AI 与工作的未来：与马克·安德森的对谈

如果没有 AI，我们现在就会为经济前景而恐慌。实际上，在人口增长放缓的背景下，我们已经经历了 50 年非常缓慢的技术变革。时机奇迹般地恰到好处。我们将在真正需要 AI 和机器人的时候拥有它们。剩余的人类工人将变得稀缺，而非贬值。

If we didn't have AI, we'd be in a panic right now about what's going to happen to the economy. We've actually been in a regime for 50 years of very slow technological change in the face of declining population growth. The timing has worked out miraculously well. We're going to have AI and robots precisely when we actually need them. The remaining human workers are going to be at a premium, not at a discount.

我们正在经历的这一刻有多重要？

How big of a deal is the moment in time that we are living through right now?

这是一个非常、非常具有历史意义的时刻。AI 就是点金石。现在我们拥有了一项技术，能将世界上最常见的东西——沙子，转化为世界上最稀缺的东西——思想。

This is a very, very historic time. AI is the philosopher's stone. Now we have a technology that transfers the most common thing in the world, which is sand, converted into the most rare thing in the world, which is thought.

我们与最前沿的 AI 创始人进行了大量交流。最领先的创始人正在思考，能否创建由创始人包揽一切的整个公司。

We spent a lot of time with the most cutting-edge AI-forward founders. The most leading-edge founders are thinking of can you have entire companies where the founder does everything.

人们担心年轻人将没有工作可做，AI 正在取代他们。每个人都想谈论失业，但实际上你应该关注的是任务消失。工作的存续时间比单个任务更长。

There's all this concern that young people's jobs are not going to be there for them. AI is replacing them. Everybody wants to talk about job loss, but really what you want to look at is task loss. The job persists longer than the individual tasks.

你对产品经理、工程师和设计师这三个特定角色的未来有何看法？这三个角色之间正发生一场三方对峙。每个程序员现在都认为他们也能成为产品经理和设计师，因为他们有 AI。每个产品经理都认为自己可以成为程序员和设计师。每个设计师都知道自己可以成为产品经理和程序员。实际上，他们都有点道理。结果是，擅长两件事的叠加效应超过两倍，擅长三件事的叠加效应超过三倍。你将成为跨领域组合中的超级相关专家。

What's your sense of just the future of three very specific roles? Product manager, engineer, designer. There's like a Mexican standoff happening between those three roles. Every coder now believes they can also be a product manager and a designer because they have AI. Every product manager thinks they can be a coder and a designer. And then every designer knows they can be a product manager and a coder. They're actually all kind of correct. What happens is the additive effect of being good at two things is more than double. The additive effect of being good at three things is more than triple. You become a super relevant specialist in the combination of the domains.

人们还没有完全理解这变化有多大。在我看来，那些真正想提升自己、发展职业生涯的人，现在应该把每一刻空闲时间都用来与 AI 对话，说：‘好了，训练我吧。’

People aren't fully grasping how much this is changing. And people who really want to improve themselves and develop their careers should be spending every spare hour, in my view, at this point talking to AI, being like, 'All right, train me up.'

今天的嘉宾是马克·安德森，科技和商业领域最具影响力的人物之一。他发明了网页浏览器，建立了全球最大的风险投资公司。他还是一位多次创业者，几乎投资了每一代科技公司，同时也是对技术过去和未来最清晰、横向和富有洞察力的思想家之一。在这场非常特别的对话中，我们聊到了我们正在经历的这一刻有多么独特和重要，他教给孩子们哪些技能以在 AI 未来中茁壮成长，产品经理、设计师和工程师在未来几年会怎样，AI 中的护城河在哪里，最 AI 原生的创始人正在做哪些不同的事情，以及更多只是触及这场深刻而重要对话表面的话题。你将从这次聊天中更聪明地了解当今世界正在发生的事情以及未来的方向。非常感谢我的 newsletter 社区和 X 上的关注者为本对话建议了主题和问题。如果你喜欢这个播客，别忘了在你最喜欢的播客应用或 YouTube 上订阅和关注。这非常有帮助。如果你成为我 newsletter 的内部订阅者，你将免费获得超过 20 款令人难以置信的产品，包括 lovable、replit、bolt、gamma、nad、linear、superhuman、devon、posthog、dscrip、whisperflow、perplexity、warp、granola、magic patterns、raycast、chappd、mob 和 stripe atlas 的一年免费使用权。请访问 lenny'snewsletter.com 并点击产品通行证。在简短赞助商广告之后，有请马克·安德森。

Today my guest is Marc Andreessen, one of the most seminal figures in tech and in business. He invented the web browser, built the world's largest venture firm. He's also a multi-time founder and an investor in essentially every generational tech company and is also one of the most clear-minded, lateral, and insightful thinkers about both the past and the future of technology. In this very special conversation, we chat about how unique and significant the moment that we are all living through right now is, what skills he's teaching his kids to thrive in the AI future, what happens to product managers, designers, and engineers in the coming years, where moats exist in AI, what the most AI-native founders are doing differently, and so much more that is just scratching the surface of this very deep and important conversation. You are going to walk away from this chat being smarter about what is going on in the world right now and where things are heading. A huge thank you to my newsletter community and focus on X for suggesting topics and questions for this conversation. If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. It helps tremendously. And if you become an insider subscriber of my newsletter, you get a year free of over 20 incredible products, including a year free of lovable, replit, bolt, gamma, nad, linear, superhuman, devon, posthog, dscrip, whisperflow, perplexity, warp, granola, magic patterns, raycast, chappd, mob, and stripe atlas. Head on over to lenny'snewsletter.com and click product pass. With that, I bring you Marc Andreessen after a short word from our sponsors.

本期节目由 DX 赞助，DX 是由顶尖研究人员设计的开发者智能平台。要在 AI 时代蓬勃发展，组织需要快速适应。但许多组织领导者难以回答紧迫的问题，比如哪些工具有效？它们如何被使用？什么真正驱动价值？DX 提供了领导者应对这一转变所需的数据和洞察。借助 DX，Dropbox、Booking.com、Adion 和 Intercom 等公司深入了解了 AI 如何为开发者提供价值，以及 AI 对工程生产力的影响。要了解更多，请访问 DX 网站 getdx.com/lenny。即 getdx.com/lenny。如果你是一位创始人，创业最难的部分不是拥有想法，而是在不被后台工作淹没的情况下扩展业务。这就是 Brex 的用武之地。Brex 是为创始人打造的智能金融平台。使用 Brex，你可以获得高额度公司卡、便捷的银行服务、高收益国库券，以及一支为你处理手动财务任务的 AI 代理团队。他们会做你不想做的事情，比如报销费用、扫描交易以发现浪费、生成报告，全部按照你的规则。借助 Brex 的 AI 代理，你可以在保持完全控制的同时更快行动。美国三分之一的初创公司已经在使用 Brex。你也可以在 brex.com 开始使用。

