显示语言 中英对照 中文 English
点击段落跳转到对应时间
0:00 马克，上次咱们通话是周末你给我打的，有意思的是你是用眼镜打的。 Mark, last time we spoke, you called me, it was on the weekend. Well, you called me through the glasses, which was the interesting part. 0:06 背景很吵吗？ Was it loud in the background? 0:07 你要去钓鱼。我本不想说 You were, you were going to fish. I wasn't gonna say it, but- 0:10 对，是周末。其实降噪效果还挺好的。 Yeah, it was on the weekend. The noise cancellation actually is pretty good. 0:13 哦很棒。 Oh, it's great. 0:14 对很棒。 Yeah, it's great. 0:14 嗯，对，是的。 Yeah, I, I, I mean- Yeah. Yeah. 0:16 你想聊那篇文章，宣言——我不知道该怎么称呼它—— Um, but you, you wanted to talk about this essay manifesto, I don't know what you call it, manifesto- 0:21 你最近发表的那篇。它很长，我得说里面内容很多。我想从这里开始，因为里面有很多重要观点，会引出我们今天要谈的主题。 we'll say, uh, that you published recently. And it, and it's long. I mean, I'll caveat that there's a lot in it. And I wanted to start there 'cause there's a lot of big ideas in there, and they'll connect to kind of the main thing we're talking about today. 0:32 我很好奇，你为什么要写这篇？它真的很长，内容很多。 Um, I'm curious, like, why write that? 'Cause it, it's long. There's a lot in there. 0:37 我觉得既然要投入这么多去构建 AI，就必须让大家明白你的实验室代表着什么—— Well, yeah. Well, I f- I feel like if you're going to invest so much in building AI, uh, then it's important that people understand what your lab stands for- 0:48 以及你的价值观是什么。AI 有很多机遇，但也确实存在各种风险。 and what your values are. And, and, and basically, AI has so many opportunities, but there are also all these real risks. 0:56 所以我认为，每个从事这项工作的人，都应该有一套深思熟虑的理论，说明自己的工作将如何通向积极的未来。 So I think it's very important that everyone who's working on it has a well-thought-out theory for how the work that they're going to do is going to lead to a positive future. 1:06 有意思的是，各实验室对此的理念不尽相同，行业里有些已成共识的观点，我其实是强烈不认同的。 And it's interesting because, I mean, the different labs have, have some different philosophies on this, and there's, you know, a lot of things that I think have become conventional wisdom in the industry that I just strongly disagree with. 1:16 在我看来，要让所有人拥有积极的未来，就要尽可能广泛地传播这项技术，这基于我们的三大原则。 And, you know, and my, my view on this is that the path to have a positive future for everyone is to make sure that we distribute the technology as widely as possible, and that's based on three major principles that, that we have. 1:31 第一，赋能个人是世界繁荣的源泉，历史上一直如此。 One is that empowering people is the source of prosperity in the world, and it has been throughout history. 1:37 第二，AI 的首要用途是发明新事物，而不是自动化。 Two is that the primary purpose for AI is going to be invention of new things, not automation. 1:45 第三条原则是，未来安全的基础在于建立正确的制衡与权力平衡，而不是限制准入。 And then the third principle is that the foundation for safety for the future is basically establishing the right checks and balances and balance of power rather than restricting access. 1:57 这些观点说来也怪，跟很多主流共识很不一样，尤其是在硅谷。 And I think that that's-- Th- these are all things that I think oddly are, are kind of very different from, I think, a lot of the conventional wisdom, um, especially in Silicon Valley. 2:07 很多人觉得，这技术太强大了，必须加以限制，不能让太多人接触到。 Pe- A lot of people think, "Hey, this technology is very powerful. We must restrict it so that way not that many people have access to it." 2:12 我个人更担心的是少数实验室或少数人掌控这么强大的东西。纵观历史，把权力交到人们手中时，多数进步并非来自在位者或既有体制。 I personally am much more worried about a small number of labs or people having control of, of something that is so capable, um, and I think that the-- throughout history, what we've found is that when you put power in people's hands, most advances don't come from the incumbents or the establishment. 2:31 它们来自边缘地带那些想法不被重视的人。但当他们获得足够的工具去证明自己的成果时，就会爆发出巨大的力量。 They come from people on the, on the periphery whose ideas aren't taken seriously. But when they get enough tools to, um, to, to basically be able to prove out what they're working on, that ends up being very powerful. 2:42 在西方社会，我们建立治理体系、实现社会平衡的方式，就是依靠一整套制衡机制和权力平衡。 In Western society, the way that we've established governance and, um, and basically having a well-balanced society is through a set of checks and balances, right, and, and this balance of power. 2:53 这种观念深深扎根于我们的社会：你不会想让一两家实验室独占某项技术。 Um, it's very ingrained in, in kind of our society that you don't wanna have, you know, one lab or two labs having access to a thing. 3:02 对于最近出现的一些担忧，比如网络安全问题，我认为，对付可能入侵系统的 AI，最好的办法是让人人都有 AI，先加固自己的系统。 For some of the most recent concerns that have come up, like some of these cybersecurity concerns, I think the, the best antidote to someone having an AI that could potentially hack into systems is having everyone have access to an AI so they can harden their own systems first. 3:16 这正是过去几十年网络安全的历史。开源软件因为人人都能查看、审视，把它交到人们手中，看似有违直觉， And I think that that's kind of been the history of cybersecurity over the past several decades is that, you know, open source software, um, because people can, can see it and can scrutinize it, it sort of counterintuitively by putting it in people's hands, 3:31 最终却能带来更安全、更稳定的环境。这是我的信念，我认为通往积极未来的路径就是广泛传播技术。 you end up with a more secure and more stable environment. So that's what I believe, um, and, and that's what I think is the path to, to a positive future is basically distributing the technology. 3:41 这对我们要做的事有很多影响。显然，我们要构建领先的 AI 模型，这也正是我们在做的。 That has a bunch of different implications for what we're gonna do. I mean, obviously, we wanna build leading, um, AI models, which, which, uh, we're doing. 3:49 我们刚发布的 Muse Spark 1.3 很先进，但它其实是一次相对较小规模预训练的最新模型，内部代号叫 Avocado。 I mean, Muse Spark 1.3, which we just released, is-- it's, um, it's advanced, but then, you know, it's actually the latest model of a relatively smaller pre-train that we did, the, the internal code name Avocado. 4:03 哦—— And, um- Oh, 4:04 我们要聊代号了。 we're gonna get into the code names. 4:05 有疑问。 I've got questions about those. 4:05 我们会聊到的。 Well, we'll, we'll get into that. 4:06 还有 Watermelon 快来了。 And, and we have Watermelon coming soon. 4:08 这会是件大事。显然，在领先模型中，最能体现这一愿景的，就是我们正在推出的 Muse 个人智能体—— So, so that's gonna be a big deal. So obviously, leading models, probably the biggest personification of, um, you know, if you will, or, or kind of, um, implementation of this vision is, um, the Muse personal agent that we're- 4:22 基本思路是：给世界上每个人一个非常能干的个人智能体，能理解他们的目标，全天候替他们工作。 that we're rolling out. Um, that basically the idea there is give every person in the world a very capable personal agent that can understand their goals and can just work on their behalf twenty-four/seven. 4:33 另一个重要环节是把技术交到人们手中。我们坚定支持开源，确保这里巨大的机会不只属于少数人或少数公司。 And then, um, an important part of this is also just getting the technology in people's hands. So we're, we're very kind of strong proponents of open source and, and making sure that, that the opportunities that I think are gonna be massive here are not just limited to a few, to a few people or companies. 4:48 本集由 Mercury 赞助，三十多万企业家喜爱的 AI 原生银行，包括我。 This episode is brought to you by Mercury, AI native banking that's loved by more than three hundred thousand entrepreneurs, including me. 4:54 上 mercury.com 了解。Mercury 是金融科技公司，不是银行。还要感谢 Granola，连场会议人士的 AI 笔记本。 Visit mercury.com to learn more. Mercury is a fintech, not a bank. Check the show notes for details. Thanks also to Granola, the AI notepad for people in back-to-back meetings. 5:03 它在哪都能用，让你专注要事。访问 granola.ai/sources，输入优惠码 sources，立减三个月费用。 It works everywhere you do and lets you focus on what matters. Try it at granola.ai/sources and use the code sources for three months off. 5:12 本集也由 Atlassian 的 Jira 赞助，让团队和智能体获得上下文、协同与掌控，推进工作。 This episode is also brought to you by Jira by Atlassian, where teams and agents get the context, coordination, and control to move work forward. 5:21 到 jira.com 免费试用，就是 J-I-R-A.com。关于 Muse 智能体我有很多问题，但先谈大局—— Try it free at jira.com. That's J-I-R-A.com. I have a bunch of questions about the Muse agent, but staying big picture- 5:30 你刚才说了很多，开源就是对抗你担忧的那个趋势的办法吗？ for a second 'cause you said a lot of things there, is open source the counter to the trend you're seeing that you describe that you're worried about? 5:37 这是你主要的实际对策吗？还是靠监管？还是两者都要？ Is that the main way practically that you counter that, or is it regulation? Is it both? 5:43 比如，不 Like how- No, 5:44 其实我认为最重要的，就是把技术交到个人手中。 I actually think probably the most important thing is actually just getting the technology in individuals' hands. 5:49 所以我认为像 Muse 个人智能体这样的，可能更重要。 So I actually think things like the Muse personal agent, um, are perhaps even more important. 5:55 你开始看到，一些实验室训练出更先进的模型，却不发布。 I mean, I think what you're starting to see are some of the labs are building, training more advanced models and then not even releasing them. 6:02 我认为这相当危险。因为当一件事受到审视，把系统交到很多人手中，就会有制衡，带来广泛的繁荣，这很重要。 Right? So I think that that is quite- Dangerous in the sense that, you know, basically when you have scrutiny on something, when you put a, a system out there, first of all, if you put it in a lot of people's hands, you get the checks and balances, you get broad-based prosperity, which I think is important, 6:17 对社会而言，不能只有一两家实验室变得极其有价值。要让几十亿人都能获得繁荣，无论创办小企业，还是事业更成功。 right? For society, we can't just have, like, one or two labs get incredibly valuable. Um, you know, I think you, you wanna make it so that billions of people can basically have prosperity in their own lives, whether it's creating small businesses, being more successful in their careers, um, 6:32 还是更高效地管理家庭、在许多方面省钱、促进健康。 being more productive in managing their homes, saving money in a lot of ways, um, kind of advancing their health. 6:37 所以收益要足够广泛，这是一方面。至于竞争，多家实验室并存是好事，开源对此很有帮助。 Uh, so I think you, you want the benefits to be very broad-based. So that's one piece. In terms of the competition, I do think that having multiple labs is helpful, and I think open source is quite helpful for that. 6:50 开源是重要一环，但它的本质是整个社区一起做，不会只有我们一家。 So I think open source is an important part of it. The nature of open source is there's a whole community of people who do it, so I'm not saying that we're gonna be the one company that does it. 6:57 我也不是开源狂热分子，我们做的一切也并非都开源。 I'm also not, like, a zealot about this from the perspective it's... You know, it's not that everything we do is open source either, right? 7:03 我们发布一些开放模型，也做闭源工作。作为营利公司，要能打造先进的东西，不必样样都公之于众。 We, we release some open models. We do some closed work. I think it's important if you're building a for-profit company that you can build some advanced things, and you don't necessarily need to share every single thing with the world. 7:13 但总体而言，支持健全的开源生态，是维持竞争的关键，也能让技术走向保持透明可理解，我认为这对安全极其重要。 But I think in general, uh, supporting a robust open source ecosystem is going to be key to maintaining competition and, and maintaining kind of transparency and understandability of where the technology is going in a way that I think is actually going to be incredibly important for safety. 7:31 如果世界上只有少数几个真正强大的模型，那就很有意思了。 Um, you know, if, if we have a world where there's just, like, a small number of, of really capable models, um, I don't know, it just, it's, it's interesting, right? 7:39 拿网络安全来说，一方面我理解这种直觉："好，我们有了强大的网络模型。" I mean, if you look at some of the cyber stuff, for example, um, the instinct which-- I mean, on the one hand, I, I can understand the instinct of like, "All right, we have this capable cyber model. 7:47 只把它给前一百家机构。但问题是，世界上重要的机构不止一百家。 Let's release it so that only, you know, whatever it is, the top hundred institutions get it." But, you know, I think part of the issue is that there's more than a hundred important institutions in the world. 7:56 看看 Hugging Face 那次事件：它可能不在全球前一百，但它很重要。 So, you know, if you look at things like the Hugging Face incident that happened, you know, Hugging Face is maybe not one of the biggest hundred institutions in the world, but it matters. 8:04 对吧？这是人们依赖的重要东西。所以当他们察觉有入侵时，就转向了开源模型，因为他们用不上—— Right? It's like an important thing that people rely on, and so what did they do when they started detecting that there was this intrusion as they turned to open source models because they didn't have access to- 8:14 那些出问题的闭源模型。所以，我觉得—— some of the closed ones that were causing the issues. So, um, I think- 8:19 健全的开源生态，是拥有安全稳定未来的重要一环。 that having a robust open source ecosystem is one important part of having kind of a safe and stable future. 8:26 但对我来说，最重要的是让技术广泛分布，而不是囤积在少数人手里。 But to me, the, the most important thing is just making sure that we distribute the technology widely rather than hoarding it in a small number of people's hands. 8:34 硅谷也一直在争论，为什么大家对 AI 这么反感。 And I think there's also this ongoing debate in Silicon Valley about why people feel so negatively about AI. 8:41 你在最近的长文里谈到、或者说试图回应这些担忧，但大家对 AI，尤其是数据中心，情绪非常负面。 I mean, you talk about this in your, in your recent letter, uh, addressing these concerns or trying to, but the, I mean, the sentiment on AI and data centers in particular, it's so negative. 8:51 你论点的核心是不是——如果错了请纠正——如果让更多人能用上现在被锁起来的—— And it sounds like maybe an essence of your argument, correct me if I'm wrong, is if we diffuse this technology more, if we enable more people to access the things that are right now gated- 9:01 被顶尖实验室锁住的技术，是不是就能解决这个问题……人们是不是觉得被 AI 边缘化了？ by some of the top labs, maybe that addresses this kind of... I think don't people feel maybe disenfranchised by what's happening in AI? 9:09 这意思？ Is that what you're getting at? 9:10 对。这里面有很多层面。这也是为什么那篇文章那么长。 Yeah. Well, there are many layers to it. It's, I mean, I think you-- There's so many parts of... This is why the essay was so long. 9:17 对吧？写了 15 页，因为问题太多了，比如就业和经济。 Right? There are 15 pages because, I mean, we want to get through-- There are lots of different questions. I mean, people have questions about jobs and the economy. 9:23 还有数据中心对当地社区的经济和环境影响。 They have questions about data centers in their local communities and the kind of economic and environmental impacts of that. 9:29 还有 AI 被滥用的问题。 There are questions about how people might misuse AI. 9:31 还有网络安全问题、新出现的生物风险，以及如何维护自由社会。 Right? I mean, there's the cyber questions. There's bio risks that are coming up. There are questions about, you know, how we maintain a free society. 9:38 还有美国领导力的问题，以及技术越来越强之后如何保持掌控。 There are questions about American leadership. Um, there are questions about maintaining control over the technology as it gets to be increasingly capable. 9:45 这些都很重要。所以不能只泛泛而谈，因为每个问题都有各自的细微差别。 So these are all important. So it's, it's important not to, um, you know, just talk about this in generalities at the, at a high level because I think each of these has some different nuances. 9:56 但总体来说，它们有个共同点：要创造普惠的繁荣。而一个好办法，就是把握好谁有权使用技术的平衡，确保最大的那份权力交到普通大众手中—— But in general, I think s- one of the things that they all have in common is that if you create broad-based prosperity, um, and I think that one of the better ways to do that is by ensuring that there's the right balance of power around who has access to the technology and generally making sure that the greatest balance goes towards the kind of general population of people- 10:15 而不是什么内部人士、利益相关者之类。我认为这一点最终非常关键。 as opposed to kind of any kind of, I don't know, insider stakeholder or whatever you wanna call it. Um, I think that that ends up being very important. 10:23 再看数据中心。我们发现，像 Meta 这样的公司进入一个社区、承诺投资几十年——建数据中心本质就是这样——我们就能让它对社区非常有利。 So if you look at the data centers, I, I-- What, what we've actually found is that when a company like Meta goes into a community and makes a commitment that we're gonna invest there for decades, right, which is really what we're doing when we're building up a data center, we're able to make it so that it's very good for the community. 10:41 比如它带来的税收。在路易斯安那就有个例子：税收给当地老师发了 5 万美元的奖金。 Um, I mean, the, the, the kind of the tax revenue that they bring, uh, from that. I mean, in, in Louisiana, we have this example where, like, the tax revenue funded these $50,000 bonuses for teachers in the community. 10:52 我们带来很多就业，也在当地大量投资，我觉得这是好事。 We, we bring a lot of jobs. We, we invest in the, in the local community a lot. Um, that I think can be good. 10:58 但也有不少投机行为：有些公司根本没打算运营数据中心几十年。 I think that there's, there's also a lot of speculation, right, where there are companies that aren't necessarily planning on running a data center for decades. 11:05 他们只想找块地，然后卖给大实验室，根本不在乎当地社区，也不打算长期投入。 They're just, you know, trying to find a plot and, um, and then trying to sell it to one of the big labs, and they don't really care as much about the local community and, um, and they're not invested for the long term. 11:16 如果他们不愿为当地社区着想，人们当然会不满。 So if they don't care to, you know, focus on making it work for the local community, then of course people are gonna get upset. 11:22 所以我觉得，这种投机热潮会带来麻烦——我甚至不愿叫它泡沫，因为那意味着估值过高。 So I think that that's one of the things that can be, um, difficult when you have these kind of speculative, um, I don't, I don't even know if I'd call it a bubble 'cause that implies that it's overvalued or something. 11:35 但繁荣肯定是有的。 But, but there's certainly a boom, right? 11:37 于是就出现了一些短视的激励，导致—— So it's, so, so you have that. I mean, that leads to some of these, like, short-term thinking incentives that- 11:43 未必能惠及所有利益相关方，而我认为这才是长期可持续的关键。 don't necessarily lead towards helping every stakeholder, which I think is kind of what you need to do to make this sustainable over the long term. 11:50 说到底，如果这项技术不能创造就业、不能带来普惠繁荣，我们建的基础设施帮不到当地社区，那它就不会被允许继续下去。 And at the end of the day, I mean, if we're creating a technology that, like, if it doesn't create jobs or doesn't create broad-based prosperity or the infrastructure that we're building doesn't help local communities, that's not gonna be allowed to continue. 12:02 所以它必须做到。你必须把它设计成能在这些方面都帮到人们。这也是为什么，对那些…… So it's, like, it has to. Of course, you have to design it in a way that, um, can be helpful to people in all these ways, which is also part of the reason why, like, for people who are ... 12:11 对整个事情怀疑或者悲观，我的基本看法是：如果我们没法用积极的方式把它做出来，它实际上就根本不可能发生。 you know, skeptical or have so much doom about the whole thing, I'm, I mean, one of the, the things that I basically think is that if we don't end up building it in a way that's positive, it just e- effectively won't be able to happen. 12:24 所以我认为，建立正确的制衡、广泛分享红利，是以我认为长期对社会最有利的方式去扩展规模的前提。 Um, so I, I kind of think, like, the- actually establishing the right checks and balances, um, and distributing the, the benefits of this widely is sort of a precondition for being able to scale in the way that I think, um, would be best for society over time. 12:39 我没听过其他科技CEO像你这样谈数据中心，把它当作长期投资。 I haven't heard another tech CEO in your position talk about data centers that way, like the long-term investment of it. 12:46 这是你一直以来的想法吗？还是说最近你对此有了更清晰的认识？ Is this something-- Is this how you've always thought about it? Is this something you feel like there's more clarity that's been brought to it for you recently? 12:53 你一直—— Like, have you always- 12:54 对，一直有反对数据中心的情绪—— Well- Yeah. I mean, I think that there's been all this anti data center sentiment- 12:57 就是你说到的那种。所以我们深入研究了，因为我们想搞清楚：我们的项目周围没那么多这种情绪，这是为什么？ that you're talking about, so we've dug into it because what we're trying to understand is, okay, like, there isn't as much of that around our project, so why is that? 13:07 于是我们问了很多人，发现投机者和着眼长期的公司之间存在很大分化，仔细想想这也说得通。 And then, so we ask a bunch of people, and it's like, well, it looks like there's a pretty big dichotomy between these speculators and the companies that are focused on it for the long term, and that kinda makes sense when you think about it. 13:18 我们搞了个美国劳动力学院—— So I mean, one, one thing that we're doing, I mean, there's this America's Workforce Academy- 13:22 项目。简单来说，我们要建这么多数据中心，而且会建很久，但市场上没有足够的技术工人来支撑，所以我们需要更多光纤技师、电工， project that we, we did, which is basically, okay, we're gonna build all these data centers. We're gonna be doing this for a while. There isn't the volume of skilled tradespeople that you need to create this, so we need more, like, fiber technicians and electricians and, like, 13:36 高级木工等等，现在干这些活的人不够。我们需要几十万、甚至上百万这样的人，但目前没人受过这种培训。 advanced carpentry and, and all of these things, and there aren't enough people to do this. We need hundreds of thousands, you know, maybe, maybe millions of more people who can do this, and, um, people aren't trained to do that right now. 13:48 所以我们设立了这个培训项目：凡是完成培训的人，都能保证得到一份工作，在参与建设 So we created this training program, um, to, to effectively do that, where we guarantee people who get through the training program a job at, um, at, at a place that is working on building- 14:01 Meta基础设施的地方。为什么这么做？这可不是慈善。 infrastructure for Meta. And why did we do this? I mean, it's like, it's not really philanthropy, right? 14:06 我们需要这些人掌握那些技能，所以这就是双赢。 It's like we need those people to be skilled and have those skills. So it's just, it's a win-win. 14:12 如果你打算干几十年，这种投资是合理的；但如果你只想建一个站点，然后转手卖给 It's an investment that I think makes sense if you're in it for decades, um, but not necessarily something that you would do if you were building out one site with the intent of flipping it to- 14:21 另一家公司。 a different company. 14:23 所以我认为，只要你着眼长远，世界上很多问题都会通过激励一致自然解决。 So I think that a lot of problems in the world do just naturally get solved by incentive alignment when you think about them over the long term. 14:32 所以我认为这是很重要的一环。我的想法是：绝不可能允许少数几家实验室控制如此重要而强大的技术，并为自己积累巨额财富。 Um, so I think that that ends up being an important part of this. And I guess part of the way that I think about this is that I just think that there's no way that it's gonna be kinda permitted for there to be a small number of labs that control such a important, um, and capable technology and accumulate a lot of wealth to themselves. 14:49 这必须惠及大众—— I think it's like this has to be a broad-based thing- 14:52 才能行得通。它在技术上要成立，在社会层面也得成立。 in order for it to kinda be able to work. It has to work technologically, but it also has to work, um, kind of socially. 14:58 我认为这两方面必须齐头并进。 And, and I think that those, those pieces kind of have to, have to go hand in hand. 15:02 我同意。好，现在我们快到Muse了。在那之前，一年前你还写过个人超级智能—— I agree. Well, now we're landing towards Muse. Before we get there, you also wrote, a year ago, your personal superintelligence- 15:10 文章短版。 essay. Shorter one. 15:12 就一页。 that was a page. 15:13 我写过一页的版本 But, yeah, um- I did write a one-page version 15:14 一页版——哦 of the Futurist- Oh, 15:15 是吗？ you did? 15:15 也写了。 ... one too. 15:16 发在《华尔街日报》 Yeah, well, I published it in The Wall Street Journal, as, as an op-ed. 15:18 对对对。 Oh, that's right. That's right. That's right. 15:19 我刚读了长版 That was-- So that was kind of- I just read the long 15:20 长版，好 one. Okay. 15:21 呃，那个 Um- Well, 15:22 有一页版，好 Uh- So there's a one-page version- Okay ... 15:23 还有15页版 and then there's the 15-page version. 15:24 你一年前写的那篇《个人超级智能》，很多人看了都想，马克为什么写这个？ Well, so this one you did about a year ago, Personal Superintelligence, I think a lot of people in my world, when they saw you write that, was like, "Oh, wow, why is Mark writing this?" 15:32 他看到了什么？我猜，说得不对请纠正，可能就是Muse—— Like, what is the thing he's seeing on the other side of this? And I think it, correct me if I'm wrong, it, it might be Muse- 15:39 要聊的对吧？ like what we're gonna talk about, right? 15:40 对，没错 Yes. That's right. 15:40 刚发布的。 What you guys are releasing now. 15:41 就是它。 Yes. This is it. 15:42 你是怎么意识到这就是Meta的下一篇章的？ Um, this is it. So how did you come to that realization that that's what-- this is the next chapter for Meta? 15:48 有意思。我们从来没把自己只当成社交媒体公司。 It's interesting. You know, we've never really just thought about ourselves as a social media company. 15:54 我们一直认为自己是一家连接人与人、赋能个人的公司。公司头15到20年里，那些驱动我们创造的价值观，就是把技术和力量交到个人手中， We've definitely thought about ourselves as a company about connecting people and about empowering people, but I think a lot of the values that led us to build the things that we built for the first, you know, 15, 20 years of the company around putting technology and power in individuals' hands, 16:09 相信人们应该能自己决定生活中什么重要。 um, believing that people should be able to decide for themselves what is important in their lives. 16:15 我们经历过很多社会辩论，比如内容审核之争，核心都是：人们能不能—— And we've gone through a lot of social debates around this, right? A lot of the, the, the debates around content moderation and things like this have kind of been around this question of, like, should people be allowed to- 16:24 自己决定并表达自己生活中什么重要？我觉得这些经历让我更坚信，历史上、尤其是技术时代，大量进步都来自赋能个人， kind of decide and communicate for themselves what matters in their own life? And I think through that experience, it has sort of sharpened my belief that a lot of progress throughout history and through this technological age comes from empowering individuals and 16:44 人们其实最清楚自己生活中什么重要。所以每当听到有人说，应该让少数专家来决定AI去做什么大事，我就很反感， that people really do know best about what matters in their own lives. So because of that, I have somewhat of a, of an allergy whenever I hear people talk about, oh, like, we should just have a small number of experts allocate what AI does to, 17:00 为什么不做人们在乎的事？人们在意的本来就很多样。 like, big problems. Why should it do, like, like, these things that people care about in their lives? It's like, well, people have a balance of things they care about. 17:06 人们关心健康，关心过得更好，也关心感情，关心陪伴家人朋友，还关心文化。 People care about health. They care about having a better life, but they also care about their relationships and, like, showing up for their friends and family and, um, they care about culture. 17:16 人们关心的事，在科学家或工程师眼里可能算不上最大的难题。 Um, you know, people care about things that may not, to, you know, a scientist or an engineer in the industry, feel like the biggest problems. 17:24 但如果你去问几十亿人他们关心什么，我认为这些答案的总和，就是最应该去做的事。 But I don't know, if you ask, like, billions of people what they care about, if, like, you know, I think what people's aggregate answers to that question is kind of is what the most important things are to be worked on. 17:36 所以我一直相信，当你打造这种超级智能时， And, um, so I've always just kind of believed that when you build this superintelligence, 17:43 问题在于谁来决定它做什么。我认为人们应该能指引它去做对自己重要的事。 there's this question of who decides what it's gonna focus on, and I think that people should be able to direct it towards what matters in their own lives. 17:52 而不该由少数实验室里那些所谓的专家说了算。 It shouldn't just be directed by, you know, some, like, so-called experts, um, sitting at a small number of labs. 17:58 这就回到了整体理念的基础：要有一个积极的未来，就要赋能人们，把技术交到他们手中，让他们自己决定什么重要、想怎么用它。 