不是非此即彼。这也是我停戴 Oura 戒指的原因之一。 No, it's not. Like, it doesn't have to be the either/or. This is one of the reasons I stopped wearing my Oura ring. 301:33 我戴 Oura 睡眠戒指戴了很多年，四年，然后有一天我突然想：“我不需要知道自己—— So I wore the Oura sleep ring for many years, four years, and one day I was just like, "I don't need to know that I had- 301:43 对。 Yeah 301:43 ……昨晚没睡好。” ... a bad night of sleep." 301:44 对。 Yeah. 301:45 我自己知道啊。这增加了一种，我不想叫它焦虑，我没多少焦虑，但它就是增加了一种提醒—— Like, I know. This added, I don't wanna call it anxiety, I don't have a lot of anxiety, but this added just like reminder- 301:55 嗯哼。 Mm-hmm 301:55 ……提醒我昨晚没睡好，这到底有什么实际用处？我为什么需要所有这些数据？我为什么需要优化一切？ ... that I had a bad night of sleep, what purpose is it actually playing here? Why do I need to have all these stats? Why do I need to optimize everything? 302:05 以至于，又一次，在听了《Diary of a CEO》里那个小段子之后，我就想：“靠，我有点想要反着来。”比如—— To the point where, again, after that single anecdote, you know, from Diary of a CEO, I was like, "Fuck, I kinda want the opposite." Like- 302:13 对。 Yeah 302:13 ……我平时几乎滴酒不沾。大概一个月一杯葡萄酒。我就想，你知道吗？我应该开始，呃，呃—— ... I don't barely drink at all. Like, a glass of wine once a month maybe. Like, do you know what? I should start, uh, uh- 302:21 嗯。 Mm 302:21 ……就在周末，来一杯。也许两杯。 ... just on the weekends, get a glass. Maybe two. 302:24 对。 Yeah. 302:25 而且，我是说，我从来不是那种为饮食纠结的人。呃，这是娶了一个……的巨大特权之一。 And also, I mean, I've never been the person to fret about my diet. Um, this is one of the great privileges of being married to... 302:37 这里我插一句题外话。永远别管一个四十多岁的女人叫“wonderful woman”（贤淑女士）。我妻子气坏了——…… And I'll do a quick aside here. Don't ever call a woman in her in her 40s a wonderful woman. My wife was so pissed- ... 302:44 不仅因为去年她生日那天我上了这个节目，这已经够糟了，还因为我在节目里叫她“wonderful woman”，因为那听起来就像她叫南希、六十五岁—— that not only last year did I come on her birthday, that was bad enough, but that I called her a wonderful woman on the show, because that sounded like her name was Nancy and she was 65- 302:55 对，对。 Yeah. Yeah 302:55 ……住在罗切斯特郊区。 ... in the suburb of Rochester. 302:57 对。 Yeah. 302:57 所以我就想，我这是要援引第五修正案了。比如，不，我不知道啊，我是丹麦人，我怎么会知道？但这招没管用。所以我改说“awesome woman”（了不起的女人）。 So I'm like, I was trying to play, plead the, the Fifth here. Like, no, I don't know, I'm Danish. How, how would I know? But that didn't go over well. So I'm gonna say awesome woman instead. 303:06 嗯哼。 Mm-hmm. 303:07 如果你，娶了一个了不起的女人，她不仅会做饭，还享受做饭，呃，你知道吗？ If you- marry an awesome woman who also not only knows how to cook, but enjoys it, um, do you know what? 303:14 这是一种非常传统的劳动分工，在人类社会里稳定存在了大概几万年，也许这也不是世界上最糟糕的主意。 There's a very traditional division of labor here that has been quite stable through several tens of thousands of years in human, uh, societies that perhaps also wasn't the worst idea in the world. 303:29 我也知道有很多男人喜欢做饭，也有女人不喜欢，但就刻板印象而言，呃，它确实帮到了我。 And I also know there are plenty of men who like to cook, and there are women who don't, but as stereotypes go, uh, that certainly helped me. 303:38 比如，我，我觉得，呃，如果我不是日常吃得相对健康，我可能没法对食物这么放松。 