Today's episode is brought to you by DX, the developer intelligence platform designed by leading researchers. To thrive in the AI era, organizations need to adapt quickly. But many organization leaders struggle to answer pressing questions like which tools are working? How are they being used? What's actually driving value? DX provides the data and insights that leaders need to navigate this shift. With DX, companies like Dropbox, Booking.com, Adion, and Intercom get a deep understanding of how AI is providing value to their developers and what impact AI is having on engineering productivity. To learn more, visit DX's website at getdx.com/lenny. That's getdx.com/lenny. If you're a founder, the hardest part of starting a company isn't having the idea. It's scaling the business without getting buried in back office work. That's where Brex comes in. Brex is the intelligent finance platform for founders. With Brex, you get high limit corporate cards, easy banking, high yield treasury, plus a team of AI agents that handle manual finance tasks for you. They'll do all the stuff that you don't want to do, like file your expenses, scour transactions for waste, and run reports, all according to your rules. With Brex's AI agents, you can move faster while staying in full control. One in three startups in the United States already runs on Brex. You can too at brex.com.

马克·安德森，非常感谢你来到这里，欢迎来到播客。

Marc Andreessen, thank you so much for being here and welcome to the podcast.

太好了，Lenny。谢谢你。很高兴来到这里。

Awesome, Lenny. Thank you. It's great to be here.

我想从一个宏观问题开始。我有无数方向想聊，但我认为这能给我们一点参考框架。我们正在经历的这一刻有多重要？

I want to start with just a big picture question. I have a billion directions I want to go, but I think this is going to give us a little bit of a frame of reference. How big of a deal is the moment in time that we are living through right now?

这是一个非常、非常具有历史意义的时刻。我认为 2025 年可能是我整个职业生涯乃至人生中最有趣的一年，而且我预计 2026 年会超越它。

This is a very, very historic time. I think 2025 was maybe the most interesting year in my entire career and probably life, and I think I would expect 2026 to exceed that.

哇，这很说明问题。

Wow, that says a lot.

是的，我见过一些世面。所以，感觉有两件事正在发生。一是许多人对你所说的全球传统机构的信任，我认为现在正在全面崩溃。顺便说一句，有很多数据支持这一点。所以，我认为有很多人们长期依赖的结构、秩序和机构已被证明无法应对挑战。与之相应的是，国家和全球的对话已经变得，可以说，解放了。所以，我们在言论自由、思想自由、人们公开讨论甚至几年前可能无法讨论的事情的能力方面，经历了一场不可思议的革命，并且这种自由正在急剧扩大。我认为这已经踏上了通往更广泛话语的单行道。此外，还有这些极其巨大的地缘政治转变正在发生。

Yeah, I've seen some stuff. So, it feels like two things are happening. One is the trust that a lot of people have had in what you describe as legacy institutions around the world is, I think, in full-scale collapse right now. By the way, there's a lot of data to support that. And so I think there's a lot of structures and orders and institutions that people have just relied on for a long time that have just proven to not be up for the challenge. And then kind of corresponding with that is the national and global conversation have become, let's say, liberated. And so, this sort of incredible revolution that we have in freedom of speech, freedom of thought, ability for people to openly discuss things that maybe they couldn't discuss even a few years ago, is just dramatically expanded. And I think that's now on a one-way train for just a much broader range of discourse. And then, there's also just these incredibly massive geopolitical shifts that are happening.

显然，美国在剧烈变化，欧洲在剧烈变化，中国在剧烈变化，拉丁美洲也在剧烈变化。那里正在上演非常戏剧性的事件，几乎全世界都是如此。我认为很多假设正被摆到台面上重新审视。而且所有这些事情同时发生。所以，你看到所有这些国家和行业都越来越动荡，而 AI 这项新技术又将真正影响一切。同时，公民能够充分参与并辩论这些事情。所以，这三大宏观趋势正在同时碰撞。我认为我们可能才刚刚处于这三者的起点。这些都感觉像是历史性的时刻转变，其规模或许堪比 1989 年柏林墙倒塌，或者二战结束。确实有这种感觉。

And obviously, the US is changing a lot, Europe is changing a lot, China is changing a lot, Latin America, by the way, is changing a lot. Very dramatic events playing out down there right now, kind of all over the world. I think a lot of assumptions are being pulled out into the daylight and re-examined. And then it's the fact that all these things are happening at the same time. So you've got all of these countries and industries where things are increasingly in upheaval, but you have AI as this new technology that's going to really affect things. And then you've got citizens being able to fully participate and argue things out. So it's like those three big mega things are all colliding at the same time. I think we're probably just at the very beginning of all three of those. And those all feel like historical moment shifts, comparable in magnitude to maybe the fall of the Berlin Wall in 1989, maybe the end of World War II. It certainly feels like that.

天哪，真是生逢其时。

Good God, what a time to be alive.

是啊。

Yeah.

就 AI 这块而言，很多人都在试图弄清楚该怎么做。你认为 AI 对世界或听众的影响中，还有什么没有被充分认识到？

In terms of the AI piece, which is where a lot of people are trying to figure out what to do, what do you think isn't being priced in yet in terms of the impact AI is going to have on say the world or just people listening?

我认为，戴上技术帽子来看，很明显这些东西现在真的管用了。三年前的 ChatGPT 时刻，其实才过去三年，对吧？当时的大问题是：这玩意儿极其有趣和富有创意，我们有了能写莎士比亚风格十四行诗和说唱歌词的机器，这太神奇了。但随后有个大问题：你能利用这项技术进行推理，并在真正重要的领域（如医学、科学、法律等）解决问题吗？结果证明答案是肯定的。过去 12 个月，尤其是最近三个月，已经真正证明了 AI 能够进行真正的推理。你现在到处都能看到：AI 正在开发新的数学定理。在假期期间，AI 编程达到了临界质量。世界上最优秀的程序员，包括 Lisbald 等人，第一次基本上说：‘是的，AI 现在编程比我们强。’这极其强大。我认为我们都假设 AI 现在将在任何存在可验证答案的领域变得非常擅长推理，这将包括许多非常重要的领域。所以，技术感觉正在快速前进，而且会运作得非常好。

I think at this point it's pretty clear with our technology hats on that this stuff is really working now. When there was the ChatGPT moment three years ago, it was only three years ago, right? The big question was: this is incredibly fun and creative, we have machines that can compose Shakespeare and sonnets and rap lyrics, that's amazing. But then there was this big question: can you harness this technology for reasoning and problem solving in domains that really matter, like medicine, science, law, and so forth? And it turns out the answer to that is yes. The last 12 months, and especially the last three months, have really proven that AI can do real reasoning. You're seeing it all now: AI is developing new math theorems. Over the holiday break, the AI coding thing really hit critical mass. The world's best programmers, including like Lisbald's, for the first time basically said, 'Yeah, AI is now coding better than we can.' That is incredibly powerful. I think we all assume AI is now going to get really good at reasoning in any domain where there are verifiable answers, and that will include many very important domains. So the technology feels like it's moving fast and it's going to work really well.