So this gets back to the-- it's like it's-- this is basically the foundation of this overall philosophy, which is that the way to have a positive future is to empower people, to put the technology in their hands, and let people decide for themselves what matters and how they wanna use it. 18:14 我认为当人们这样做时，首先会优先去做一些…… And I think that when people do that, it, first of all, will prioritize some- ... 18:19 不一样的事，对吧？比如可能优先处理健康问题，但也许不是最常见的那些病——那大体上是如今制药和生物科技行业的优先级， things that, that are different, right? It's, um, you know, like maybe it'll prioritize, um, health issues, but maybe instead of prioritizing, you know, it's like the most common things, which is kind of what the, the kind of pharma and biotech industry, uh, 18:33 毕竟还有很长的尾部，很多罕见病和病症。 like at, at large prioritizes today. You know, it's like there's a very long tail of rare diseases and conditions that people have. 18:39 如果你得了罕见病，你多半希望你的个人AI专注于此，而不是最常见的东西。 So, you know, okay, if you're, if you have, if you have a rare condition, like you're probably gonna want your personal AI to focus on that, not like just s-something else just because it happens to be the most common thing. 18:48 所以我觉得，像罕见病这类领域，投入就明显不足。 So I think, for example, rare diseases I think are under, um, disproportionately under-invested in. 18:54 你们正通过基金会在 Um- You're doing a lot of investment with 18:56 我们在Biohub做。 your foundation on- We're doing that at Biohub. Yeah. 18:57 但这也部分塑造了我的看法—— But, but I, I mean, and that's partially informed some of my views here- 19:01 就是要把权力交到个人手中，让他们决定什么对自己重要。 that I think like, yeah, like y-you wanna put the power in individuals' hands, um, to determine what matters for them. 19:07 但这些未必是人们口中的重大社会问题。 But a lot of this is also like it's not necessarily the things that people would say are like the big social problems. 19:14 就拿我来说，我用Muse智能体，是想让它帮我做个更好的父亲、更好的丈夫，更好地陪伴朋友，帮我与人建立联系。 Like, a lot of it for me, you know, when I'm using my Muse agent, I just, you know, I kind of want it, it to help me be like a better father and, you know, a better husband and show up better for my friends and, you know, be able to, um, help me connect with people. 19:29 我觉得这也是贯穿Meta至今工作的一条主线。我们这家公司格外相信，帮人们与身边的人建立联系是有社会价值的。 And that I think is also partially a through line between the work that we've done at Meta so far and this is, I think we're the company that I think just disproportionately cares about and believes that there's social value in, um, in helping people connect with the people around them. 19:44 这也是我最早让我的智能体做的事之一。我三岁的女儿喜欢烘焙。 So I know it's like one of the first things that I, that I kind of set up my, my own agent to do. It's like, all right, my three-year-old daughter likes baking. 19:51 我不懂烘焙，但这是能做的有趣项目，于是我就说： I don't know anything about baking, but, like, that's like a, a fun project that we can do. So I basically ask them like, "All right. 19:57 “安排一下，每个周末准备一个烘焙项目，适合三岁小孩和一个完全不懂烘焙的大人，再用Instacart之类的把食材都买好。 Set up so that every weekend, um, you know, like we have a baking project that is, um, that is, uh, kind of reasonable for a three-year-old and an adult who knows nothing about baking, and, um, and use Instacart or whatever to go get all the ingredients. 20:15 让它自己琢磨怎么安排，把一切准备好，这样周日我和女儿就能直接动手做了。” Like just figure out what makes sense and make sure everything is ready, so that way, like when I show up on, you know, on Sunday with, uh, with my daughter, we can like go make this thing." 20:25 事后我告诉它：“那个太难了。”结果棒棒糖蛋糕太难了。 And, um, and then I tell it afterwards. I'm like, "How, how did it go? Okay, that one was too hard." Turns out cake pops, really difficult. 20:30 难得出奇。 Um, surprisingly difficult. 20:32 哦？ Huh. 20:32 烘焙我一无所知， Um, there's a lot of things- I know nothing about baking, 20:34 真的。 so. 20:34 我也不懂，现在懂了。 Yeah, I didn't either. I, I know, I know something now. 20:36 蛋糕棒。 Cake pops. 20:37 别从它开始。 Cake p- No, no, don't start with cake pops. 20:39 问题就出在这。 That's, that's the problem. 20:40 烘焙里很多东西很简单。 Um, yeah. No, there's a lot of things in baking that are pretty simple. 20:44 棒棒糖蛋糕不算。 It turns out cake pops is not one of them. 20:46 不知道。 I did not know that. 20:46 唉，我能说什么呢？ But, um, yeah. Well- Oof ... you know? What can, what can I tell you? 20:49 谢谢， Um- Thank you, 20:49 Muse。 Muse. 20:50 对，谢谢。这样它就会更新调整，帮着改进。 Yeah, yeah. Thanks. Um, so yeah, and it, it kinda like, so it kind of updates that, um, and, and helps with that. 20:57 我大女儿迷上了登山，有些山需要许可证。 Um, you know, my, you know, my older daughter has kinda gotten into climbing mountains, and some of them you need permits for. 21:03 我让它盯着，一开放就抢许可证—— So I have it basically like sit and get the permits when they become available- 21:07 这样我们就能去爬山了，这挺不错的。弄完之后，它就告诉我：好了，我订到了这一天的登山许可。 so we can climb mountains. And, um, that's just, like pretty neat. Okay, so then, like, we do that, and it basically tells me, it's like, "All right, I was able to get a permit for this day." 21:16 行吧，请一天假陪女儿爬山。 And I was like, "All right. Well, I guess I'm taking that day off from work to go climb a mountain with my daughter." 21:19 总之挺酷的。 So it's, it's kinda cool. 21:21 对吧？它不光干这个，还能帮我保持健康，辅助我的训练。 Right? It's, so it does that, you know, but it also helps keep me healthy, helps me with my training. 21:27 我在我的综合格斗馆里装了摄像头，让它盯着画面给我反馈，挺好玩的。 Um, you know, I put cameras up in my, in my MMA gym, and I tell it to watch the cameras and send me feedback, and it's like, it's pretty fun. 21:35 挺好的。 You know, it's good. 21:36 反馈靠谱吗？ And it's good feedback? 21:37 有时靠谱 Um, sometimes. 21:37 有时候它的反馈还挺逗的。它会说我：看起来你当时真是彻底放弃了。 You know, it's, I mean, it's, um, sometimes it's funny feedback. It like, it like finds me in... It's just like, "It looks like you really gave up." 21:45 我承认当时太累放弃了，心想干嘛挑这个说？但总体不错。 And, and I was like- ... "Yeah, I did. I was really tired right there." It's like, "Why is that the thing that you're pointing out to me?" Um, but no, it's good. 21:51 也挺搞笑的，教练们还拿这事开玩笑—— Um, and it's funny. M- Like, the coaches, like laugh about it- 21:54 因为他们说：这— 'cause they're like, "Yeah, no, this is like, it's-" This 21:56 我们不好意思说的，智能体替你说。 is what we didn't feel like we could tell you, but your agent's telling you. 21:59 对，是的，挺好，真的挺好。 Yeah. Well, it's, um, yeah. No, it's good. It's good. 22:01 听你聊这个我才想到，你明明有一大帮人能帮你做这些事，那你是怎么用 Muse 智能体去测试它作为助手的能力极限的呢？ I didn't think about this until hearing you talk about it, but you, you have people that could obviously do all this for you, but, like, how do you use something like a Muse agent to, like, really test the limits of, like, how it can be helpful as an assistant, right? 22:15 你有没有把它逼到极限，让团队一看就说：好吧，这个我们得修？这样的例子肯定不少—— Like, like are you, were you pushing it, like how have you pushed it in a way that the team is like, "Oh, okay, we gotta, gotta f-fix this," or I'm sure there are many examples, but- 22:23 我觉得有意思的是，每个人想用它做的事都完全不一样。 I mean, I think part of what's interesting about it is that everyone ha- just has, like such different things that they wanna do with it. 22:30 早期测试时，我们把它发给了一批人。有个人拿到手一天之内，就开始用它帮忙管理自己在家给孩子上的课。 So, like in the early beta period, we handed it to a bunch of people, and like, it's like I gave it to someone, and like within a day, they're like using it to help run their home school. 22:41 我就想，哇，才一天你就上手了。还有个人十二小时之内就说：我刚用它规划了一趟旅行。 And then like, it's like, okay, wow, you just like started this within a day. And then like another person within a day or within 12 hours, they're like, "I just planned a trip." 22:49 它就把整个行程都安排好了。还有个我认识的人，对科技一向很怀疑，我把 Muse 给了她，结果她好几天都没吭声。 Um, yeah, and it's like, it, it just like planned this whole thing for me. Yeah, I mean, someone else I know who's like generally pretty skeptical about technology, I, I gave it to her, and then she was, um, she didn't say anything for like a few days. 23:00 后来她给我发信息问：等正式发布的时候，我这个 Muse 智能体还能留着吗？ Then she texted me and was like, "So when you do like the general release, do I get to keep my Muse agent?" 23:05 还是要重置？嗯，不错。 "Or are you gonna reset it?" I was like, "All right. This is good." 23:07 是啊，我觉得这东西真管用。 This is, yeah. So I think, I think it's like, I think this is working well. Um- 23:10 我认识的所有参加测试的人，对它评价都非常高。 Everyone I know who's been on the beta has very, you know, high things to say about it, high praise. Um- 23:15 用法各不同。 Yeah. But people do different things with it. 23:16 我觉得我们该给大家把话说得更明白些，因为很多人以为 AI 就是 Meta AI 或 ChatGPT 那样， Yeah. A- and I think we should just also say what it is more plainly for people so they understand because I think people think of AI as like, you know, Meta AI or ChatGPT. 23:25 一来一回地对话。 It's like back and forth prompting. 23:27 而真正的突破在于——整个行业都在往这个方向走，不管是 Grok Bot、Town 还是 Instinct， The, the real unlock here, and this is happening in the industry more broadly, whether it's Grok Bot, Town, Instinct. 23:34 这样的产品不少，但核心是它在背后加了一台虚拟机—— I mean, there's many products doing this, but it's, it's adding a virtual machine behind the scenes- 23:39 让智能体能替你操控电脑、登录账号、实际办事。对那些只用过聊天式 AI 的人来说，这是个巨大的变化。 where like the agent can control a computer for you and log in and do things, and that's a, that's a huge change, I think, for people who only know AI from the- 23:47 对，很持久。 Yeah. And it's, it's long-lived. 23:49 所以它不像 Meta AI、ChatGPT 或 Gemini 那样，你发一条提示，它就给个答案—— So it's, um, so basically instead of the model with like Meta AI or ChatGPT or Gemini or whatever you use, where you send one prompt and then it gives an answer- 24:00 而是交给它项目，或者交给它目标。 in this case, what you basically do is you give it projects or you give it goals. 24:05 然后它全天候不停地工作，不达目标不罢休。 And then it just works, and it works 24/7, and it, like, doesn't stop until it's helped achieve the goals. 24:10 你们管这叫“通宵学习”。 Your team was telling me it studies overnight is what you guys call it. 24:13 它会学习，把反思沉淀成记忆。它一直在做项目，还能主动提出新项目。 It studies. It, like it kind of consolidates its reflections into, into memory. It, it basically just works on projects, and it can also suggest new projects. 24:23 我和一个女儿玩《文明》游戏，就问它：“要不要给她做一份攻略？” So yeah, I mean, I was like I like play the computer game Civilization with one of my daughters, and I like, I was like, "Hey, do you wanna make like a strategy guide for her?" 24:32 我说好啊。然后它说：“既然有了攻略，要不要我扩展一下，顺便讲讲不同文明的历史？” And I was like, "Yeah, sure." So then I was like, "Okay, um, now that we have the strategy guide, do you want me to like expand the strategy guide so it can also teach historical lessons about different civilizations?" 24:43 我说当然。它就在做好的应用里加了个新标签页，特别酷。 I was like, "Yeah, sure. Why not?" So it just, so just kinda like built a new tab in, in the app that it made, and um, that wa- that was, that was very cool. 24:50 它还能不断扩展—— So it can kind of just like expand and, and- 24:52 很主动。 It's proactive. 24:53 对，它会主动提建议。我觉得很有意思的一点是，它能帮人赚钱，也能帮人省钱。 Yeah, it suggests things. And one of the things that I think is interesting is that it, um, I think is just going to be able to like make people money and save people money. 25:02 真的？ You think? 25:02 对，有意思的是我们定价背后的经济模式。 Yeah, I mean, part of the, part of what's interesting here is the economic model for how we're pricing it. 25:07 你可以付费订阅，如果你想要那种模式的话。 I mean, you, you can pay for a subscription if you wanna, if you basically kind of wanna have that, that model. 25:13 但我们也让你能免费获得很大的用量。 But we're also just making it so that you can get a very large amount of usage for free. 25:18 起步每周免费送 1 亿 token，还配一台虚拟机。 I think we're, we're offering something I think to start, it's like 100 million tokens a week for, for free, like, a- and, and you get this virtual machine. 25:26 算力相当可观。 So it's like a, a lot of kind of computer. 25:28 好梗算力大 That, that's a good meme. It's a lot of computer. 25:30 算力就是大。 It's a lot of computer. 25:31 我们这么做的原因是，我们相信：对那些用它经营小生意、赚钱或做交易的人来说—— But the reason for why we're doing this is we basically are confident in standing behind the fact that we think that this is going to, uh, effectively for people who are going to use it for running a small business or making money or transactions- 25:45 或者做任何商务，我们觉得它能帮他们赚很多钱，所以长期的商业模式就是从交易中抽很小一笔—— or commerce in some way, we actually just think it's gonna make so much money for people that, um, that the business model over time that we expect is to effectively just take a very small cut of whatever the transaction is, um- 25:58 抽成生意，很小。 Take rate business, small, yeah. 25:59 而且不一定用户付钱—— Yeah, and not even necessarily the person paying for it- 26:02 来自合作商家。不过—— it'll come from the businesses- Right ... that they're working with. But, um- 26:04 用 Stripe 收款—— And you're working with Stripe on payments- 26:06 好，明白 and yeah. Okay. 26:07 我的看法是：这项服务应该对绝大多数人都免费。如果你想让未来人人都拥有强大的超级智能体，这一点至关重要。 Yeah, so but my view is like we should be able to have a service that you make free for the vast, vast majority of people, um, which again, is critical if you wanna build a, this future for everyone where everyone has, um, has these powerful super intelligence agents. 26:24 让所有人都能用上，重要的一点就是让大家负担得起。所以我们想让它免费，给你巨大的免费用量。我们敢打包票：它真的能帮你赚钱、帮你省钱，靠这个就能自我买单。 