Like, I, I don't think, um, I would've been able to just be as relaxed about food if I didn't eat relatively healthy on a, on a regular basis. 303:46 但总的观点就是：退出那场优化游戏，同时也不走向反面。你也不必做沙发上摊着的肥宅。你，你可以动起来。 But the overall point being just getting out of that optimization game without turning into the opposite. You don't have to be a fat slob on the couch either. You, you can move around. 303:59 对。然后我们在谈的更大的话题是，呃，生命的有限性。 Yeah. And then the bigger point we're talking about is, um, the, the finiteness of life. 304:05 对。 Yes. 304:05 这一点我同意你。 And I agree with you on this. 304:06 这其实是我做进 Omarchy 里的一个功能。就是 memento mori。 This was actually a feature I built it into Amaci. So memento mori. 304:12 嗯。 Yeah. 304:12 记住你终有一死。 Remember death. 304:14 对。 Yeah. 304:14 在 Omarchy 里，你点一下时钟，往下拉就能打开日历。 If you pull down the calendar in Amaci, you open it by clicking the clock. 304:18 嗯哼。 Mm-hmm. 304:18 打开之后，第一个好玩的地方是，这是我从网上学来的。有个叫 Years Progress 的账号，它每天更新，显示“2026年已过去63%”。不知道你见没见过那个。 It opens up, and the first funny thing, I took it off the internet. There's this, uh, account called Years Progress, and it, every day it updates, and it just show 63% done with 2026. I don't know if you've seen that one. 304:30 见过。 Yeah. 304:30 它就是一个进度条不断填满。我把它做进了日历里。所以现在它会显示：2026年我们已经过完了63%。 So it just fills up. So I built that into the calendar. So right now it says we're 63% done with 2026. 304:35 嗯哼。 Mm-hmm. 304:35 但如果你双击它，它会问你：“你哪年出生的？你预计活多久？” But then if you double click that, it'll ask you, "When were you born? How long do you expect to live?" 304:40 嗯哼。 Mm-hmm. 304:40 默认是90岁。我填了1979年。 And the default is 90. So I put in 1979. 304:43 对。 Yeah. 304:43 然后在“这一年过完了多少”那行下面—— And then right below the line of like, how long are we done with the year- 304:47 哦，太妙了。 Oh, that's beautiful 304:47 ……多了一行，显示你的人生已经过完了多少。 ... is a new line that shows how long are you done with your life. 304:50 对。 Yeah. 304:50 我记得我是62%。 I think I'm at 62%. 304:52 对。 Yeah. 304:53 鼠标悬停上去，它会显示 memento mori。 I'm like, if you hover over it, it says memento mori. 304:55 太美了。 That's beautiful. 304:57 不少人觉得这有点阴间，但这正是它做成彩蛋的原因。它不是明摆着的。你得点好几下才能找到。我喜欢这个提醒。我喜欢它提醒我时间是有限的，我应该充分利用它。 Fair number of people found it morbid, but this was why it is an Easter egg. It's not like exposed. You gotta click a few things to get there. I like the reminder. I like the reminder that time is finite, and I should make the most of it. 305:09 这也是有孩子的特权之一。你会真切地意识到，那些陈词滥调全是真的。我的天，时间过得太快了。 And this is also one of the privileges of having children. You really realize that the cliches are all true. Oh my God, it goes so fast. 305:19 嗯哼。 Mm-hmm. 305:19 我记得我现在十三岁的孩子一岁、两岁的时候。我就想：“这明明就是昨天啊。”它之所以成为陈词滥调，是因为这是共同的体验。陈词滥调是好东西。 And I remember my now 13-year-old being one, being two. I was like, "This was just yesterday." And the reason it's a cliche is because it's a shared experience. Cliches are good. 305:31 它们提醒你，你是人类共同体验的一部分，别人在你之前也经历过。 