我认为没有被充分理解的是，行业中的很多人有一种一维的看法：由于技术奏效，AI 就会横扫世界并改变一切。我认为这是错误的框架，基于对我们过去 80 年所生活的世界的不完整理解。我特别要指出两点。第一，过去 30 或 50 年，在美国和西方，我们感觉像是处在一个技术巨变的时代，但实际上，如果你去寻找统计或分析证据，基本上找不到。经济学家有一种衡量经济中技术变革速度的方法：生产率增长。过去 50 年的生产率增长实际上非常低，而不是很高。我们都感觉它很高，但实际上它一直很低。事实上，在我有生之年，美国的生产率增长速度大约是 1940 年至 1970 年间的一半，大约是 1870 年至 1940 年间的三分之一。所以从统计上看，在美国和西方，经济中的技术进步实际上已经大幅放缓。因此，AI 正在冲击一个实际经济中长期以来几乎没有技术进步的环境。然后还有另一件不可思议的事情正在发生：人口崩溃。这是一种西方现象，并日益全球化，人类繁殖率正在快速下降。许多国家，包括美国，繁殖率低于 2，这意味着世界上许多国家，包括中国（这非常重要），实际上将在下个世纪人口减少。所以，你有一个前提条件：世界上技术进步很少，而且世界将人口减少。AI 将进入一个这两点都成立的世界。这极其重要，因为我们实际上需要 AI 发挥作用来提高生产率增长，而生产率增长是我们需要提高经济增长的。而且我们确实需要 AI 发挥作用，因为我们将需要机器来做所有那些我们没有人去做的工作，因为我们在未来一百年里真的会人口减少。

I think the thing that is not well understood is that a lot of people in the industry have a one-dimensional view: as a result of the technology working, AI just sweeps the world and changes everything. I think that's the wrong frame, based on an incomplete understanding of the world we've been living in for the last 80 years. I would call out two things in particular. One is that it has felt to us in the US and the West for the last 30 or 50 years like we've been in a time of great technological change, but actually if you look for statistical or analytical evidence of that, you basically can't find it. Economists have a way of measuring the rate of technological change in the economy: productivity growth. Productivity growth for the last 50 years has actually been very low, not very high. We all feel like it's been very high, but what's actually happening is it's been very low. In fact, the pace of productivity growth in the US in my lifetime has been running at about half the pace it ran between 1940 and 1970, and about a third the pace it ran between 1870 and 1940. So statistically, in the US and the West, technology progress in the economy has actually slowed way down. So AI is hitting an environment in which we have had almost no technological progress in the actual economy for a very long time. And then there's this other incredible thing happening: the demographic collapse. It's a Western phenomenon, increasingly global, where the rate of reproduction of the human species is in rapid decline. Many countries, including the US, have a reproduction rate under two, meaning many countries around the world, including China, which is a really big deal, are actually going to depopulate over the next century. So you have this precondition: there has been very little technological progress happening in the world, and the world is going to depopulate. AI is going to enter a world where those two things are true. This is incredibly important because we actually need AI to work to get productivity growth up, which is what we need to get economic growth up. And we actually need AI to work because we're going to need machines to do all the jobs that we're not going to have people to do, because we're literally going to depopulate the planet over the next hundred years.

所以我认为这些因素的相互作用将比很多人想象的要有趣得多，坦白说也更复杂。

And so I think the interplay of these factors is going to be much more interesting and frankly more complex than a lot of people have been thinking.

我想顺着孩子这个话题聊下去。我知道你有一个孩子，我特别喜欢通过人们教给孩子什么、引导孩子往哪个方向走，来观察他们的思维方式和价值观。你有没有在引导你的孩子学习某些特定技能，甚至规划职业方向？

I'm going to follow this thread about kids. I know you have a kid and one of my favorite lenses into how people think and what they value is what they're teaching their kids, what they're steering their kids towards. Are there specific skills or even careers that you're steering your kid towards?

我是这样想的……我们有一个 10 岁的孩子，而且我们实际上在家教育，所以我们对此思考很多。我认为思考 AI 对个人影响的方式是……很多人只关注工作得失这种直白甚至过于简单的观点，这个我们可以聊，但在个人和孩子层面，有两件具体的事。首先，很明显 AI 会让擅长做事的人变得非常擅长做事。它会成为一种工具，整体拉高平均水平。这已经在发生了：任何需要写作、设计或写代码的人，如果今天已经很擅长，用了 AI 之后突然就变得非常厉害。教育体系有望基于这一点来教学。但还有另一件事正在发生，尤其是在编程领域：真正优秀的人正在变得极其出色。你可以想到超级赋能个体——那些非常擅长编程、拍电影、做音乐、搞艺术、做播客或风险投资的人——他们能利用 AI 变得极其出色且超级高效。我相信你也有这样的朋友。真正优秀的程序员正在经历这一点：他们不是比以前好两倍，而是好十倍。所以对于单个孩子来说，问题是如何让他们成为超级赋能个体，深入他们所做的任何事情，从而充分利用 AI 变得极其出色。这才是真正的机会，也是我们努力的方向，我会鼓励父母朝这个方向努力。

The way I think about this... we have a 10-year-old and we actually homeschool, so we think a lot about this. I think the way to think about the impact of AI on people as individuals is... a lot of people focus on a straightforward or overly simplistic view of job gains and job losses, which we could talk about, but there are two specific things at the level of an individual person and individual kid. First, it's pretty clear that AI is going to take people who are good at doing things and make them very good at doing things. It's going to be a tool that raises the average across the board. You see that playing out already: anyone who needs to write something, design something, or write code, if they're pretty good at it today, they use AI and all of a sudden they're very good at it. The education system will hopefully teach based on that. But there's this other thing happening, especially in coding: the really great people are becoming spectacularly great. You think of the superempowered individual—someone really good at coding, making movies, songs, art, podcasting, or venture capital—who can harness AI to become spectacularly great and super productive. I'm sure you have friends in this category. Really good coders are experiencing this: they're not twice as good, they're 10 times as good. So at the unit of an individual kid, the question is how to get them in a position where they're a superempowered individual, deep in whatever they do, so they can fully use AI to be spectacularly great. That's the real opportunity, and that's what we're shooting for and what I'd encourage parents to shoot for.