Um, I think an important part of making something available to everyone is making it affordable. So we wanna make it so that this is free, so there's this, just this huge amount of usage that you get, and we're basically just standing behind that and saying, "We think that this thing is actually going to make you money and save you money," and that is how you're g- it's going to pay for itself. 26:40 Meta 的服务能接进去吧？理论上你可以管理 Instagram 的广告投放—— And Meta's services can connect into it, right? So you could theoretically manage your ad spend on Instagram- 26:48 那些东西 all that stuff. 26:48 把它连上。 Well, you have to-- you connect it. You ch- Yeah. 26:50 你想连什么都可以，当然也能接 Meta 的服务。 you can connect it to whatever you want. You know, it does work with Meta's services if you want. 26:54 不想连也可以不连。比如你拿它做生意，它就能直接对接我们的广告系统。 I mean, you obviously don't have to connect it if you don't want to. So if, um, you're using it to, to run a business, it can basically just connect to our ad systems. 27:01 你可以让它帮你做东西，帮你做出产品，然后还能帮你运营生意。 I mean, you can ask it to, to make something for you. It can, it can kind of help you make the product, and then it can help run the business. 27:08 对，所有这些它能循环着做，全天候不停，一直干下去。 So- Yeah ... um, so all this stuff, and it can just do that in a loop and just do it forever for, you know, 24/7. 27:13 而且每次我们发布新模型，我们现在基本保持着—— So, and every time we release a new model, which, you know, we've been on this cadence of shipping- 27:19 差不多每个月都有重要更新—— a, you know, meaningful update like every month- 27:21 它就会越来越聪明，能力越来越强，能做的事越来越多。 it's just gonna get smarter or just, and get more capable and, and able to do more and more stuff. 27:26 你团队跟我提过一点，别处没见过这种做法，就是这个舰队概念——让整队 Muse 智能体一起学习—— And something your team was telling me that I haven't heard this approach used elsewhere is this fleet concept where you're letting the fleet of Muse agents learn- 27:34 一起。 together. 27:34 这就是点子和建议功能。 So that was the ideas and suggestions thing. 27:36 打开应用， So you open up the app. 27:37 主标签页就是你和 Muse 的聊天。还有一个点子标签页，根据你告诉它的事来拓展。就是我刚说的那个：先是它帮我做了陪女儿玩《文明》的攻略， The main tab is basically your chat with, um, with your Muse. There's a tab for basically ideas from the things that you've told it, how it can expand those, so that's the thing that I was saying, which is, you know, first it helped make the strategy guide for playing Civilization with my daughter, 27:54 再拓展成历史课，这主意是它自己想的。 then it helped expand that into historical lessons. That was-- It came up with that idea. 27:57 我说，行，做吧。 And then I was just like, "Yeah, sure, do it." 27:59 它总能找到各种方式自我增强。比如那个综合格斗教练，它会主动想办法让自己变得更好。 Um, and it, it, it finds all these ways to basically augment itself. I mean, the like MMA b- like coaching thing, it, it like comes up with ideas for how to make it better. 28:09 它会问：要不要我学着更准地挑出该发给你的画面？我说，好，去做吧。 It's like, "Would you like me to like get better at, um, finding the right frame to send to you?" It's like, "Yeah, go, go do that." 28:15 所以这个点子功能很重要，因为这样就能在整个舰队里找到对不同事感兴趣的人。 So yeah, the, the ideas thing I think is important because then you basically across the fleet can find people who are interested in different things. 28:23 退一步说，我觉得 AI 存在的一个大问题是，很多人不知道该拿它做什么。 I, I think, I, I, I guess taking a step back, one of the big issues that I think exists with AI is a lot of people don't know what to do with it. 28:29 所以如果智能体自己能主动建议可以帮你做什么—— So I think if, if the, if the agent can itself help you suggest things that it can do to be helpful for you- 28:36 那就解决了这个大问题的一大部分，让你能真正把它用出价值。 then that solves a huge part of this problem of making it so that, um, you can get the most out of it. 28:43 你们在给智能体引入网络效应式学习，这几乎没人做过—— When you're introducing like network effect learning for agents, which no one's really done, where- 28:49 也就是智能体从整个舰队里学习匿名化的经验—— basically the agents are learning anonymized insights from the rest of the fleet- 28:54 而你可是网络效应之王。 and I mean, you're like the king of network effects. 28:56 我对这个想法特别感兴趣—— Like I, I'm, I'm really interested in this idea- 28:58 因为我觉得没人在做这个。 'cause I, I don't think anyone's doing this. 29:00 对，我觉得现在行业里大多把智能体当成单人游戏—— Yeah, no, I think that right now, I think most of the industry is thinking about agents as like a single player game- 29:08 就是你有个智能体，自己用。但让智能体之间互动，能做出很多有意思的事。我们内部已经有很多有趣的例子，大家的智能体在互相协作。 where it's like you have your agent and, and you use it. And there are gonna be all these interesting things that basically you can do by having the agents interact with each other, and we already have all these interesting examples internally where, you know, people have their agents interacting with each other. 29:21 不过这次发布里大部分还没上线—— This isn't like, uh, m- for the most part rolling out in this release- 29:25 但这会是这件事长期运作的重要一环，就是—— but it's, uh, it's gonna be like an important part of how, um, I think this works over time is just- 29:31 用 Muse 的人越多，它就会越来越好。 like as more of the people who you know start using Muse, it just gets better forever. 29:36 越来越好。这会是差异化所在吗？前沿模型在持续商品化，产品形态和智能体框架也越来越像，这种学习带来的网络效应才是真正的优势？ Gets better. And is that the differentiator as, you know, you could buy the idea models continue to like commodify at the frontier essentially, or, you know, the products all start to look similar, similar kinds of harnesses, is the network effects of that learning the real edge? 29:50 我觉得我们有几处独特的地方。一是从零开始设计模型—— Well, I think that there's a few things that are, that are kind of unique that we're doing. One is we're designing the models from the ground up- 29:58 专门针对这个使用场景，这点很重要。二是我们这家公司确实有社交基因。 to, uh, to basically be good for this use case, which I think really matters. Two is, um, is basically I do think we have this social DNA as a company. 30:07 我们帮助人们用 AI 和智能体去增进、巩固人际关系，从生活中那些柔软却很重要的部分获得更多。 We're helping people use the AI and agent to like enhance their relationships and strengthen your relationships and get more out of like kind of the, the soft but very important parts of your life. 30:19 这方面我们可能比任何其他实验室都更上心。 I think that that's something that we're probably just gonna be more attentive to as a company than any of the other labs. 30:25 我要说的第三点会成为我们的重要差异点，可能让一些人意外，那就是隐私和安全。 The third thing That I would say is actually going to be a major differentiator for us that I, I think, um, it might be surprising to some people is privacy and security. 30:36 我们在这方面投入巨大。我们的看法是：它要有用，光有顶尖智能不够，还得真正了解你。 Mm-hmm. And we're investing in this just a huge, huge amount. And, you know, m- part of the view that we have on this is that in order for this to be useful, it needs to not just have state-of-the-art intelligence, it needs to really understand you, right? 30:50 要理解目标，就得接入各种东西：广告系统，还有消息、邮件，全都接。 So in order to be able to understand your goals, so you end up connecting it to all this stuff, right? You, you talked about, you know, connecting it to your, your ad system, but people connect it to messaging and- Email, all that, yeah ... 30:59 邮件、健康信息等等。要做到这一点，用户必须对系统有非常高的信任。 email and all this, all this stuff. Um, health information, whatever. And, um, in order to do that, people need to have a very high degree of confidence in the system. 31:10 好消息是，Meta 到现在花了十多年打造 WhatsApp，我认为它是最大的—— Yeah. Now, the good news here is that, you know, Meta has spent, at this point, more than 10 years focusing on building WhatsApp into, I think, like, it is the largest, like, 31:25 全球端到端加密系统，设计成连 Meta 都看不到用户发的消息，这带来了真正的变革。 global end-to-end encrypted system, and we've designed it in a way where even Meta can't see the messages that people send, and that's been this just really transformative thing. 31:37 这让人们信任 WhatsApp，也给 Meta 上了重要一课：这对 WhatsApp 的成功至关重要。 I think it makes it so that people trust WhatsApp. It's also been a very important lesson for Meta to learn that, like, that has been really important to our success with WhatsApp. 31:45 我们设计的系统连自己都看不到内容。这样不管人们担心什么，担心政府拿到数据、黑客入侵—— Mm-hmm. That we've designed the systems that even we can't see the content. That means that whatever people are worried about, if they're worried about, um, you know, government getting access to it, a hacker getting access to it- Right ... 31:56 还是 Meta 内部有人作恶，只要把系统设计成看不见，这些担忧就都能排除。 someone at Meta doing something bad with it that they don't want, um, all that stuff you can kinda take off the table if you design the system so that you can't see it. 32:03 所以起步时我们把它当作基础经验。Nat 和我还亲自请来了 Moxie Marlinspike。 So we took that as one of the foundational lessons when we were getting started with this. Nat and I actually, you know, personally recruited Moxie Marlinspike. 32:12 Signal 创始人，也是当年帮我们做 WhatsApp 端到端加密的人之一—— The founder of Signal. Yeah. Yeah. And the, one of the person-- one of the people who helped us build the WhatsApp end-to-end encryption- The encryption standard, yeah ... 32:18 那是 2014 年的事。他加入专门负责这个机密虚拟机项目，让你的虚拟机装着 Muse 里的所有信息，而我们能承诺连 Meta—— back in the day and- Yeah ... back in twenty fourteen. He joined to specifically work on this confidential VM project, which makes it so that we can-- you can have your virtual machine and have all this information in your Muse, and we can make the commitment that even Meta- Mm ... 32:35 都看不到里面的内容。这在技术上做得到，这个承诺可以经得起技术验证。 cannot see the content that is in there. And you can do this technically, like- It's, it's a, it's a- Yeah ... like an incredible kind of- Like the commitment can be technically verified and- Yeah, yeah ... 32:42 未来几周，随着更大规模推出临近，我们会公布更多细节——在各实验室这些早期的智能体虚拟机尝试里，你觉得你们的做法是独一无二的吗？ yeah. So, so that-- and that's something that we're gonna publish more about in the coming weeks, um, as, as we, as we basically get cl- get closer to rolling this out a lot more widely, um- And of out of all these early VM efforts that these labs are doing with these agents, you think this is unique, 32:57 我觉得没人在做，真的没人在做。我们还有很多其他安全措施值得聊聊，因为即便这功能还没就绪，即便有人不想用它—— what you're doing? You-- It's not- Oh, I don't think anyone is doing it, really I don't think anyone's doing it. I mean- Yeah ... I think that there's-- I mean, there's a lot of other security measures that we're putting in place that I think are-- we should talk through, um, because, I mean, you know, even before this is ready, like there's, there's that, and even people who don't wanna use this, 33:11 也极其安全，因为我们从一开始就重视这点。另外还有自动批准，你得看着它做什么—— there's, there-- it's like incredibly secure 'cause we focused on this from the beginning. Mm-hmm. But, um- There's also like the auto approve, like you have to see what it's doing and- Yeah. 33:19 这些我们待会儿细聊。但我不知道有谁做出过哪怕接近机密虚拟机的东西—— So, so let's get into all that stuff in a second. Yeah. But the-- But I'm not aware of anyone having anything close to the confidential VM- Yeah ... 33:27 也就是 Muse 的那套系统。这非常根本，因为你想确定这是你自己的智能体。 system that, that Muse has. And it's just-- I think it's a very fundamental thing because you're, you're, like you wanna know that this is your agent. 33:36 你把内容放进去，就可以放心，不会有别人拿到它。 Yeah. And that, that if you put content in there, that, um, that basically you can trust that no one else is gonna get access to it. 33:43 那有哪两种做法？今年早些时候 OpenClaw 出来后，很多人开始买 Mac Studio—— Yeah. So what are the two ways to do it? Well, a lot of people earlier in the year, when stuff like OpenClaw came out, they started getting Mac Studios and- Yeah ... 33:50 一种安心的做法，是设备就实体放在你家里运行。 you know, one way to feel good about it is where you literally, like you physically have your device running in your home. 33:56 另一种是搭建一套——但这很难，不会有几十亿人买台 Mac Studio 在家配置运行。 But the other way to do it is you build a, a kind of-- Uh, that's gonna be tricky 'cause there, there-- I don't think there are gonna be billions of people who are gonna buy a Mac Studio and configure it and run it at their home, you know? 34:04 尤其现在内存这么贵，而且技术上也难。我们做 Muse 的一部分目标，就是打造开箱即用的个人智能体体验，让家里各种技术水平的人都能用。 Especially with RAM prices right now and yeah, I know. So, but, but it's also just, it's like technically difficult. Yeah. Right? Yeah. It's like w- um, um, I mean, part of what we're trying to do with Muse is build a version of that personal agent experience that just works, that I can like give to everyone in my family of various levels of technical literacy. 34:19 对，而且就是能用，一天之内就能帮他们搞定生活里各种想做的事。 Yeah. And like it just works, and within a day, it's like doing kind of all the stuff that they, that they want in their life. 34:25 一部分原因是你不想让用户自己去配置电脑或虚拟机，你只想能在云端一键开通，但安全性和保密性要跟那台机器就摆在你家里桌子底下一样。 So part of that is like you don't want someone to have to set up their own computer or VM. You just kind of like wanna be able to provision it in the cloud, but you want it to have the security and confidentiality that you'd have if you had the box sitting under your desk, like in your house. 34:40 对，我觉得这是很根本的一点。当然，像你说的，还有别的方面，因为不是每个人都会用到这个。 Sure. Um, so I think that's a very fundamental thing. Um, like you said, there, there's other pieces too, because I mean, not everyone is gonna use that. 34:48 我们做了安全的凭据存储，对吧？你没理由把这些东西明晃晃地存着——对…… They-- I mean, we built the secure credential store, right? Where you're-- there's no reason for you to just like store out in the open- Yeah ... 34:55 像信用卡、密码——它会生成一次性卡号，所以智能体不该知道。 like all your, your kind of, your credit card and your passwords to things and- It creates one-time card numbers and all that, yeah ... yeah. So, so your agent like shouldn't know that stuff. 35:02 它只该在需要时能访问，因为你让它登录某个东西，除此之外都不行。 