They remind you that you are part of the human experience, and others have been here before you. 305:38 我得问一下，你，你刚才说了不起的女人，呃，而，而且，呃，厨艺很棒。 I have to ask, you, you mentioned the amazing woman, um, and, and, uh, great cook. 305:43 了不起。 Amazing. 305:44 了不起。 Amazing. 305:44 这个一定要记住。不是贤淑（wonderful），是了不起（amazing）。 Very important to remember. Not wonderful, amazing. 305:47 了不起。那么这事儿……你之前说过可颂的坏话还是怎么的。我在一条推文里看到的。所以丹麦可颂就是更好？ Amazing. So there's this... You talked, you talked some shit about c- croissants or something. I saw this s- in a tweet. So is the, the Danish croissant is somehow better? 305:59 其实，丹麦的可颂也不怎么样。在我看来，你唯一能找到像样可颂的地方，说出来好笑，居然是 7-Eleven，这让我觉得特别逗，因为美国的 7-Eleven 是那种——至少在芝加哥—— Actually, Danish croissants are not particularly good. In my opinion, the only place you can find a half-decent croissant, of all places, is 7-Eleven, which is hilarious to me, because 7-Eleven in the US is the place where- At least in Chicago, 306:10 你去买杯冰沙、顺便祈祷别被枪击的地方。 you'd go to get a slushie and try to avoid getting murdered. 306:15 对。 Yeah. 306:15 但在丹麦，7-Eleven 是那种非常高档、特别讲究的便利店，而且它们碰巧做的可颂还不错。 But in Denmark, 7-Eleven is this very upscale, super nice convenience store, and they just happen to make decent croissants. 306:25 对。 Yeah. 306:25 我一直想不通的是，我在美国住了他妈二十年，居然没吃到过一个能达到哥本哈根 7-Eleven 水平的可颂。 And it always baffled me that I lived in the United States for 20 freaking years, I have yet to have a single croissant that's even at the level of a 7-Eleven croissant in Copenhagen. 306:39 这听起来不太可能是真的。 That seems improbable. 306:41 你搞明白为什么了吗？ Have you understood why? 306:43 没有，我，我是认真考虑过雇个调查记者把这事查个水落石出的。因为我听到的理论五花八门。 No, I've, I've legitimately considered hiring an investigative journalist to get to the bottom of this. Because the number of theories I've heard have been all over the place. 306:53 是水的问题，是黄油的问题，是手艺的问题，是这、是那、是别的什么。我就觉得，这些说法没一个对得上。如果是黄油的问题，进口不就行了？ It's the water, it's the butter, it's the skill, it's the this, it's the that, it's the other thing. And I'm like, none of it adds up to me. If it's the butter, can't you just import it? 307:03 我记得我在这边的一些超市里见过丹麦黄油，那个 Lurpak。所以不可能是黄油。那是手艺问题吗？不，不可能是手艺。 I thought I saw Danish butter in some, uh, supermarkets over here, the Lurpak. So that can't be it. Can it be the skill? No, it can't be the skill. 307:11 美国住着一大堆法国人和比利时人，全世界最好的可颂师傅。到底是什么原因？然后我就听到第二种说法—— There are plenty of French people and Belgian people, the best, uh, croissant makers in the world, who live in the US. Like, what is the reason? And then I get the second thing, which is- 307:20 对。 Yeah 307:20 ……“不对不对，你只是没去对地方。”我实打实去过十五家所谓的—— ... "No, no, you just haven't gone to the, to the right place." I've literally gone to 15 of quote-unquote- 307:25 对。 Yeah 307:25 ……“对的地方”，结果可颂全是垃圾。 ... the right place, and all the croissants were shit. 307:28 对。 Yeah. 307:28 所以在我离开这个星球之前，我必须把美国为什么做不出一个像样的可颂这件事查个水落石出。 