所以我听到的核心就是“能动性”，我们在 Twitter 上经常看到这个词——培养能动性，让他们不等着别人告诉该做什么，而是自己想办法。

So what I heard there is essentially agency, this word that we see on Twitter all the time—building agency, them not waiting for someone to tell them what to do, figuring out what to do.

是的。“能动性”这个词在过去几年变得非常流行，尤其是在加州。这很有意思，因为一开始我很难理解——能动性，他们在说什么？他们指的是主动性、愿意直接去做事。Demo Bird 有一个很棒的说法：活玩家。你可以成为事件的主要参与者。起初我觉得这有点显而易见，但后来我意识到这不再那么明显了，因为我们的社会很大程度上建立在规则之上。每个人都被默认教导要遵守所有规则，如果你打破规则，大家都会抓狂。我们已经陷入一种状态，很多人自然认为你应该训练孩子遵守所有规则。学校系统，从幼儿园到高中，随着时间的推移越来越强调这一点。但不对，你应该——尤其是对于你的孩子——鼓励能动性。这是有道理的。我昨晚刚和 10 岁的孩子聊过。我提出了一个概念：要领导，必须先学会服从；要发号施令，必须先学会听从命令。

Yeah. This term agency has become very popular, certainly in California for the last couple years. It's really interesting because I had a lot of trouble with it early on—I'm like, agency, what are they talking about? What they're talking about is initiative, willingness to just do things. The demo bird has a great term: live player. You can be a primary participant in events. At first I thought that's kind of obvious, but then I realized it's not so obvious anymore because so much of our society is based on rules. Everybody gets taught by default you're supposed to follow all the rules, and if you break the rules, everybody freaks out. We've worked our way into a state where the natural assumption for a lot of people is that you want to train kids to follow all the rules. The school system, K through 12, has gotten more and more focused on that over time. But no, you should actually—especially at the unit of your kid—encourage agency. There's something to be had. I just had this conversation with my 10-year-old last night. I rolled out the concept: in order to lead, you must first learn to obey; in order to issue orders, you must learn how to follow orders.

你知道，我们试图在他的生活中保持一定的结构，而不是纯粹的自主性。但我的意思是，有些规则是重要的，等等。不过，生活中有一个巨大的溢价，那就是成为一个能够完全承担责任、完全掌控局面、运营组织、领导项目、创造新事物的人。也许这在过去 30 年的文化中有所削弱。现在有一个术语重新流行起来，这是健康的。而这就是我对 AI 用于孩子的看法。AI 应该是一个有自主性的孩子撬动世界的终极杠杆，让他能够说：“好吧，我实际上可以成为一个主要的贡献者。”无论是发展物理学的新领域、编写代码、成为艺术家、写小说，无论是什么，我都能充分参与世界，真正改变事物。这个想法与这项技术的结合，对我来说感觉非常健康。

And you know, you kind of try to keep him with some level of structure in his life, not just pure agency. But yeah, I mean, so look, some rules are important and so forth. But yeah, no, look, there is a huge premium in life on being somebody who is able to fully take responsibility for things, fully take charge, run an organization, lead a project, create something new. And maybe that has been a little bit diminished in our culture over the last 30 years. It's healthy that there's now a term for that coming back into vogue. And again, that's how I view AI for kids. AI should be the ultimate lever on the world for a kid with agency, to be able to say, 'Okay, I can actually be a primary contributor.' Whether that's developing new areas of physics, writing code, being an artist, writing novels, whatever that thing is, I can fully participate in the world, I can really change things. The combination of that idea with this technology feels very healthy to me.

那句名言是什么来着？给我一个杠杆，我就能撬动世界。

What is that quote about? Give me a lever and I'll move the world.

我就能撬动世界。是的，完全正确。嗯，你提到这个很有趣。早期的科学家，包括艾萨克·牛顿，都超级痴迷于炼金术这个概念。他发展了牛顿物理学和微积分，但他真正痴迷的是炼金术，而他永远无法让它成功。炼金术是把铅变成金，把非常普通的东西变成非常稀有和珍贵的东西。他花了几十年试图找到贤者之石，那个能把铅变成金的机器或过程，但他从未成功。没有人成功过。现在，有了 AI，我们实际上拥有了一种将沙子转化为思想的技术。

And I'll move the world. Yeah, that's exactly right. Well, it's actually funny you mentioned that. The early scientists, including Isaac Newton, were super obsessed with this concept of alchemy. He developed Newtonian physics and calculus, but the thing he was really obsessed with was alchemy, which he could never get to work. Alchemy was the transmutation of lead into gold, turning something very common into something very rare and valuable. He spent decades trying to figure out the philosopher's stone, the machine or process that would transmute lead into gold, and he never figured it out. Nobody ever figured that out. Now, with AI, we literally have a technology that transforms sand into thought.

这让我大开眼界。

Just blew my mind.

没错，世界上最常见的东西——沙子，被转化为世界上最稀有的东西——思想。所以 AI 就是贤者之石。它确实是。它只是一个极其强大的工具。这就是我如此兴奋的地方。再次，这就是我们对我们 10 岁孩子所做的。我们要做的一件主要事情是确保他知道如何利用并从贤者之石（即 AI）中获益。这是我们教他的一切的核心。有一个流传的梗说硅谷人不让他们的孩子用电脑。可能有一小部分人这样，但我认为恰恰相反。你在硅谷越深入，就越要确保你的孩子完全理解并知道如何使用它。这当然是我们所处的模式，也是我会鼓励父母思考的模式。

Right, the most common thing in the world, sand, converted into the most rare thing in the world, thought. And so AI is the philosopher's stone. It actually is that. It's just this incredibly powerful tool. And that's where I get so excited. Again, this is what we're doing with our 10-year-old. A primary thing we want to do is make sure he knows how to leverage and get benefit out of the philosopher's stone, which is AI. That's central to everything we're teaching him. There's this meme going around that Silicon Valley people don't let their kids use computers. There may be a handful of people like that, but I think it's more the other way around. The more you're plugged into stuff in Silicon Valley, the more important it is to make sure your kids fully understand this and know how to use it. That's certainly the mode we're in, and that's the mode I would encourage parents to think about.

我不知道你的孩子在家上学。这太有趣了。这几乎是对当今教育的一种表态。也许你有什么想法吗？对于那些可能不在你那个收入阶层、但想帮助孩子成功的人，也许在家上学，也许不是。你有什么建议？

I did not know your kid was homeschooled. That is super interesting. It's almost a statement on education in today's day. Maybe is there any thoughts there? For folks that maybe aren't in your tax bracket that want to help their kids be successful, maybe homeschooled, maybe not. What advice would you have?