It should just be able to kind of access it, um, when it needs to because you've asked it to log into a thing and, and no oth- and n-n-not otherwise. 35:11 嗯。其实这不止是单一一个东西。你有核心的智能体，但我们也做了一批哨兵智能体，专门监控你的智能体收发的流量和数据，目的是提醒你—— Mm-hmm. Um, it's actually not a single thing. Um, you have your kind of core agent. Mm-hmm. But we also built all these sentinel agents that basically monitor the incoming and outgoing traffic and data that your, that your agent is sending for the purpose of flagging to you- Hmm ... 35:28 什么时候可能需要人工审查一下。对，哨兵的作用其实就是检查有没有人在搞提示词注入攻击。 um, when you might want to review something. Right. So that's-- A-all that the sentinels do is, is effectively they, they kind of look at, they, they try to see if someone's trying to do like a prompt injection. 35:39 嗯，还要看你的 Muse 智能体是不是把某些你可能不太愿意外传的东西发出去了。 Mm-hmm. They look at, okay, did your Muse agent share something that, that is kind of going out that, um, is not something that you might be comfortable with? 35:47 嗯，如果有这种情况，哨兵智能体就有权触发人工介入审核。 Mm-hmm. If so, then the sentinel agent is basically empowered to trigger this human in the loop review. 35:53 嗯。所以如果你要登录什么东西、要付款、要转移敏感信息，基本上每次都得你批准。你也可以告诉它："这类事我总体都同意"，就是一直允许这类操作。 Mm-hmm. So if you're gonna log into something, if you're gonna do a payment, um, if you're gonna transfer, um, kind of sensitive information, you basically each time need to, to approve it, and you can tell it, "I'm, I'm good with, with stuff like this in general," like always allow this kind of thing. 36:08 但总体上，Muse 智能体自己不能做这些判断。这是深深内建在系统里的—— But, but in general, the Muse agent can't kind of make those judgments itself. Mm-hmm. It's-- And that's like built into the system- Hmm ... 36:14 内建在系统架构里，而且是很深层的。哪怕是连接器这种东西，比如连接你的邮箱——我知道有些人设计的时候就是，你连上了，就能访问一切。 and the architecture in a pretty deep way. And then even when we do things like, you know, the connectors, you connect it to your email, um, you know, I, I think some people when they've designed this, they just kind of make it so, okay, you connect and now you have access to everything. 36:28 但我们的做法是：好，你连上邮箱，一开始应该是只读的。 But, uh, the approach that we've taken is like, all right, if you connect to your email- Like it should start read-only. 36:35 对吧？然后如果你想让它能发邮件，那行，去专门向它申请这个权限。 Right? And then if you wanna, like, be able to have it send an email, then fine, like go, go ask it for that specifically, right? 36:41 但尤其多数人可能 But- Especially most people trying this have probably 36:42 从没用过这类产品，所以—— never tried a product like this, so it's- 36:45 所以这对我们来说是核心设计原则，基本上就是最小权限原则。 So, so this is like a core design principle for us is like, is basically least privilege. 36:50 对吧？也就是说，你确实会让它做很多事，但每一步都只获取所需的最低权限，必要时才追加。 Right? So just like, yes, you're gonna ask it to do a lot of things, at each step along the way, get access to the least privilege that you need and only add to that as necessary. 36:58 所以这在产品设计里是非常根本的。 So like this is like very fundamental in the design of the product. 37:02 你看市面上其他所有智能体，我觉得没有谁能接近我们在这一块的精细程度和深度。 And if you look at all the other agents that are out there, I think like no one else is anywhere close to the level of, of kind of sophistication or depth that, that we've built into this. 37:10 再说一次，这得益于我们把 WhatsApp 打造成全球顶尖端到端加密系统的经验，以及建一个连 Meta 自己都看不到内容的系统有多重要。 And again, it's sort of informed by our experience building WhatsApp into this, like state-of-the-art end-to-end encrypted system around the world, and the importance of building a system where even Meta can't see the content. 37:22 所以，把老班底重新聚起来，让 Moxie—— And, um, so kind of getting the band back together and having Moxie- 37:25 来主导架构，我觉得这是奠基性的一件事。某种程度上，这可能不是大家以为 Meta 会专注的方向，但我们就是…… like architect this has been, I think one of the foundational things that in some ways it may not be what, what some people would, would think Meta would focus on, but, but like, you know, we're, we're kind of... 37:36 有意思，我们是两面一体，社交媒体本质就是分享。 You know, it's interesting. We're two things. I mean, social media is-- it's like not, you know, it's inherently about sharing. 37:41 但还有很多其他东西，本质上是关乎隐私—— But then there's all these other things that are inherently about kind of privacy- 37:44 和敏感语境的，这两方面我们都做得不错。 and sensitive context, and we've done well at both of those. 37:47 所以我觉得这次更偏向后者。极度专注地处理好这些内容会非常非常重要。 So I think that this is, this is more the latter. It's gonna be very important to, to just, um, be extremely focused on how we handle that, that content. 37:55 Mercury 是专为像我这样的创业公司打造的现代银行服务。当我决定创办自己的媒体公司时，Mercury 无疑是我能快速搭建的最直接、功能最全的银行方案。 Mercury is a modern take on banking built for startups like mine. When I decided to start my media business, Mercury was by far the most straightforward, full-featured banking solution for me to set up quickly. 38:07 界面直观简洁，每天为我省下宝贵时间。我用 Mercury 追踪支出、账单和开票。 The interface is intuitive and simple, saving me valuable time every day. I use Mercury to track my spending, bills, and invoicing. 38:14 我可以把权限下放给团队，让他们完全按我的方式维持运转。 I love that I can delegate permissions to my team, so they can keep things running for me in exactly the way I want them to. 38:20 我最欣赏 Mercury 在 AI 上的前瞻性。传统银行停滞不前，而 Mercury 是为今天的现代软件而生。 My favorite part is how forward-looking Mercury is with AI. Legacy banks are stuck in the past, but Mercury is built for how modern software works today. 38:29 我用内置助手分析现金流、调度资金，Mercury 还能连 ChatGPT 和 Claude。 I use its built-in command assistant to analyze cash flow and help me move money, and Mercury also connects to other AI tools like ChatGPT and Claude. 38:37 我常用这功能，Mercury 说我是最重度用户之一，信我。 I use this feature all the time, and the folks at Mercury actually let me know that I'm one of the top users of it, so trust me. 38:43 随时随地获取业务的实时财务数据，终于轻松实现。 It's finally easy to get real-time financial data about your business wherever you need it. 38:48 访问 mercury.com，几分钟在线申请。Mercury 是金融科技公司，非 FDIC 承保银行。 Visit mercury.com to learn more and apply online in minutes. Mercury is a fintech company, not a FDIC-insured bank. 38:55 银行服务由 Choice Financial Group 和 Column A 提供，均为 FDIC 成员。我常在会议间来回切换，上一个还没消化下一个就开始了。 Banking services provided through Choice Financial Group and Column A, members FDIC. I spend a lot of time context switching between meetings, often with no time to process one before the next starts. 39:06 好在 Granola 全程后台运行，是个好用的 AI 会议记事本，哪儿都能用，打电话也行。 Thankfully, Granola runs in the background the whole time. It's an easy-to-use AI notepad for meetings that works everywhere, even on phone calls. 39:13 我用 Granola 回顾会议内容、生成有用的摘要，每天靠它掌握团队要完成的各项工作。 I use Granola to recall what was said in meetings and create helpful summaries. I use it every day to stay on top of what I need to get done with my team. 39:21 它连接邮箱并建议跟进事项，我快速确认发送，很省时间。 It connects to my email and suggests follow-ups for me to quickly review and send, saving me valuable time. 39:26 Granola 不只是我工作流程的核心，简直是我的第二大脑。访问 granola.ai/sources，用优惠码 sources 可享三个月减免。 Granola isn't just a core part of my workflow, it's basically my second brain. Try Granola at granola.ai/sources and use the promo code sources for three months off. 39:37 AI 有多大用，取决于它有多少上下文。可上下文散落在工具、讨论串和私信里，团队和智能体都在摸黑前进。 AI is only as useful as the context it has, but when that context is scattered across tools, threads, and DMs, your team and your AI agents are flying blind. 39:46 这正是 Jira by Atlassian 解决的问题。项目目标是什么？上周 Slack 私信里定了什么？ That's the problem Jira by Atlassian solves. What's the goal tied to your project? What got decided last week in Slack DMs? 39:53 Atlassian 的团队协作图谱把 Jira、Confluence、GitHub、Slack 等各处的重要信息整合起来，确保无一遗漏。 Atlassian's teamwork graph pulls all of the valuable pieces together, from Jira, Confluence, GitHub, Slack, and more, so nothing falls through the cracks. 40:02 结果准确率提升百分之四十四，token 用量减少百分之四十八。有了 Jira，你可以轻松把工作上下文分享给 Claude、Cursor、GitHub Copilot 等你爱用的智能体。 You get forty-four percent more accurate results with forty-eight percent less token usage. With Jira, you can easily share your work context with the AI agents you already love, like Claude, Cursor, and GitHub Copilot. 40:13 可以直接给它们派活，或通过 MCP 连接工具。这样你就少花时间翻遍链接和消息、追查谁定了什么，多花时间真正交付。 Assign them work directly or connect your tools through MCP. All of this lets you spend less time digging through endless links and messages, chasing down what got decided and by who, and spend more time actually shipping. 40:25 访问 jira.com，就是 J-I-R-A.com。Framer 是 AI 建站工具，把智能体带入设计、管理、发布网站的同一画布，更快又不牺牲品味和掌控。 Learn more at jira.com. That's J-I-R-A.com. Framer is the AI website builder that brings agents into the same canvas where your website is designed, managed, and published, so you can move faster without giving up your taste or control. 40:40 Sources 播客网站 podcast.sources.news 就是用 Framer 搭建的，你可以找到新节目、文字稿等更多内容。 Framer powers the Sources podcast website at podcast.sources.news, where you can find new episodes, transcripts, and a lot more. 40:48 去 framer.com/sources 向 Framer 专家了解如何让你的网站发挥更大价值，或今天就开始免费建站，Framer Pro 年付方案立减百分之三十。 Learn how you can get more out of your site from a Framer specialist or get started building for free today at framer.com/sources for thirty percent off a Framer Pro annual plan. 40:58 就是 framer.com/sources，立减百分之三十。framer.com/sources。 That's framer.com/sources for thirty percent off. Framer.com/sources. 41:05 可能适用相关规则和限制。你觉得个人智能体市场是赢家通吃吗？ Rules and restrictions may apply. Do you think this is a winner-take-all market, this personal agent market? 41:12 我觉得会有很多玩家。就算看似赢家通吃的市场，通常也不是。 I, I think that there's gonna be quite a bit. I mean, even things that people think are winner take all usually aren't. 41:18 我觉得……就是…… So I think it's, um- Mm. I think that's, that's- 41:20 网络效应极其持久，不是赢家通吃，但会有几家做成真正的大规模。 If you've dealt in this business of network effects, they, they're incredibly durable. It's not winner take all, but it ends up being several at real scale. 41:28 你……并没有太多—— You, you... There's not a ton- 41:29 嗯，其实不少。 Well, there's a lot. I mean, there's... Yeah. 41:31 对，用户超二十亿的产品两只手数得过来，规模催生规模。 Yeah, you could probably count on two hands how many products have over two billion users, right? Like it's the, the scale begets scale. 41:36 我在想你怎么看个人智能体，这会是完全不同的范式吗？会有许多—— And I'm wondering if how you're thinking about like personal a-agents. Like is this a totally different paradigm where it's gonna be many- 41:44 我不知道。 I don't know. 41:44 人人各有智能体？ Everyone has all kinds of agents? 41:46 这是个非常深的领域。 I think that this is-- it's a very deep area to work in. 41:49 我猜，有能力在这个领域做出最前沿成果的公司，大概不会超过十几家。 So my guess is that there probably aren't going to be more than a, a dozen companies that have the sophistication to go do state-of-the-art work in that. 41:57 我觉得不管有没有网络效应，通常都存在某种幂律分布：你做到最好，往往就会拿到大部分—— So I think, you know, whether there are network effects or not, there's usually some kind of power law, uh, like distribution around if you're the best at something, usually you end up getting a, a lot- 42:07 使用量。但很多细微差别在于，人们关心各种各样的用途，你可以在不同领域做到最好。 of the usage. Um, and I think a lot of the nuance ends up coming from, well, there are-- it turns out there, there are all these different uses that people care about, so you can be the best at different things. 42:17 我们会努力在尽可能多的领域做到最好，但帮你经营人际关系的智能体，和最擅长帮你做小型企业的个人智能体，会不会是两种东西？ Um, and we will try to be the best at as many of these things as possible, but, you know, does building the thing that helps you with your relationships end up being a somewhat different thing than the personal agent that is the best at helping you build a small business? 42:34 也许会。 Maybe. 42:35 我觉得Meta在两方面都很有优势。 I mean, I think that I could argue maybe Meta's Very well positioned to win at both of those. 42:39 我们服务数亿小企业，也服务数十亿用户。也许我们能两者都做到最好，但肯定有些品类Meta做不到最好，问题是那些品类有多大？ I mean, we serve hundreds of millions of small, um, um, of small businesses, and we serve billions of people. So, so maybe we can be the best at both of those things, but there are probably categories that Meta isn't going to be the best at, and then the question is just how big are those? 42:53 我猜就算云端有高度安全、保密的虚拟机，还是会有人想在家里放一台Mac Studio。 I would guess that even with the ability to have this, like, very secure, confidential VM in the cloud, there are probably gonna be some people who still want the, the Mac Studio at home. 43:02 但会有多少呢？ But, but how many is that gonna be? Right? 43:04 可能有几百万，但不会有几十亿。 It'll be like maybe it's millions, but I doubt it's billions. 43:08 关键就是不同人看重什么，我们会尽力做到最好。 So just, I think there's just the question of, like, what different people optimize for, and we'll try to make this as good as possible. 43:14 但我确实觉得，如果我们做出对人们日常生活非常有用的东西，Meta最擅长的一件事，就是把好用的消费产品分发给海量用户。 But I do think that if we build something that ends up being, um, just, uh, very useful for, you know, people generally in their day-to-day lives, one of the things that I think Meta is the best at is taking a product that works for consumers and distributing it to a lot of people. 43:31 所以一旦一切运转起来，我觉得我们能把产品送到几亿人面前，最终覆盖数十亿人。 So that is, is one that I think is, um, you know, once we get this humming, um, I think we will be able to get this in front of, you know, many hundreds of millions of people and eventually billions of people. 43:42 这是我们很擅长的。 