So before I leave the planet, I have to get to the bottom of why the United States cannot make a proper croissant. 307:36 而且我甚至不是要求顶级法国可颂、顶级比利时可颂那种世界上最好的可颂。我要的只是哥本哈根 7-Eleven 那种可颂。 And I don't, I'm not even asking about peak French croissants, peak Belgian croissants, which are the best croissants in the world. I'm just asking about the 7-Eleven croissant from Copenhagen. 307:48 这要求过分吗？ Is that too much to ask for? 307:49 这，呃，真是个迷人的悬案。呃，什么……往大了说，你最喜欢的……这让我特别想知道，你最喜欢的一顿饭是什么？比如，如果，如果我给你最后一餐，然后把你干掉。 That's, um, fascinating mystery. Uh, what... Zooming out, what's your, what's your favorite... This makes me wanna know, what's your favorite meal? Like, if you, if I gave you a last meal and were to murder you after. 308:02 这完全取决于它是午餐还是晚餐。如果是午餐—— It all depends whether it's lunch or it's dinner. If it's lunch, it- 308:05 就一顿饭。什么……我，我觉得既然你都要被干掉了，午餐还是晚餐真无所谓。 It's one meal. What's, what's... I, I feel like if you're getting murdered, it doesn't really matter if it's lunch or dinner. 308:11 嗯，这取决于你一天里几点被干掉。 Well, it depends on what time of day you're getting murdered. 308:13 哦，懂了。 Oh, I see. 308:14 如果我，如，如果是，如果是下午行刑，那我就只能吃午餐。 If I, i- if it's a, if it's an afternoon execution, I only get lunch. 308:18 好。 Okay. 308:18 如果吃午餐，那就是在，呃，Louisiana，哥本哈根北边的那个博物馆。 And if I get lunch, it's at, um, Louisiana, the museum in north of Copenhagen. 308:29 我两天前刚和 Jamie 去过，我们经常去。博物馆本身很美。但我们不是冲着博物馆去的。 I was just there with Jamie, uh, two days ago, and we go there all the time. The museum is lovely. We don't go for the museum. 308:37 嗯哼。 Mm-hmm. 308:37 我们是冲着自助餐厅去的。这本身听起来就很疯。比如—— We go for the cafeteria. That in itself sounds crazy. Like- 308:40 确实疯。 Yeah, it does 308:40 ……你要，你要开半小时车去一家自助餐厅？ ... you're gonna, you're gonna drive half an hour to go to a cafeteria? 308:43 嗯哼。 Mm-hmm. 308:43 可它偏偏就是我这辈子吃过最好的午餐。那里的厨师简直好得离谱。 Well, it just so happens to be, it is the best lunch I've probably ever had in my life. The chef there is just unbelievably good. 308:53 而且不只是食物惊人，对我这种系统思维的大脑来说更惊人的是，我们在那儿的时候，我估计任何时候都有两百来号人在等餐。 And not only is the food amazing, what's even more amazing to my system thinking brain, is that at any one time when we were up there, I think there were probably 200 people waiting for their food. 309:04 哦，靠。 Oh, shit. 309:04 而我们的餐五分钟就上来了。他们菜单很小，显然是大批量做的，而且好吃得离谱。 And our food was delivered in five minutes. They have a small menu, and clearly they make it in bulk, and it's unbelievably delicious. 309:15 就是一家博物馆的自助餐厅？ Just a cafeteria at a museum? 309:17 是一家博物馆的自助餐厅，碰巧有我这辈子在任何地方吃过的最好的午餐。 It's a cafeteria at a museum that happens to have the best lunch I've ever had anywhere in my life. 309:20 那到底是什么样的食物？什么样的午餐—— Well, what kind of food are we talking about? What's, what kind of lunch- 309:23 就是北欧菜。所以是，是……有意思的是，我看着那份菜单就想，你知道吗？如果我小时候看到这份菜单，我会说：“我不在这儿吃。” It's, like, Nordic cuisine. So it's, it's... The funny thing is the menu, when I look at it, I was like, do you know what? If I saw this menu as a kid, I'd go like, "I'm not eating here." 309:31 好吧。 