这是一个挑战，它回到了你最初关于教育的问题。有两种完全不同的思考教育的方式。通常的思考方式是在国家层面，在美国可能是州层面，即如何教育所有孩子？这非常重要，你需要像全国 K-12 学校系统这样的大规模系统。但还有另一个问题，对于单个孩子（n=1），你能做什么？我给你这个问题的终极答案。几个世纪以来，人们就知道，在 n=1 的单位上教育孩子的最佳方式是一对一辅导。如果你有一个孩子，目标是最大化这个孩子，那么一对一辅导能带来最好的结果。历史上每个皇室都知道这一点，每个贵族阶层都知道。有很多惊人的例子：亚历山大大帝由亚里士多德辅导，他征服了世界。几个世纪以来，许多伟大的国王、王后、皇室和贵族都一直采用这种方法。也有统计和分析证据表明这是正确的。

This is the challenge, and it goes to your original question about education. There are two completely different ways to think about education. The way it's usually thought about is at the level of a nation, a national level issue or maybe a state level issue in the US, which is basically how do you educate all the kids? That's incredibly important, and you're going to need some large-scale system like the national K-12 school system. But then there's this other question at n equals 1 for an individual kid: what can you do with an individual kid? I'll give you the ultimate answer to that question. It's been known for centuries that the ideal way to teach a kid at the unit of n equals 1 is one-on-one tutoring. If you have an individual kid and the goal is to maximize that kid, by far you get the best results with one-on-one tutoring. Every royal family in history knew this, every aristocratic class knew this. There are amazing examples: Alexander the Great was tutored by Aristotle, and he took over the world. Many great kings, queens, royal families, and aristocrats over centuries always had this approach. There's also statistical and analytical evidence that this is correct.

教育领域有一个重大问题：如何提高教育成果？事实证明这非常困难，只有一种方法始终有效——Bloom 两西格玛效应。一对一辅导能常规性地将学生成绩提升两个标准差，把一个孩子从第 50 百分位带到第 99 百分位。有辅导老师时，孩子处于一个紧密的反馈循环中，始终处在能力的前沿，获得实时纠正。但除了最富有的人，这从来都不经济可行。AI 提供了实现这一点的真实前景。一个对某件事特别感兴趣的孩子可以和 LLM 交谈，问无数问题，获得即时反馈。你甚至可以告诉它简化解释或考考你。人们今天就能做到。所以对于家长来说，有一个巨大的机会来用 AI 辅导补充传统教育。会有大量初创公司，可汗学院也在大力推动。我认为大方向是混合模式：学校加一对一 AI 辅导。有一所新私立学校叫 Alpha，其理念就是实体学校加大量 AI 辅导。我认为这里有一个神奇配方，会广泛适用。对此感兴趣的家长，现在正是认真思考并考察选项的好时机。

There's this massive question in the field of education: how do you improve educational outcomes? It turns out it's very hard, except for one method that always works—the Bloom two sigma effect. One-on-one tutoring routinely raises student outcomes by two standard deviations, taking a kid from the 50th percentile to the 99th. With a tutor, the kid is in a tight loop, constantly on the leading edge of what they can do, getting real-time correction. But it's never been economically feasible for anyone except the richest. AI provides the real prospect of doing that. A kid super interested in something can talk to an LLM, ask infinite questions, get instantaneous feedback. You can even tell it to dumb it down or quiz you. People can do this today. So there's a massive opportunity for parents to augment traditional education with AI tutoring. There will be tons of startups, and Khan Academy has a big push. I think the broad answer is a hybrid approach: schools plus one-on-one AI tutoring. There's a new private school called Alpha based on this philosophy—in-person schools plus heavy AI tutoring. I think there's a magic formula here that will apply broadly. For parents interested, now is a great time to think hard about it and look at options.

这很有意思，因为有人担心年轻人的工作会消失，AI 正在取代他们。但另一方面，你描述的情况表明今天的学习者会进步很快、学得更多。你在这个分歧上怎么看——年轻人是麻烦大了，还是最终会赢？

It's interesting because there's concern that young people's jobs won't be there, AI is replacing them. On the flip side, what you describe suggests learners today will move fast and learn more. Where do you sit on this divide—are young people in big trouble or will they win in the end?

工作替代、失业的说法非常简化——过于简单的模型。我们经历了 50 年非常缓慢的技术变革，速度是前一时期的一半，是 100 年前的三分之一。经济中几乎没有技术进步，因此工作更替也极少。即使 AI 将生产率增长提高两倍，也只会让我们回到 1870 到 1930 年间的工作更替水平。那时人们认为世界充满机遇。孩子们在新领域发展新职业，创造新产品和服务。即使 AI 将经济变革速度提高两倍，也会转化为更高的经济增长和就业增长。一些任务层面的替代会发生，但会被增长和创新的宏观效应淹没，到处都会出现招聘热潮。此外，这一切发生在人口增长放缓、人口萎缩的背景下。未来 10-30 年，许多国家的人力工作者将因人口萎缩而变得稀缺。将人口下降与移民减少（由于民族主义抬头和对移民率的担忧）结合起来，剩下的人力工作者将更受重视，而非贬值。因此，更快的生产率增长、更快的经济增长、更慢的人口增长和更少的移民，意味着那种反乌托邦式的无工作场景会少得多。我认为它很可能完全被超越了。

The job substitution, job loss thing is very reductive—an overly simplistic model. We've been in a regime of very slow technological change for 50 years, at half the rate of the previous era and a third the rate of 100 years ago. We've had almost no technological progress in the economy and remarkably little job turnover. Even if AI triples productivity growth, it would take us back to the same level of job turnover as between 1870 and 1930. Back then, people thought the world was awash with opportunity. Kids developed new careers in new areas, building new products and services. Even if AI triples the pace of economic change, it will translate to higher economic growth and higher job growth. Some task-level substitution will happen, but it will be swamped by macro effects of growth and innovation, leading to hiring blooms everywhere. Also, this is happening with declining population growth and increasing population shrinkage. Human workers in many countries over the next 10-30 years will be at a premium because of shrinking populations. Combine declining population with less immigration—due to rising nationalism and concerns about immigration rates—and remaining human workers will be at a premium, not a discount. So the combination of faster productivity growth, faster economic growth, slower population growth, and less immigration means much less of this dystopian no-jobs scenario. I think it's probably totally outpaced.

所以，我听到的是你并不太担心失业问题。关键在于时机刚好合适，人口减少，所有这些因素都必须协调一致，才能避免 AI 导致大规模失业，对吗？

So, what I'm hearing is you're not super worried about job loss. Is the key here that the timing kind of just works out, this population decrease, you know, like all these kind of have to line up for there not to be this massive job loss with AI?