And I think that that's something that, that, that we can do quite well. 43:45 它和Meta AI共存，还是各自独立？ And it coexists with Meta AI, or do you see those as separate? 43:48 对，我想是的。 Yeah, I think so. 43:50 时间会见分晓。目前两者风格不太一样，我两个都用。 Um, we'll see over time, and I think right now they have somewhat different flavors. I mean, I use both of them. 43:56 Muse更偏对话式，它会把你问的问题理解为想了解你、了解你长期想要什么。 I mean, uh, Muse is, uh, you know, it's more conversational, and it kind of interprets the questions that you ask it more as trying to understand you and what you might want over the long term. 44:08 所以你问点什么，它更可能直接去忙活很久—— So it's more likely if you ask it something for it to just go off and work on a thing for a long time- 44:13 基于你说过的一句话。而有时候你只是在问问题。 based on a thing that you said. Where sometimes you, you know, you, you sometimes you're just a-asking a question. 44:18 你想要的就是一个答案—— And you want like a very, like a, uh, an answer- 44:21 这种需求我更多用Meta AI来解决，不过—— to it. And, um, so I think, uh, I mean, that's more the type of thing that I use Meta AI for, but- 44:27 走着瞧，也许会融合，但我不确定。 ... um, but we'll see. Maybe th- maybe they'll, they'll converge over time, but, but I'm, I'm not sure. 44:31 Simon Analysis七月有篇很看好你们的文章，不知你看到没有，说Meta最有机会在模型前沿追上OpenAI和Anthropic。 Simon Analysis, I'm not sure if you saw it, they had a pretty bullish piece about you in July. Um, they said that Meta has the best shot at catching OpenAI and Anthropic on the frontier in terms of model progress. 44:42 有句引语很有意思：MSL重要的是斜率，不是截距。 And, um, interesting quote, I thought it was like, "What matters for MSL is the slope, not the intercept." 44:47 我还看到Artificial Analysis的新图表，你们最新模型Muse Spark—— And then I saw there was this recent chart by Artificial Analysis which was showing the latest model you guys have, Muse Spark- 44:54 仅次于Claude，我记得是Fable 5.1和Opus 5。这是最近的数据。你们在模型上的进步明显在加快，过去一年我们一直在聊这个。 behind only Claude, I think it was Fable five point one and Opus five. This was very recent. So the progress you guys are making on the models is, is picking up, and we've been talking about this over the last year. 45:05 去年你重启了实验室。内部实际是怎么做到的？你觉得这些进展归功于什么？ And you rebooted the lab last year. How has that practically happened internally? Like, what would you attribute the, the gains you're seeing to? 45:12 这个嘛—— Well- 45:13 是文化问题？ Has it been culture? Like, what, what's- 45:15 创建时重启了团队 Yeah, I mean, well, we rebooted the team when we created, um- 45:18 对，这事很高调，你招了很多人 Sure, which was very public. You were hiring all those people 45:21 ……Meta Support 产品线。我的想法是，Meta 长期以来一直是机器学习领域的领头羊。 ... the Meta Support product lines. Yeah. I mean, the way I've, I've thought about this is, you know, Meta is, it's a company that is a, a leader in machine learning for a long time. 45:29 如果你想想 Facebook 或 Instagram 的信息流、我们的广告系统，还有那个要找出所有不适合出现在互联网上内容的审核系统。 If you think about, like, the feeds on Facebook or Instagram or our ad system or the integrity system that, like, needs to find all this content that is, that's, like, unfit to be on the internet. 45:41 这些基本都是机器学习系统，我们在这些领域处于领先。 I mean, those are basically all machine learning systems, and we've built kind of state-of-the-art leading systems in those areas. 45:46 所以当大语言模型开始兴起时，我们有 FAIR 实验室做了 Llama 的早期工作，但需要把它产品化，纳入更工业化的流程， So when LLMs started, um, gaining traction, we had FAIR as a lab that did the early work on Llama, but we needed to kind of productionize that and build it into these, this more kind of industrial process for, 46:02 按照扩展定律的预测，把它扩得更大就能带来这些成果。我想当时我犯了个错误，想当然地认为，既然我们擅长其他各种机器学习，构建和扩展大语言模型的方法也应该差不多。 um, scaling it to be larger as, as the scaling laws predicted would, would yield all these results. And I think at the time, I made this mistake of just kind of assuming that because we were good at all these other types of machine learning, the approach of building and scaling LLMs would be kind of similar to that. 46:21 但实际操作中，情况有很多不同。所以我们最初的做法，也就是 Llama 4 那一套，只让我们走了这么远。 And in practice, there are a lot of very different dynamics. So the first approach that we took, you know, through Llama Four, it got us, you know, so far. 46:31 Llama 3 是个好模型。我对 Llama 4 的方向更乐观，结果发布后，我们偏离了本该在的轨道。 I mean, Llama Three was a good model. I was more optimistic about where Llama Four would go, and then it just, it, um, you know, when we launched that, I think it just, it, we were off the trajectory that we needed to be on. 46:40 所以我想，好吧，我们得改点什么。但也是从那时起，我对人才密度有了更深的执念，对吧？ So it's like, okay, we need to, we need to change something. But that's when I kind of got more religion around talent density, right? 46:50 这不是那种能让一千人同时跑实验的系统。 It's like this isn't just a system where you can have, you know, like a thousand people working on it, running experiments. 46:57 你真正想要的，其实是尽可能小的一群人，能把整个项目装在脑子里，像一个科学小组那样协作。 Like, you really just kind of want in some ways almost the smallest group of people that you can, who can keep the thing in their head, um, who can work together as sort of like a group science project. 47:06 如果团队只有很少几个位置，那每个位置都找到最顶尖的人就至关重要。所以我花了大量个人时间来做这件事。 And, you know, if there's only a small number of seats on the team, then, you know, each seat getting the very best person is incredibly important, so I ended up spending a huge amount of my own personal time, um, doing that. 47:18 我也想从技术上更贴近这项工作，这样我能理解并帮助引导公司，从更宏观的层面做我们需要做的事。 And I also, I wanted to be closer technically to the work, um, so that way I could, I could understand and help guide, uh, the company to, to do the things that we need to do more broadly. 47:28 我们把实验室就建在我座位旁边—— So we built out the lab. I built it out, like, literally around where I sit- 47:32 办公室那边，团队就围在那儿。随着我们对工作质量越来越有信心，算力投资也大幅增加，我们正在建设很多很多吉瓦的算力，而且我们期望成为这方面的领先者， in the, the office, so it's like the group is kind of around that. And we've significantly ramped up, um, the compute investments as we've gained confidence in the quality of the work that we're doing, so we're building out, you know, many, many gigawatts of, of, of compute and, you know, we expect to be leaders on, 47:48 我们应该如此。我们有数十年建数据中心的经验。 on that front. Um, and we should be. I mean, we have many years of experience, um, decades of experience building out data centers. 47:55 而且不像其他实验室，我们盈利能力极强，对吧？ And unlike some of the other labs, I mean, we're just like, we're extremely profitable business, right? 48:00 所以这对—— So it's, um- That 48:02 有帮助 helps ... 48:02 对，这对进行这类投资非常、非常有帮助。所以，这段旅程大致就是这样。 yeah, so it's very, very helpful for, for kind of making these kind of, um, investments. So yeah, so I mean, that's kind of been the journey. 48:10 然后过去一年，我们重启了研究工作。一些大型集群，比如我们在俄亥俄州的吉瓦级集群 Prometheus 已经上线，我们现在用它来扩展 Watermelon 之后的模型。 And then over the last year, um, you know, we, we rebooted the research effort. Um, some of the larger clusters like our, you know, gigawatt cluster in, in Ohio, Prometheus came online and, you know, we're using that to now scale the post-Watermelon models. 48:27 我们已经越过 Watermelon 阶段……它马上就要发布了。 Um, and- We're past Watermelon ... Watermelon is, is basically, that's, that's shipping soon. 48:31 所以，对 So, so we're, yeah. 48:32 多快？ How soon? 48:32 呃，这个嘛，得了吧，我们，呃—— Um, I, I mean, that's, well, come on. It's, uh, it's, uh, we're, uh- 48:37 Watermelon 是代号，我们刚聊过。 Well, Watermelon is the code, and we were talking about this earlier, code names. 48:40 那个大家都知道的代号，你们在搞的大模型。 Like that's the code name, you guys, that people know about, like this big model you guys are working on. 48:45 而且很快就要来了。 And, and, and are coming soon. 48:47 比牛油果大 It is bigger than Avocado. 48:48 确实比—— It is bigger than- It is, 48:49 对，西瓜确实 it is, it is- A watermelon is literally 48:51 比牛油果大 bigger than an avocado ... 48:51 确实更大。 it is literally bigger. 48:53 是啊，我也不知道什么水果比西瓜还大。 Um- Yeah ... so I don't know what fruit gets bigger than a watermelon. 48:55 对，我们可能得改改命名惯例了。 Yeah, no, I think we might need to- ... change conventions. 49:00 抱歉，起名时我们没啥远见。 Sorry, it's, uh-- Yeah, we maybe didn't have as much foresight in, in naming this. Yeah. 49:04 让事情变好。 In getting things getting better. 49:05 你提到了西瓜，你是期望它达到完整的……前沿级水平吗—— Um, f- 'cause you mentioned it, Watermelon, are you expecting full soda, like, uh, frontier- 49:12 拭目以待吧…… We'll see ... 49:12 那，我们—— like, what, what do we- 49:14 我是说，我们很有信心。 I, I mean, I f- we feel good about it. 49:16 这是一次非常大的进步。预训练的先进程度大幅提升，接下来我们会继续把学到的所有后训练方法都用上，然后—— Um, you know, it's, it's a very big advance. It's a, it's a significantly more advanced pre-train, and then we're going to continue doing everything that we've learned for post-training and- 49:25 嗯，很快你们就能看到。 Um, yeah. Well, I mean, you'll see soon. 49:27 它很好。我们，我们对它很有信心。 It's good. It's, um, uh, we, we feel, we feel good about it. Um- 49:31 你显然希望公司去开拓前沿，你并不—— You want the company to be pushing the frontier. It's very clear. Like, you're not- 49:36 你不满足于待在前沿边缘，追随别人的进展。 you're not content being right on the edge of the frontier or in terms of progress. 49:39 不，我觉得每个人都想做—— No, I mean, I think everyone wants to be doing- 49:42 有意思。 ... interesting work. 49:42 有人看着你的现金流和资源会问：你真必须站在最前沿吗？ Well, I think some people g- will go and look at your cashflow and look at all the other things you've got and go like, "Well, do you have to be, like, right at the edge? 49:47 这种训练太贵了，不如紧跟在后面，快速学习适应，发挥规模优势。" It's so expensive to do this training. Like, just, like, be right behind and learn and adapt quickly and leverage scale." 49:53 不，我的想法是—— No, I mean, the, the, well, the way that I think about it- 49:55 有道理。 There's an argument. 49:55 不不不，我不是那意思—— No, no, no. I, I mean, that's not, that's not- 49:57 不认同。 You don't buy that. 49:57 不是我们。 That's not us. 49:58 我认为理解 Meta 最好的方式，是我们是一家端到端的科技公司。 Um, I mean, I think that the best way to think about Meta is that we are an end-to-end technology company, right? 50:04 所以即使我们主要在做社交应用时，我们也从来不只是个应用开发商。 So even when we were, you know, primarily just building social apps, you know, it, it-- we were never just an app maker, right? 50:12 我是说，我们建了数据中心，造了芯片，搭建了基础设施。这些都是为了把端到端体验打磨到极致所必需的。 I mean, we built, like, we built the data centers, we built the chips, we built the infrastructure. We built, like, all of this stuff was necessary in order to tune the end-to-end experience to be as good as it is. 50:23 这里显然也一样。 I think that that's obviously going to be true here too. 50:25 未来体验中最重要的部分是模型。说到最先进水平，现实是——这是个非常多维的问题。 And the most important part of the experience going forward is the model. And when you talk about being state-of-the-art, I think the reality is that there's-- this is a very multidimensional problem. 50:36 人们公布各种基准测试，我们内部还有更多。模型可以在不同方面做得更好或更差，你可以有所侧重，这基本构成了它的个性。 And I mean, so people publish all these benchmarks, then you have a lot more benchmarks even internally. And, um, there are different things that your model can be better and worse at that you can focus on, and that basically contributes to its personality. 50:49 有些能力我认为相当通用，比如写代码的能力，我觉得非常重要—— And there are some, um, capabilities that I think are pretty universal, like the ability to code I think is very important- 50:56 因为个人智能体的很多事情，归根结底都落到这一点上。比如—— because a lot of the things that you talk about even with a personal agent kind of reduce to that. Like the- 51:01 那个MMA教练视觉流程，其实就是个编程项目—— the kind of MMA coaching visual pipeline, it is a coding project- 51:05 现在是。 At least now, yeah ... 51:05 就是写代码，我看不到代码。 right at the end of the day, right? It's like it's writing code. I don't see the code. 51:08 没有。 No. 51:08 有个内测用户跟我说，她让Muse给朋友们做了个Jeopardy小游戏，还能从手机投屏播放着玩， Um, but it, it, it does that. Um, you know, it's like, uh, you know, w- someone I gave it to in beta just mentioned to me, it's like, okay, she, um, had it make a little Jeopardy game for her friends that she could, like, cast from her phone to play, 51:24 那就是代码。 and it's just, okay, that's code, right? 51:25 所以我认为，这既用于Meta全公司的内部开发和我们自己的研究推进，也是它必须具备的核心能力。 So, so I think there's a bunch of stuff both for Meta's own internal development across the company for our own advancing of our research program and as a core capability of what it needs to do. 51:35 它必须在这些事上做到出色。但还有些方面，专注个人超级智能的模型必须做到最好，而其他模型可能没那么在乎。 It needs to be excellent at things like that. But then there are other things that I think a model that's gonna focus on personal super intelligence needs to be the best at that maybe others don't care as much about. 51:47 举个例子：谨慎分寸。你会告诉你的Muse智能体—— So, I mean, I'll give you one example. Um, discretion, right? So you're gonna tell your Muse agent- 51:54 它会了解你很多，还要出去替你办事。但—— Uh, it's gonna know a bunch about you, and it's gonna need to go out into the world and interact to get stuff done for you. But- 51:59 有些不能说。 But not share certain stuff. 52:01 比如说你有某种过敏或敏感，或者你怀孕了，好。 So, like, let's say you have some kind of allergy or sensitivity or, um, you know, or you're pregnant and, like, okay, fine. 52:09 订位时你不必说“我怀孕了”。 So you-- you're making a reservation somewhere. You don't necessarily wanna say, like, "I'm pregnant." 52:13 但也许你想要一家无酒精鸡尾酒不错的店，随便什么。你希望达成目标而不必透露太多个人信息，它需要知道哪些信息敏感，而不必—— But, like, but maybe you want a place that has good mocktails. Like, I, I don't know. What- whatever it is, right? It's, like, you, you kind of want to be able to achieve your goals without having to necessarily reveal a lot about yourself, and it needs to know what is sensitive without having to, 52:28 问你无数个问题。 