Okay. 309:32 就是，我不知道—— It's like, I don't know- 309:33 这挺精致的嘛。 This is pretty fancy 309:34 ……根茎类蔬菜和—— ... root free fruit and- 309:35 对。 Yeah 309:35 ……还有洋葱和—— ... and onions and- 309:36 对。 Yeah 309:36 ……各种健康的东西。听起来就难吃。然后你坐下来一吃，天使都在唱歌。但还不止于此。是黄油。 ... all sorts of healthy stuff. That sounds disgusting. And then you sit down and eat it, and the angels sing. But it's even more particular than that. It's the butter. 309:45 对。 Yeah. 309:46 他们那儿的打发黄油—— The whipped butter that they have there- 309:49 对。 Yeah 309:49 ……全世界最好的黄油。 ... the best butter anywhere in the world. 309:50 我觉得这肯定跟可颂悬案也有关系。 I think that has to be connected to the croissant mystery too. 309:52 肯定，肯定有点关系。但真的——……就算在巴黎，我也没吃到过这么好吃的打发黄油——不如哥本哈根 Louisiana 的黄油好吃。 There, there m- there must be some connection. But literally- ... even in Paris, I've never had whipped butter that tasted- As delicious as the butter does at Louisiana in Copenhagen. 310:04 说真的，我愿意开三十分钟车—— Literally, I would drive 30 minutes- 310:06 对。 Yeah 310:06 ……就为了去 Louisiana 的自助餐厅吃面包配黄油。 ... to have bread and butter at the cafeteria in Louisiana. 310:10 那你是怎么发现的？就是某天去了博物馆然后—— Well, how'd you discover that? Just one day you went to the museum and- 310:13 我想我们就是去了博物馆—— I think we just went to the museum- 310:15 然后你们尝了一下。我的天。 And you tried it. Holy shit 310:15 ……然后吃了午餐，我们就想—— ... and we had the lunch, and we were like- 310:18 这也太离谱了。 This is weird 310:19 ……这都能跟我这辈子去过的最好吃的餐厅媲美了——……我去过 Noma 几次，该尝的都尝过，而这家绝对在同一水平。 ... this is rivaling the best tasting restaurants- ... I've ever been to in my life, and I've gone to Noma a couple of times and tried what there is to try, and this is right up there. 310:31 好。那晚餐呢，既然你提到了？是不是得另想一家？ All right. What about dinner, since you mentioned? Is it something different that you have to think? 310:36 嗯，我觉得哥本哈根有很多好选择，但其实我会选西班牙的一家。就是马贝拉有个叫 Puente Romano 的地方。 Well, I think there's a lot of good options in, uh, Copenhagen, but actually I'll pick a place in Spain. So there is, um, this place in Marbella called Puente Romano. 310:49 那是一家酒店，在海滩边他们有一家吃新鲜海鲜的海滩餐厅。 It's a hotel, and down by the beach they have this beach restaurant with fresh seafood. 310:56 嗯。 Mm. 310:57 好吃得离谱。我想不起来它叫什么了，但你查一下就知道，Puente Romano 那家海滩海鲜餐厅简直绝了。 Unbelievably delicious. I'm trying to remember what it's called, but yeah, if you look it up, the beach seafood place at Puente Romano is incredible. 311:08 美食里头，有多少是那个地点、和你在一起的人，对吧？或者说你第一次在那儿的体验？ How much of great food is, like, the place and the p- the person you're with, right? Or the, or, or the experience when you first are there? 311:18 我是说，氛围肯定加分，但我觉得我的味蕾也确实—— I mean, the ambiance certainly helps, but I think my taste buds are also just- 311:22 好，那，那—— Okay, that's, that's- 311:22 ……对味道很敏锐—— ... in tune for- 311:23 所以是真材实料—— So it's legit- 311:23 ……的味道。 ... flavors 311:24 ……就是好吃。 ... it's just good food. 311:24 就是真的好吃。 It's just really good food. 311:27 因为就拿那家自助餐厅来说，人特别多。