是的。你看，如果我们没有 AI，我们现在就会对经济前景感到恐慌。对吧？因为我们面临的将是人口减少的未来，而人口减少如果没有新技术，就意味着经济萎缩。所以经济本身会随时间收缩。机会减少，没有新工作，没有新领域，也没有新的消费需求来源。因此，你会非常担心进入一个严重衰退或停滞的时期。本质上，你会看到一些非常反乌托邦的情景，经济在慢慢自我消亡。所以你会担心与大家以为他们担心的相反的事情。我们之所以不担心那个，唯一的原因是我们现在知道我们有技术可以替代人口增长不足，以及可能出现的移民不足。所以我认为时机奇迹般地恰到好处，我们将在真正需要 AI 和机器人的时候拥有它们，以防止经济萎缩。我认为这从根本上来说是一个好消息。

Yeah. Well, look, if we didn't have AI, we'd be in a panic right now about what's going to happen to the economy. Right? Because what we'd be staring at is a future of depopulation, and depopulation without new technology would just mean that the economy shrinks. So it would mean that the economy itself shrinks over time. The opportunity diminishes. There are no new jobs. There are no new fields. There's no new source of consumer demand for spending on things. And so you would be very worried about going into a period of severe decline or stagnation. Essentially, you'd be looking at very dystopian scenarios of an economy kind of self-euthanizing itself over time. And so you'd be very worried about the opposite of what everybody thinks they're worried about. The only reason we're not worried about that is because we now know that we have the technology that can substitute for the lack of population growth and also for the lack of immigration that's likely. So I would say the timing has worked out miraculously well in the sense that we're going to have AI and robots precisely when we actually need them to keep the economy from shrinking. And I just think that's fundamentally a good news story.

要说到人们担心的大规模失业问题，从另一方面来看，你必须考虑高得多的生产率增长率。你必须考虑每年 10%、20%、30%、50% 的生产率增长率，这比地球历史上任何经济体的生产率增长率都要高出几个数量级。我们有可能达到那种水平。我和其他人一样，也有乌托邦式的幻想。如果 AI 在一夜之间彻底改变一切，那么让我们来设想一下乌托邦的情景。你会得到更高的生产率增长，更高的技术变革水平。相应地，你会经历一场巨大的经济繁荣，经济大幅增长，然后随之而来的是价格崩溃。因此，受 AI 影响或被 AI 商品化的商品和服务的价格会崩溃。会出现价格通缩。由于价格通缩，人们今天购买的所有东西都会变得便宜得多，这相当于全社会财富的巨大增长。这实际上值得讨论，因为人们在这个问题上容易跑偏。如果 AI 会像乌托邦主义者或反乌托邦主义者认为的那样彻底改变经济，那么必然的经济计算结果是巨大的生产率增长。生产率大幅增长，从机械意义上讲，意味着更少的投入带来更多的产出。所以，你用更少的投入获得更多的经济产出。你用 AI 替代人类工人或其他。结果，你会在所有受影响的部门获得巨大的产出繁荣。这些过剩导致价格崩溃。价格崩溃意味着今天花 100 美元的东西现在只花 10 美元，然后只花 1 美元。这相当于给每个人大幅加薪，因为他们现在有了额外的购买力。这种额外的购买力又会转化为经济增长和新领域的发展。每个人的物质生活都会迅速变得更好。而且，顺便说一句，如果由此产生失业，那么提供社会保障网以防止人们陷入贫困的成本会低得多，因为福利计划需要支付的所有商品和服务的价格都在崩溃。医疗价格崩溃，住房价格崩溃，教育价格崩溃，其他一切价格都因为 AI 的巨大影响而崩溃。因此，在人们设想的这种乌托邦或反乌托邦情景中，不存在所有人都贫穷的情况。事实上，恰恰相反：每个人都变得更富有，因为价格崩溃，然后为那些因某种原因找不到工作的人支付社会保障网实际上更容易了。所以，也许我们最终会进入那种情景。我乐观的一面说，是的，也许 AI 真的那么强大，也许经济的其他部分真的可以改变以适应它，也许那会发生。但结果将是一个比人们想象的要好得多的消息。再说一次，我刚才描述的一切只是对非常基础的经济学的直接外推。我没有做任何大胆的预测。这只是一个直接的机械过程，如果你有更高的生产率增长率——这必然是更高技术增长率的结果——它就会自行展开。所以，明确地说，我认为我们面对的世界不会像乌托邦主义者或反乌托邦主义者认为的那样彻底改变。我认为它会更加渐进，原因我们可以讨论。但我认为这个渐进的过程将是一个好消息的过程。即使它快得多，它也会是一个好消息的过程。它只是以我描述的那种方式成为一个好消息的过程。

To get to the mass job loss thing that people are worried about, on the other side of things, you'd have to look at far higher rates of productivity growth. You'd have to look at rates of productivity growth that are 10, 20, 30, 50% a year, something like that, which are orders of magnitude higher than we've ever had in any economy in the history of the planet. It's possible that we get that. I have my utopian temptation along with everybody else. If AI radically transforms everything overnight, then let's play out the utopian scenario. You get a much higher level of productivity growth. You get a much higher level of technological change. Corresponding to that, you'll have a massive economic boom. You'll have massive growth in the economy, and then corresponding with that, you'll have a collapse in prices. So the price of goods and services that are affected by or commoditized by AI, the prices of those goods and services will collapse. There'll be price deflation. And as a consequence of price deflation, everything that people are buying today gets a lot cheaper, and that's the equivalent of a gigantic increase in wealth right across the society. This is actually worth talking about because people get sideways on this issue. If AI is going to transform the economy as much as the utopians or dystopians think it will, the necessary economic calculation of what happens is massive productivity growth. The consequence of massive productivity growth, what that literally means mechanically, is more output requiring less input. So you get more economic output for less input. You're substituting AI for human workers or whatever. And as a consequence, you get this massive boom in output with much lower input costs. The result of that is you get lots of goods and services in all those affected sectors. The result of those gluts is you get collapsing prices. The collapsing prices mean that the thing today that cost you $100 now costs you $10 and now costs you $1. That's the equivalent of giving everybody a giant raise, because now they have all this additional spending power. That additional spending power then translates to economic growth, the development of new fields. Everybody is materially much better off very quickly. And then by the way, to the extent that you do have unemployment coming out the other side of that, it's now much cheaper to provide the kind of social safety net to prevent people from being impoverished, because the prices of all the goods and services that a welfare program has to pay for are all collapsing. The price of healthcare collapses, the price of housing collapses, the price of education collapses, the price of everything else collapses because of this incredible impact that AI is having. And so in this kind of utopian or dystopian scenario that people have, there's no scenario in which everybody's just poor. In fact, it's quite the opposite: everybody gets a lot richer because prices collapse, and then it's actually much easier to pay for the social safety net for the people who for some reason can't find a job. So maybe we end up in that scenario. The optimistic part of me says, yeah, maybe AI is that powerful and maybe the rest of the economy can actually change to accommodate that, and maybe that'll happen. But the result of that is going to be a much better news story than people think it's going to be. And again, everything I've just described is just a very straightforward extrapolation on very basic economics. I'm not making any bold predictions. This is just a straightforward mechanical process that plays itself out if you have higher rates of productivity growth, which are necessarily the results of higher rates of technological growth. And so, to be clear, I think we're looking at a world that's not like radically transformed the way that maybe the utopians think it will be or the dystopians think it will be. I think it'll be more incremental for reasons we can discuss. But I think that incremental process is overwhelmingly going to be a good news process. And then even if it's much faster, it's also going to be a good news process. It'll just be a good news process in the other way that I described.