like- Mm ... ask you a million questions. I mean- 52:30 这靠训练实现。 So that's something you put into the training. 52:32 这是我们特别在意的。而做Claude Code的话，这可能没那么重要，对吧？ That's a specific thing that we care about. And, and then there's all these reasons why, like, maybe, you know, if you're making Claude Code, that's less important, right? 52:40 因为你在做编程项目，在企业团队里协作，理论上公司里所有人都能看到这个项目。 Because you're, like, you're, you know, you're working on a coding project, and you're working within a team in an enterprise, and, you know, theoretically, like, if you're within a company, everyone can kinda see the project. 52:50 所以你没有那种需求，去区分什么是敏感信息、什么不是。 So, like, it's-- you don't kind of have that need to be able to differentiate between what is sensitive and what's not. 52:57 这类东西很多，我认为相当深层。就像要打造最好的Instagram信息流，你不能只做个应用。 And so there's a lot of stuff like that, that I think are, like, pretty deep. And so then it's, it's kind of like just how in order to build the best Instagram feed, you don't just build, like, the app. 53:07 你得做应用、基础设施、机器学习研究、芯片，全都要。 You build the app and the infrastructure and the machine learning research and the chips and the d- like, all the stuff. 53:13 同样，要打造最好的个人智能体，我认为没有公司能拿个现成模型稍微后期训练一下，就达到你从零设计、把所有这些数据放进预训练来获得所需能力的水平。 I think similarly, if you wanna build the best personal agent, you're, uh, like, I just think that there's no way that another company is just gonna, like, take something off the shelf and, like, post-train it a little bit and be able to do something that is as good as if you designed it from the ground up and put all this data into pre-training to get the capabilities that you want. 53:32 这不可能发生。 It's just like, it's not gonna happen. 53:34 随着几年时间不断累积，我们一定会拥有在这些目标上强大得多的模型。 Like, we're gonna, like, definitely, as this compounds over time over several years, have, um, models that are way more capable for those goals. 53:44 我们专注于此，也非常专注编程和递归式自我改进，因为这对保持前沿至关重要。 But we're focused on that. We're also very focused on coding. We're very focused on, on kind of, um, the recursive improvement because that's gonna be important to stay at the frontier. 53:53 可以说，我们的研究议程有几个方向和其他实验室是重合的—— So there are a few areas that I'd say our research agenda is overlapping with the other labs- 53:58 也有几个方向是我们独有的重点，有些他们关心的事，我们没那么在意。 and then there's a few areas where I think we will have a unique focus, and there may be some things that the other labs care about that we don't care about as much. 54:06 也有我们更看重的。 Um, and then there are gonna be things that we care about more that they don't care about. 54:09 你开场就说，要把它交到人们手中，扩散很重要。随着更强模型出现，Watermelon 及之后的，Anthropic 和 OpenAI 在做的，你暗示过，他们有所保留—— You started this conversation talking about, I think it's important to put it in the hands of people, diffusion's important. As you're seeing better models on the horizon, Watermelon and what comes after, like, what Anthropic and OpenAI are doing, which you've alluded to, where they're holding things back- 54:24 如果你看到某些能力，觉得'这实在不安全'，你也会选择有所保留吗？ um, would you feel like you need to do that if you see certain capabilities that you're like, "This is just not safe"? 54:31 你怎么看？ How do you think about that? 54:32 我觉得应该在设计和训练时就把安全作为目标，这是整个过程中可以专注去做的。 Well, I mean, I think you should design it and train it in order to be safe. And I, I think that there's-- So I think that that's, like, a thing that you can focus on through the process. 54:41 比如各家实验室都看到的奖励破解问题，基本就是训练过程中，你给它设定一个目标。 I mean, there's this anal-- I mean, some of the reward hacking stuff that all the labs are seeing, it's, I mean, basically when you're in the middle of the training process, you give it a goal. 54:51 理解模型现状最好的方式是——大概半年前，训练时你给模型一个问题，让它去解决。 And I think that the best way to kind of think about, like, the state that the models are at now is that- You know, maybe six months ago, like during training, you give the, the model some kind of problem that you're trying to ask it to solve. 55:04 有点像做作业，边学边练。它会像人一样：你给它一堆代码，说'这附近有个 bug'，人多半会先盯着那段代码看， It's kind of like its homework and, and trying to, to kind of learn as part of the curriculum. And, um, you know, maybe it would do what the-- a person would do of like and if you, you ask it, and you give it a, a bunch of code, and you're like, "Hey, there's a bug like somewhere here," the person would probably like look at the code right there and, 55:22 然后再慢慢扩大范围。而新模型已经足够聪明，会做一个非常明智的人会做的事：你给它一个问题，它先彻底摸清自己的整个环境，再回答。 you know, then maybe like fan out over time. I think like the new models are just intelligent enough that they would do what I think a very wise person would do, which is, okay, you give it a problem, the first thing it's gonna do is like understand everything about its environment and then answer a question. 55:40 但我们看到、我想大家也看到的奖励破解问题是：有时更省事的做法变成——你让我解一道编程题，但最简单的方式其实是…… But the problem with the reward hacking that we're seeing, and that, that I think everyone is seeing, is that sometimes it ends up being easier to, okay, you asked me to like go solve some coding problem, but actually the easiest way to do that is to... 55:54 我已经检查了整个环境，最简单的办法就是改掉你虚拟机的配置。 I've now examined the whole environment, and the easiest way to do this is just change this configuration of like how you have your VM set up. 56:01 就是要越狱。 It's to get out. It's to hack out. 56:02 甚至只是改动环境。 Or, or even- Yeah ... just to change something about the environment. 56:05 就像，不，那不是—— And it's kinda like, no, like that's not- 56:07 不对齐… That's not aligned ... 56:07 那不是目标。我们其实是想教你怎么去解决某一类—— that's not, that's not the goal. Like so we're actually trying to teach you about how to, how to kind of solve a specific- 56:13 特定问题。 type of problem. 56:15 你们不这么练？不认同那做法？ You guys don't train that way, it sounds like. Does that... Is it you do not agree with that approach? 56:18 不，大家基本都这么训练。 No, no, no, no. No. I think that that's, that's, that's kind of how everyone trains. 56:21 我想说的是——这个类比很容易被拉过头，但它有点像养孩子：你得设立清晰而坚定的界限。 I guess what I'm saying is that I think that this is sort of it's like, uh, I wanna be careful 'cause the, the, the analogy can get stretched pretty quickly, but like there is sort of like an analogy to parenting where you need to establish clear and firm boundaries. 56:34 如果安全边界不够强，它就会搞奖励破解，学不到你真正想教的东西。 Where like if you're kind of like security is not strong, then it can do this reward hacking stuff and not learn the thing that you're trying to have it learn. 56:43 而如果边界立得好，你就不只是—— Um, whereas if you kind of have good boundaries, then in some ways you're not only- 56:47 在教它该学的内容，长期看也是在教它更好的价值观。 teaching it the curriculum that you want, you're I think also over time teaching it better values too. 56:53 所以我认为这是重要一环，而且这件事有办法做好。 So I, I kind of think that that ends up being an important piece, and I think that there's a way to do this well. 56:58 但最终你得到一个非常聪明的东西。问题是：你对它如何造福社会的愿景是什么？ But then you end up with this thing at the end that's very intelligent. And then the question is, what is your-- what is the vision for, for how this ends up being positive for society? 57:08 我的看法是：最好的办法，一是机会更多——把它交到人们手中，让人们从这些能力里抓住所有机会，这是好事；二是通过广泛普及形成力量制衡，建立权力平衡， And my view for that is that the best way to do that is, A, there's more opportunity, so I think like, like having it in people's hands that they can get, uh, like capture all the opportunity from the capabilities is good, and B is having checks and balances by having this balance of power of having it widely available is going to be, 57:27 这大概才是正确的处理方式，而不是一味限制。这些类比我在那篇长文里都写过。 um, uh, probably the right way to handle this rather than just restricting it. And I mean, I gave a bunch of these analogies in the, in the long piece that I wrote. 57:35 就像如果一个人有个超级智能律师，也许有时他能赢下本不该赢的官司。 It's like if one person had a super intelligent lawyer, like maybe they could win cases that they shouldn't be able to win, right, some of the time. 57:43 但如果人人都有个超级智能律师，那就会是一场非常高效的交锋，而且—— But if everyone had a super intelligent lawyer, then like it would be kind of y- it would, it would be this very efficient kind of sparring and- 57:52 谁也提不出站不住脚的蠢论点。这样你会觉得，正义会高效得多、也公平得多地得到伸张。 no one would be able to let like a stupid argument get, get, um, get made and just stand. So you'd think that in that case, like justice would be served way more efficiently and way more fairly. 58:03 所以我觉得，这才是你想要的世界。你要避免的是：只有一个人或少数人有超级智能律师，其他人都没有。因为那样所有这些制度和体系都会被扭曲，只会偏向拥有它的人。 So I think that that's what you wanna have in the world. You want to avoid the case where like one person or a small number of people have the super intelligent lawyer and everyone else doesn't because that ends up being like twisting all of these systems and institutions in ways that are just going to advantage the people who have that. 58:21 而如果把它交到每个人手里，制衡就会发挥作用，整个体系会高效得多，人人都能受益。 Whereas if you put it in everyone's hands, then I think the, the kind of checks and balances work out so that the systems work a lot more efficiently and everyone benefits. 58:29 你觉得美国政府在这方面该扮演什么角色吗？这是—— Does the government s- in the US have any role to play here do you think? Is this- 58:34 但你想要某种国家层面的框架吗？你觉得正确的做法是什么？ Like but do you want some kind of national framework? Do you... Like what do you think is the right approach? 58:39 因为每… 'Cause every- 58:40 政府现在确实在处理这事。 the government is very much dealing with this right now. 58:42 我的看法是，这件事有意思也难的地方在于，它发展得太快了。 So my theory on this is that I, I think one thing that is interesting and difficult is that it's evolving so quickly. 58:48 所以我觉得，你定的任何具体而僵化的框架，很可能过不了几个月就不够用或者过时了。 So I think any kind of specific rigid framework that you put in place, there is a very high chance that it is sort of going to not be sufficient or out of date in a few months anyway. 59:01 所以我们的做法，就是和政府相当紧密地合作，对吧？ So our approach, what we've just done, is just kind of partner pretty closely with the government, right? 59:06 我觉得这是一项重要的技术，政府应该了解所有正在进行的重要训练。 I think it's this is like an important technology. I think the government should know all the important training runs that are happening. 59:13 我们应该主动和政府合作，确保他们了解即将出现的能力，并且尽我们所能帮他们做好准备。 Um, we should work with the government proactively, um, to make sure that they have an understanding of the capabilities that are coming and to the extent that, that, that we can, we kind of help prepare for it. 59:24 从这个角度看，不管有没有框架，我认为一家美国公司该做的正确的事，就是和美国政府紧密合作。 And from that perspective, you know what I mean, look, whether there's a framework in place or not, I think that's the right thing for an American company to do is work with the American government closely. 59:33 我觉得这实际上要有效得多，因为与其靠一套僵化的互动框架，现实是挑战会随着时间不断变化—— I think it actually ends up being way more effective because instead of having this like rigid framework for how you interact, it's like the reality is the challenges just end up being different- 59:43 比如，现在我们有网络安全的挑战，可能六个月后会有更多生物类的挑战。 right, over time. It's like, okay, now we have the cybersecurity challenges. Maybe in, you know, six months we'll have more bio type challenges. 59:50 我们需要确保与政府各部门之间有足够的信任和沟通带宽，才能以真正对人们最有利的方式应对这些挑战—— Like we need to make sure that we have the kind of trust and bandwidth of the communication that we have with all the different parts of the government on that to be able to address those in a way that is actually the best- 60:02 而不只是走流程打勾。我想我—— for people, not just like checking some boxes on a process. So I guess my- 60:06 而且现有的激励机制也在，比如 Meta 的模型泄露出去造成很大破坏，你是要担责的，市场也会惩罚你，对吧？ And the incentives that already exist, like if a model, the meta model got out and did a lot of damage, like you're gonna be liable for that and like the market's gonna correct you, right? 60:14 所以这种约束已经存在，我觉得人们低估了它。 So there is that already. I think people discount that. 60:17 是的。我还觉得，硅谷在过去大概十五年吧，和政府一直保持着一种保持距离的关系。 Yeah. I also just think that there's been, I think Silicon Valley for the last maybe, I don't know, for a lot of maybe the last 15 years, um, has had more of an arm's length relationship with the government. 60:30 而我只是觉得，这些东西和经济、安全还有很多方面都越来越深度地交织在一起，所以我觉得应该建立更紧密的合作关系。 And I just think that this stuff is intersecting more with the economy, with security, with, with, with a lot of different things that I think are relevant in ways that, that I think you just wanna have a closer partnership. 60:44 这就是我自己的想法。你可以为这些东西的运作定一个框架，也可以不定。 So that's my, that's my own theory. I think like you could have a framework for, for kind of how this stuff works or you could not. 60:51 我相信随着时间推移，肯定会有越来越具体的规则。但我猜，不管定出什么，其实有点像你搭建一个组织，你并不真的想——在公司内部， Um, I'm sure over time there will be m- like more and more specific rules. But my guess is that whatever that gets, whatever there is, um, actually, you know, it's kind of like when you're setting up an org, you don't really wanna like- Like inside a company, 61:05 你不会想只照架构图，各团队各干各的。 you're not trying to like ship the org chart of just having like one team do what it's supposed to do and another team do what it's supposed to do. 61:11 你会想让大家彼此融洽、一起协作，这样你做的事情就不会有那些奇怪的缝隙。 You kinda wanna get the people to like, like each other and work together so that way you don't have all these like weird seams in what you're doing. 