我，我不会把它评价成那种“天哪，我愿意在这儿坐两小时”的地方，但食物太好吃了。 Because for ex- that cafeteria, it is, there's a lot of people. Like, it, I wouldn't rate it as like, oh my God, this is just a place I wanna sit for two hours, but the food is so delicious. 311:37 这也是我在全国旅行中意识到的一点：不带偏见地去尝，什么都试试。 That's one of the things I'm, in, in traveling across the country, I'm realizing to s- sample without bias, just try things. 311:47 对。 Yep. 311:47 因为你可能会，我觉得你可能会被惊艳到。也许我能帮你找到你的可颂。好了，我们聊了死亡，聊了食物。 'Cause you might, I guess you might just be surprised. Maybe I'll find your croissant. All right, we've talked about mortality. Talked about food. 311:56 把人生终点都安排明白了，呃…… Settled end of life, uh ... 311:58 你觉得，呃，一千年以后，人类、人类文明的未来会是什么样？你觉得我们能行吗？ What do you, uh, thousand years from now, what do you think is the future of humans, human civilization? You think we're gonna make it? 312:04 一千年以后？ A thousand years from now? 312:05 对。 Yeah. 312:06 哥们儿，我连十二个月以后都预测不了。 Man, I'm having trouble predicting 12 months from now. 312:09 我知道。 I know. 312:09 一千年？ A thousand? 312:10 你觉得我们有机会吗？ You think we got a shot? 312:11 有，我，我赌乐观这一边。我赌我们会成为多行星物种。 No, I'm, I'm gonna bet on optimism. I'm gonna bet on a multi-planetary species. 312:15 我赌，呃，埃隆能把火箭送上火星，然后我们能搞明白怎么弯曲，呃，时空，然后发现一些虫洞之类的。 I'm gonna bet that, um, Elon get that rocket to Mars, and we figure out how to bend, uh, space time and, uh, discover some wormholes or, or something. 312:24 一旦做到了，你觉得，等我们有了一份靠谱的备份之后，地球上的人类会灭绝一两次然后再重新繁衍吗？ Once we do, do you think, you think all humans on Earth will die once or twice and be repopulated after we get a good backup going? 312:35 我是说，那其实是我最喜欢的《黑镜》剧集之一，就是他们被上传到八十年代某个海边小镇的那一集。 I mean, that was actually one of my favorite episodes of Black Mirror, the one where they get uploaded to the 1980s sea town somewhere. 312:44 呃，我当时就想，哇，这一刻既如此黑暗，又如此美好，因为大家共同承认：八十年代其实挺了不起的。 Um, and I just thought, like, wow, this is both such a dark moment, yet also so beautiful in the shared recognition that, like, the '80s were actually kind of amazing. 312:55 我对很多时代都没什么怀旧情结，但我确实怀念八十年代。 Like, I don't have a lot of nostalgia for many moments, but I do have nostalgia for the '80s. 313:01 对。 Yeah. 313:01 他们说，不管你身处何时何地，你十二岁或者—— And they say something about, like, whatever place and time you were in, the music you were listening to when you were 12 or- 313:07 对。 Yeah 313:07 ……或者十四五岁的时候听的音乐，就是会跟你一辈子的东西。我呢，八十年代的时候我才十岁。 ... or 14, 15, something like that, that's what's gonna stick with you. Like, no, I was only, I got to be 10 in the '80s. 313:15 所以那是我的童年早期，而我对八十年代有一种难以置信的深厚感情。所以如果我要被上传到云端—— So this was my early childhood, and I still have an unbelievably fond affiliation with the '80s. So if I'm gonna be uploaded to the sky- 313:23 你要去八十年代。 You want it to be the '80s 313:24 ……我选八十年代。 ... I'm doing the '80s. 313:26 哇。我好像还真没听人这么说过。一般都是七十年代或九十年代。 Wow. I don't think I've actually ever heard anyone say that. Usually it's '70s or '90s. 313:31 哦，还有—— Oh, and the- 313:31 因为八十年代是—— 'Cause the '80s is the- 313:32 我是真不喜欢九十年代。