我喜欢听到乐观和好消息。我还要补充一点，你一直——我在这次聊天之前研究过你，你对世界走向的预测很多次都是正确的。这就是为什么我特别兴奋能和你交谈。我给你列一个简短的清单。我想还有很多其他的事情。

I love hearing optimism and good news. I will also add that you've been — I was researching you ahead of this chat and you've been right so many times about where the world is heading. That's why I'm especially excited to talk to you. I'll give you a short list. I imagine there are many more things.

好。那么，第一，你关于网络和网页浏览器变得重要的判断是对的。你说软件吞噬世界，也对。2011 年你说 10 年内会有 50 亿人用智能手机，实际数字是 60 亿。我还看到你和 Peter Thiel 的辩论，你们争论技术是否停止进步，还是新技术会继续涌现。你主张进步会持续，他说“不，我觉得酷技术已经到头了”。你对了。我想你还有很多预测是对的。所以我喜欢听你的预测，因为我觉得它们最终会成真。

Uh okay. Okay. So, one, you were right about the web and web browsers becoming important. You were right about software eating the world. Check. You uh in 2011, you said that in 10 years we're going to have 5 billion people using smartphones. And I believe the actual number ended up being six billion. You also you had this debate with Peter Teal that I came across where you were debating whether technologies stop progressing or if new technology will continue to emerge. and you were arguing there is progress. Progress will continue. And he he was like, "No, I think we're done with cool technology." You were right. Uh imagine there are many more things you were right about. So, so again, I'm just I I love hearing your predictions because I feel like they're actually going to turn out to be correct.

首先我得说，我错的事情也很多，但你知道，我都把它们埋在后院了。从互联网上删掉，没有浏览器能找到。对，我把它们从互联网档案里抹掉了，再也看不到了。嗯，所以我也有很多错的时候。但，是的，有些我确实对了。顺便说一句，关于 Peter 那件事，我现在更认同他的观点了。今天再争论那个问题，我会很不一样，我会更尊重他的看法。这其实跟我们刚才的讨论有关——Peter 真正在说的是，我们在比特领域有很多进步，但在原子领域进步很少。这才是他论点的核心。我当时可能有点忽略或轻描淡写了，因为我太专注于让人们明白比特领域确实还有进步。但他对原子领域缺乏进步的批评是真实的，这又回到他长期谈论的一点：过去 50 年，经济的大部分领域几乎没有技术创新，尤其是涉及原子的东西。现实世界的技术变化很少，建成环境跟 50 年前没太大不同。对比一下，1870 到 1930 年世界天翻地覆，1930 到 1970 年也是，但 1970 到今天呢？没那么大变化。你随便走走，看到一堆 1960 年建的楼，1930 年建的桥，1910 年建的水坝，1880 年建的城市，我们做了什么？新城市在哪？新水坝在哪？加州高铁呢？怎么回事？所以我觉得他很多地方是对的。这也是为什么我认为 AI 不会产生那么快的影响——不会像乌托邦或反乌托邦那样一夜之间改变一切。我认为这不可能，因为 Peter 指出的原因：世界运作的很多方面都被官僚程序、规则、限制所束缚。还有政治、工会、卡特尔、垄断——所有这些经济、政治或监管结构基本上阻止了变化。举个例子，AI 对医疗系统的影响。按理说 AI 会对医疗系统产生巨大且积极的影响，但今天医疗系统很大部分是卡特尔。医生是卡特尔，护士是卡特尔，医院是卡特尔，还有推动医疗系统国有化的力量，那就成了政府垄断。卡特尔和垄断不喜欢什么？它们不喜欢快速变化。所以一个年轻人带着 AI 医疗技术来了，他们问：“这会威胁医生的饭碗吗？”如果是，那就封杀。另外，很多消费者，我在生活中看到，你可能也看到了：ChatGPT 几乎肯定比你的医生更好，但它拿不到行医执照，所以不能替代医生，不能开药，不能做手术。所以，Peter 一直很清晰地指出，我们的经济和政治体系中存在真正的结构性障碍，阻止变化达到过去那样的速度。乐观地说，也许 AI 这种新魔法技术的出现，会让我们几十年来第一次重新审视这些假设，真正问自己：这是我们想要的世界吗？我们难道不想更快地到达未来吗？这可能是乐观的看法。