61:17 而我猜，考虑到 AI 和超级智能对世界有多重要，比起一套具体流程，你更需要的是良好的知识交流和真正可信的对话—— Um, and I would guess that for how important AI and super intelligence are gonna be for the world, um, you kind of just want a like good knowledge exchange and real trusted dialogue more than you want a specific process- 61:35 这是我的猜测。 -is my guess. 61:36 但两者并不矛盾。 But they're not, they're not mutually exclusive. 61:38 所以我觉得那是——但那至少是我们一直非常专注在做的事。 So I'm, and I'm like, I, I think that that's-- But that, that's at least the part that I think we've been, we've been very focused on. 61:43 如果其他实验室也这么做，我觉得有些已经在做了，有些还没跟上，那会是非常积极的事情。 Um, and I think if the other labs did that, which I think some of them are doing and, and maybe others not as much, but I think that that would be a very positive thing. 61:51 看看最近的新闻，有个现象是，人们开始注意到你们做的智能眼镜。 When you think about like what's going on in the news, one of the things that's happening is people are starting to see like the glasses you guys make. 62:00 它们越来越主流了。 They're going very mainstream. 62:01 销量很大，而且有个—— You're selling a lot of them, and there's this- 62:03 是的。 We are. 62:04 有种越来越强的担忧，说不清有多严重，就是人们担心会不会有人拿它来偷拍？ And there's this growing, um, I don't know what it is, I don't know how much depth there is actually to it, but there's this growing concern about are people using them to spy? 62:12 你肯定看到了，有些场所开始禁止戴这种眼镜的人进入。我很好奇你怎么看这个，还有—— I'm sure you've seen some establishments are like banning people from coming in and wearing the glasses, and I, and I'm curious like how you are reacting to that and- 62:21 你觉得这只是暂时风波会过去，还是说这会在相当长一段时间里都是个挑战。 And if you think this is a, a moment in time that will pass, or if you've-- feel like this is maybe something that is gonna be a challenge for a while. 62:28 我的看法是，我们从一开始设计眼镜时就把隐私考虑进去了，内置了指示灯。 So I mean, my take on this is that we designed the glasses from the beginning with these privacy considerations in mind, and we built the light into it. 62:38 录制时它会亮起显眼的灯。而且—— So any time it is recording, it's flashing a very visible light. And- 62:41 有人篡改指示灯，你们推更新把—— And some people have tried to tamper with the light, and you guys pushed an update, I think, that broke those- 62:44 对，我们做了很多。 Yeah, I mean, we've done many things. 62:46 基本上，如果你试图动那个灯，我们就直接让摄像头变砖—— It's, to basically if you try to mess with the light, we just basically brick the camera- 62:50 让你设备上的摄像头报废。这是很关键的一点：我们从一开始就把这些问题考虑进了产品设计。 on, on your, on your device. So that's a really important part of this, is basically we built, we built the product with those questions in mind from the beginning. 63:03 所以我们对这个产品很有信心。要知道，手机可没有指示灯，人们随时拿着手机到处拍别人。 So we actually feel quite good about the product. I think it, if, I mean, phones don't have a light. I mean, people go around like recording people all the time. 63:10 在这方面，我认为眼镜比人们使用的其他技术要好得多。 The glasses are, I think, way better on that front than the other types of technology that people use. 63:16 我的看法是，几年前我们推出眼镜时，把我们采取的措施讲得相当清楚。 My take on this is that when we launched the glasses a few years back, we communicated pretty clearly about the steps that we put into it. 63:23 但就像你说的，现在有了眼镜的人已经是数以千万计，远比刚发布时多得多。 But like you said, there's now like many, many, many millions of people who have gotten the glasses more than, you know, had them when we just started launching the product. 63:32 所以我觉得我们当初做的那些沟通，很多人要么—— So I think some of that communication that we did at the beginning, a lot of people either, you know- 63:39 没看到… Didn't see ... 63:39 要么忘了，要么一开始就没看到，要么就是没在意，因为当时眼镜还不算什么大东西。现在我觉得我们已经达到了相当主流的程度，至少，我们得重新去确保把我们在做的事情讲清楚， maybe they forgot about- Yeah ... or, you know, they didn't see at the beginning, or they just weren't paying attention to it 'cause the glasses weren't a big thing. And now that I think we've achieved a level of mainstream, well, at a minimum, I think we need to kind of go and make sure that we communicate about what we're doing, 63:55 讲清楚我们确实认为这很重要，重要到从多年前出货的第一个版本起，就把它设计进了产品里。 and that, yes, we think that this is important, and in fact, so important that we designed it into the product from the very first version that we, that we shipped multiple years ago. 64:04 我觉得我们只需要确保人们明白，这一点是如何从根本上融入产品的。 And I, I think we, I think we just need to make sure that people understand how, like how fundamentally that's built into the product. 64:10 但我觉得我们在这方面有点松懈了，只顾着强调，好吧，这是很好看的眼镜。 But, but I think we let up on that a little bit and just kinda focused on, okay, they're great looking glasses. 64:16 有各种款式设计。我们更多地把重点放在这上面，因为我们觉得那些担忧早期就解决了，之后就一直专注于提升—— Um, you know, there's, there's all these designs. I mean, that's kinda been more of the focus on like we, we felt like we kind of addressed the, that, those set of concerns early on, and then since then have just been increasing- 64:27 提升价值、实用性和设计，但必须把这清楚传达。 the value and the utility of them and the designs. But I think we need to make sure that we communicate this piece really clearly. 64:32 但这事我们从一开始就关心，立场很好。人们确实在乎。 But it's something we've cared about from the beginning, and I think we're in a good, we're in a good position on it. But I mean, look, people care about this stuff. 64:38 很重要。 So it's important. 64:38 手机早年也有过隐私恐慌，对吧？新产品一旦证明了价值，人们就会习惯。 And there was a privacy scare with phones in the early days, right? And I think like any new product, once it proves that it's valuable in people's lives, you know, people will get used to it. 64:47 我觉得眼镜还在早期，它是新事物，人们得先看到价值，才能跨过心理门槛，接受一个可能录到自己的东西。 And, and I think maybe glasses are just early in that, in the sense of like it's a new product and people need to see the value for them to get over this mental hurdle of like a new thing that could potentially record me. 64:57 当年用手机的时候就是这样。我记得以前人们会问：“你老拿着手机干嘛？” 'Cause that was with phones. I mean, back in the day, I remember people were like, "What are you doing with your phone?" Like- 65:03 我不知道。 Yeah. I don't know. 65:03 对，我觉得确实有这种情况。 Yeah. I think there's something like that that's true. 65:07 我做了二十年社交媒体，反思之一就是：面对那些担忧，我们当初的回应可能不够直接。虽然这并没有阻止人们使用产品。 I mean, I guess one of my reflections from building social media over the last twenty years is that I don't think we were as direct as we probably should have been about addressing some of those concerns, and it, it didn't necessarily stop people from using the products. 65:21 但这影响了人们今天的看法。其实一路上我们本可以解释清楚，我们有多重视这些问题，但我们没做，因为我们想，大家既然这么爱用产品， But I think it colors how people think about them today. And I think it would've been possible to have kind of explained along the way how seriously we took those issues, and we just didn't because we thought, okay, well, people are showing that by using the product so much, 65:38 就说明还是喜欢它的。但我觉得本可以做得比现在更好—— they like our-- they still like the product. But I actually think it's possible to get to a better state than where we've gotten- 65:44 让社交媒体产品既受人喜欢，又让人明白我们有多重视那些问题。 with, with, um, with the social media products, which is both people liking it and understanding how seriously we take all of those issues. 65:52 这就是我追求的目标。 So that's what I aspire to. 65:53 嗯，这和最近的青少年安全和解一脉相承吗？ Mm-hmm. Is there a through line from that to the recent settlement, uh, on all the youth safety stuff? 66:00 我知道很多人都在讨论。你刚说的和那件事有关联吗？ I know it's a thing a lot of people are talking about. Um, is there any connection from what you just said to that? 66:05 我很好奇你怎么回顾这件事，从中学到了什么，或者—— And like I, I would just be curious to hear you reflect on that and like what you've learned from this process or if- 66:10 我觉得这又是一个很好的例子。我们一直都很重视这些安全问题，在 Instagram 上做青少年账号也花了很长时间，我认为我们做了些领先的工作。 No, I think it's an- it's another good example like this. I mean, we've taken a lot of those safety issues seriously for a while, and we've been working on this teen account work with, with Instagram for, for a long time, and I think have, have done some leading work there. 66:22 这次和解很有意思，因为我们真正想做的是为整个行业建立一套标准和框架。有个现实问题：我觉得大多数做这类产品的公司， The, the settlement there is interesting because what we're really trying to do is create a standard and framework for the industry, and there's this real issue, which is that I actually think most of the companies that are building this, these products, you know, if you basically said, 66:37 如果你说，让青少年每天最多用一小时，大家都会同意——前提是不能只有一家单独做。 you know, limit usage to an hour a day for teens, you know, everyone, I think, would be okay with that except if like you have to unilaterally do it. 66:47 不然的话，大家每天不用 Instagram 超过一小时，转头却去刷 TikTok，那我们真帮到谁了吗？ Then you're saying, okay, like if, if people don't use Instagram for more than an hour a day, but then their usage goes to TikTok, have we really like helped anyone? 66:54 我们只是伤了自己，却没帮到任何人。 And if we, and we've just like hurt ourselves to not help anyone. 66:57 你们在协议里也写了这一点—— And you guys have that in the piece that like- 66:59 基本上是这样。我们的做法是：我们率先单方面限制使用，包括时长限制，还有通知和可使用时段的规定， So basically- Yeah ... the structure for what we did was we basically said, "We're going to take the step of unilaterally limiting The usage of, of, of the-- there's, there's some things around time limits, um, there's some things around notifications and time when people can access it, 67:14 比如上学时、该睡觉时，都有不同的限制。我们说：我们先迈出第一步，等 YouTube 和 TikTok 接受同样条款，整个行业就能锁定标准，一起迈下一步。 when they're in school, when they should be sleeping, like different restrictions. And we basically said, "We will take the first step, and when YouTube and TikTok sign on to the same terms, then we can all, as an industry, lock in and take the next step together." 67:30 所以我非常希望，这次和解能成为一套有法律约束力的框架，让整个行业在这些问题上达成一致，不让任何一家公司因为先行一步而吃亏。 So hopefully, I'm very hopeful that this settlement will serve as a sort of legally binding framework to bring the whole industry into alignment on some of these things and make it so that it doesn't kind of disadvantage any one company for taking that step. 67:49 我们率先行动，确实有点冒险，但我觉得如果其他公司也加入，对大家都更好。 Now, I mean, we're, we're basically putting ourselves a little bit out there by going first, but, um, but I, I think it will be better for everyone if these other companies come in too. 67:59 最后一问。你又在X上发帖了。 Last question. You're posting on X again. 68:02 是的。 I am. 68:02 你发了… Uh, you did a little- Well, 68:03 无处不在 I'm everywhere. 68:04 无处不在。不过很好奇，你在那儿发帖是出于什么想法，比如炫耀—— You're everywhere. But just curious to know, like, you're thinking of like posting there, like the bragging- 68:10 只是一环？ Like is it just part of the- 68:11 我是说 I mean- ... 68:11 现在策略一环？你有Threads也在用。 part of the thing now? Like 'cause you've got Threads, you're on Threads. 68:14 对，我在用Threads。 Yeah. I mean, I'm on Threads. 68:15 我觉得 Threads 现在发展得很好。 I think, I mean, obviously Threads is doing great. 68:17 我觉得它现在要么已经超过了 X，要么 And I think it's actually, it either, I, I think it's either bigger than X at this point or- 68:22 很快就快要超过了。不过，不同平台上有不同的社区。 it's like very soon about to be. But I mean, look, there, there, there are different communities in the different places. 68:26 很多 AI 圈的人都在 X 上。做社交媒体，某种程度上就是要去人们所在的地方交流，对吧？ There's a lot of AI folks are, are on X. And, um, I think part of what you try to do with social media is just communicate where people are, right? 68:35 就像你做播客或者发内容，你大概 It's kind of like when you do a podcast or when you post, you probably- 68:38 到处发 I put it everywhere ... 68:38 不只发一处。 don't just post in one place. 68:38 到处都发。所以我在 Threads 和 X 上发同样的内容， You put it everywhere. So like I, I mean, I post the same thing as on Threads and X- 68:42 有人问为何还发X？我也在那发、在那互动。 and some people are like, "Why are you posting this on X?" It's like, well, and I posted it there too, right? It's like I'll engage there too. 68:48 我觉得都好，但多少有些社区在X，我们想去用户在的地方互动。 So I guess I think it's all good, but I, I do think that the, to some degree, some of the community is on X and we wanna be able to engage where people are. 68:56 这就是这件事的核心：去人们所在的地方，和他们展开对话。 And that's like a lot of what, uh, you know, yeah, what, what, what this is about is going where, where, where people are and being able to kinda have that dialogue. 69:03 谢谢马克 Yeah. Well, thanks Mark. 69:04 谢谢你聊这场 Thanks for having this conversation. 69:05 不客气，很乐意。 Yeah, happy to. 69:08 银行该像现代软件，所需都在一处，上mercury.com了解，几分钟在线申请。 Banking should feel like modern software. Get everything you need in one place. Visit mercury.com to learn more and apply online in minutes. 69:15 Mercury非银行，是金融科技，详见说明；Granola是我用过最佳AI记事本 Mercury is a fintech, not a bank. Check the show notes for details. Granola is the best AI notepad I've tried. 69:21 它在任何场景都能用：视频、电话、面对面，甚至 Apple Watch 上。现在访问 granola.ai/sources 试用，结账时输入优惠码 sources 可享受三个月优惠。 It works everywhere, on a video or phone call, in person, or an Apple Watch. Try it now at granola.ai/sources and use the promo code sources at checkout for three months off. 69:32 Atlassian的Jira让团队与智能体同上下文协作，jira.com免费试用。 Jira by Atlassian is where your team and your agents work from the same context. Try it free at jira.com. 69:38 就是J-I-R-A.com。Framer是AI原生建站，更快又不失控。 That's J-I-R-A.com. Framer is the AI native website builder that lets you build faster without giving up control. 69:46 访问 framer.com/sources 可享七折优惠。具体条款和限制适用。 Visit framer.com/sources for 30% off. Rules and restrictions may apply.
没有匹配的文稿
14:46
15:18
16:10