我们能聊聊这个吗？ I really don't like the '90s. Can we talk about this? 313:38 不，我说完了。不，我没说完。 No, I'm done. I'm not done. 313:40 对我来说，九十年代是虚无主义的大转折点。 To me, this was the great turning point of nihilism. 313:45 嗯。 Mm. 313:45 音乐、流派、时尚—— That both the music, the genre, the fashion- 313:50 嗯。 Mm 313:50 ……一切都从那种华丽、流行、乐观，甚至雅皮士的那一套—— ... everything turned from, like, this glamour, the pop, the optimism, even the yuppies and everything- 313:57 对。 Yeah 313:57 ……变成了垃圾摇滚、涅槃乐队和……虽然，我是说，我，我喜欢那音乐，但我不喜欢那套精神气质，我不喜欢—— ... to grunge and Nirvana and... Even though, I mean, I, I like the music, I don't like the ethics, I don't like- 314:05 对。 Yeah 314:05 ……那套道德观，它底下的东西我一样都不喜欢。 ... the morals, I don't like any of the underpinnings to it. 314:07 哦。 Oh. 314:08 那里面有一种彻底的失败主义—— There was just such a defeatism to it- 314:11 对。 Yeah 314:11 ……回头看，我回望那段时期，如果我要被重新插回矩阵，我选八十年代。 ... that in retrospect, I look back upon it, if I'm gonna get reinserted into the matrix, it's gonna be the '80s. 314:17 那是个充满乐观和乐趣—— That was full of optimism and fun- 314:20 对。 Yes 314:21 ……和色彩的年代。 ... and colors. 314:22 对。 Yes. 314:24 我记得我有一条橙色裤子—— I remember having these orange pants- 314:28 对。 Yeah 314:29 ……上面有白色的圆点。在八十年代，小孩穿这个再正常不过了。 ... with, like, white dots on them. That was just a normal thing kids could wear in the '80s. 314:35 嗯哼。 Mm-hmm. 314:35 现在我是真他妈没见过有人这么穿。呃，或者我是说，就算在九十年代，一切也都变成了那种西雅图灰。 I've never fucking seen that today. Um, or I mean, even in the '90s, it all turned just, like, Seattle gray. 314:44 嗯哼。 Mm-hmm. 314:45 大倒退。 Great regression. 314:46 所以按尼采那个，你知道，永恒轮回的说法。如果让你过《土拨鼠之日》那样永远重复的日子，你想活在八十年代，穿着带白点的橙色裤子。 So the Nietzschean, you know, this idea of eternal recurrence. If you do a Groundhog Day and you live forever, you want it to be in the orange pants with the white dots in the '80s. 314:55 也，也许橙色裤子就算了，但是—— May- maybe I'm gonna skip the orange pants, but- 314:57 不行。 No 314:57 ……但我绝对选八十年代。 ... I'm definitely picking the '80s. 314:58 这话可是你说的。 You said it. 315:01 好了，DHH，非常感谢你所做的一切。谢谢你激励了我们所有人，谢谢你为这个世界创造了很酷的东西，也，呃，谢谢你今天来聊天。 All right, DHH, thank you so much for everything you do. Thank you for being inspiration to all of us, for building cool shit in the world, and, uh, thank you for talking today. 315:11 随时乐意。Lex，再次谢谢你邀请我。 Anytime. Thank you, Lex, again for having me. 315:15 感谢收听这期与 DHH 的对话。如果想支持本播客，请查看简介中的赞助商信息，那里也有联系我、提问、反馈等的链接。 Thanks for listening to this conversation with DHH. To support this podcast, please check out our sponsors in the description where you can also find links to contact me, ask questions, give feedback, and so on. 315:27 最后，让我以拉尔夫·沃尔多·爱默生的一句话作结：“一旦你做出决定，整个宇宙都会合力助你实现。”感谢收听，下期再见。 And now, let me leave you with some words from Ralph Waldo Emerson. "Once you make a decision, the universe conspires to make it happen." Thank you for listening, and hope to see you next time.
没有匹配的文稿
24:52
13:06
11:15