So, I should start by saying I've been wrong about tons of things, but you know, I buried those out back behind the shed. Delete them from the internet. No web browser can discover them. Yes, I have them nuked out of the internet archives so they they're never seen again. Um, so, uh, you know, I'm wrong plenty of times also. Um, but yeah, I mean, look, I think, yeah, some some of those I got right. By, by by the way, I will say on the on the Peter one, I I have come I've come much more around to Peter's point of view. Um, I would probably argue that one like quite a bit differently today than I did, and I would give his view I think I think a lot more credit. Um, and and it actually goes to kind of the discussion that the kind of conversation we just had, which is the the real form of what Peter was arguing was we have lots of process in bit. We have lots of progress in bits, right? But we have we have very little progress in atoms, right? Um and and that's the real core of what he was arguing. And I think I I I think I I was a little bit I don't know missing that or kind of you know kind of glossing that over a little bit um because I was so focused on making sure people understood no there actually is still progress happening in in bits. But I think you know a lot of his critiques around the lack of progress in Adams is real and and again this goes back to this thing of like in the and he you know he's talked about this for a long time. In the last 50 years there has just been very little technological innovation in most of the economy. there's been very little technological innovation in particular anything involving atoms that you know there's been very little real world technological change there just there just hasn't been like the the the built world is just not that different today than it was 50 years ago and if you and again if you contrast that you know if you if you compare and contrast 1870 to 1930 it was a dramatically different world if you contrast 1930 to 1970 it was a dramatically different world if you contrast 1970 today it's not that different right and look you just see that you could just like walk around and it's just like oh yeah there's a bunch of buildings that were built built in like 1960, right? And there's a bridge that was built in like 1930 and there's a dam that was built in like 1910 and there's a city that was founded in, you know, 1880 and like what have we done, right? Like where are new cities? Where are new dams? Where, you know, where's where's the California highspeed rail? Like you know, you know, like what's going on here? And so like I think he is I I think he is right about a lot of that. Um, again, this is also why I think that AI is not going to have as rapid an imp. It's not going to be again this kind of utopian or dystopian view of like everything changes overnight. I think it just kind of can't happen because of the reasons that Peter articulates which is there's just there's so much about how the world works that's basically just like wrapped up in red tape like bureaucratic process, rules, restrictions. um you know the the the politics um by the way you know unions cartels opolies there there's all these structures in the world that are kind of economic or political or regulatory structures that basically prevent things from changing and so I mean let's take let's take a great example like a AI's impact on the healthare system like by rights AI is going to have a dramatic impact on the healthare system and in and in in very positive ways but you know large parts of the medical system today are they are cartels, right? And so there's like a there's the doctors are a cartel and like nurses are a cartel and like hospitals are a cartel and then there's this push to like nationalize all the healthare systems and then you've got, you know, then you've got a government monopoly, right? And it's like and and and guess what cartels of monopolies don't like is they don't like like rapid change, right? Um and so, you know, you show up as a kid and you're like, "Wow, I've got like this new technology to do like AI medicine." And they're like, "Oh, well, does it threaten Dr.'s jobs?" Well, in that case, we're going to we're going to block it. So, and I think a lot of consumers, by the way, you know, I I I see this in my life and you you'll probably see this in your life also, which is, you know, like Chet GPT is like almost certainly a better doctor than your doctor today, but like Chad GPT can't get a license to practice medicine, right? So, it can't substitute for a doctor. It can't prescribe medications, right? It can't, you know, perform procedures, right? And so there there there are these any anyway so Peter Peter I think was very articulate and has been for a long time on like no there are actually real structural impediments in the economy and in the political system that we have that actually prevent any the rates of change that are anywhere near the rates of change that people had in the past. And and you can maybe say optimistically you know maybe the presence of it of the new of the new magic technology of AI maybe it causes us to revisit a lot of these assumpt assumptions for the first time in decades to really say okay is this really the world we want to live in? Don't we actually want to get to the future faster? So maybe that would be the optimistic view.

“是时候建设了。”有人说过这句名言。我在日历里，开始工作时就写这个：“是时候建设了。”这是我早上的一整块时间。谢谢你。我喜欢你从宏观一下子落到具体。我想聊具体。这个播客的很多听众是产品经理、工程师、设计师。有很多创始人，但也有很多非创始人。很多人在做产品但不是创始人，显然很多人担心自己的职业走向。

It's time to build. Somebody famously said, I uh in my calendar, I actually have that as my when I start to work. It's time to build. That's my block in the morning of the day. Thank you for that. Okay. I love I love the way you go from just like macro to just like end of one. And I want to go to end of one. A lot of the listeners of this podcast are product managers. They're engineers. They're designers. They're not a lot of There's a lot of founders, but there's also a lot of non-founders. There's a lot of people building product that aren't founders and uh obviously a lot of people are worried about where their career is going.

这些角色中，会不会有一个消失？会不会有一个特别吃香？我该怎么跟上变化？你跟很多团队、很多产品团队都很熟。你对产品经理、工程师、设计师这三个具体角色的未来有什么看法？

Is one of these roles going to disappear? Is one of these roles going to do really well? How do I stay up to date? You're close with a lot of teams, a lot of product teams. What's your sense of just the future of these three very specific roles? Product manager, engineer, designer.

我觉得这个问题很有意思。这三个角色显然是科技公司构建产品的核心角色。我一直在用“墨西哥对峙”来形容——就是电影里那种两个人拿枪指着对方脑袋的场景。如果你看过吴宇森的电影，他特别喜欢搞三方对峙，形成一个三角形，每个人双手持枪，互相瞄准。所以现在产品经理、设计师和程序员之间就出现了这种对峙。具体来说：每个程序员现在都觉得自己也能当产品经理和设计师，因为他们有 AI；每个产品经理都觉得自己能当程序员和设计师；每个设计师也觉得自己能当产品经理和程序员。所以每个角色的人都认为，有了 AI，他们不再需要另外两个角色了，因为 AI 可以代劳。而真正的讽刺在于，他们三个最终都会意识到，AI 还能当更好的经理，所以他们会把枪口指向更高的职位——但这可能是下一阶段了。我觉得这个“墨西哥对峙”最迷人的地方在于，他们其实都有点道理：AI 现在确实是个好程序员、好设计师、好产品经理，至少能完成这三个岗位的大部分任务。所以这又回到了“超级赋能个体”的概念。如果我是个程序员，第一步是要真正理解 AI 编程的含义，以及编程未来会如何变化——从完全手写代码，变成协调十几个编程机器人。编程工作本身正在发生改变。但另一方面，我该如何成为那个超级赋能个体？如何成为一个也能利用 AI 成为优秀产品经理和优秀设计师的程序员？产品经理也一样：如何确保自己能使用编程工具，也能做 AI 驱动的设计？设计师也一样：如何利用 AI 成为程序员和产品经理？结果可能是，这些角色不再像过去 30 年那样是孤立的烟囱式岗位。但真正发生的是，任何角色中的有才华的人都会变得超级赋能，擅长做这三件事。然后这些人会变得极其有价值，因为他们能真正从零开始构建和设计新产品——这是最有价值的事情。所以我认为这就是机会所在。

This I think is a really funny question. So these three roles in particular obviously are kind of the central roles for building you know for tech companies. So, the way I've been describing it is, you know, you know the concept of the Mexican standoff, right? Which is the the movie scene where the, you know, the two guys have guns pointing at each other's heads. Um, and then there's, if you watch like John Woo movies, he loves to have he does the three-way Mexican standoff where you've got like a triangle, you know, people like, you know, and of course it's John Woo movie, they've got, you know, guns in both hands. So, they're all each each is aiming at the other two. Yeah. Um, and you got this kind of standoff situation. And so the way I've been describing this is there's like a Mexican standoff happening between those three roles between product manager, designer and coder. Specifically the following which is every coder now believes they can also be a product manager and a designer right because they have AI. Every product manager thinks they can be a coder and a designer. And then every designer knows they can be a product manager, right? And a and a coder, right? And so people in each of those roles now, you know, know or believe that with AI they they don't need the other two roles anymore, right? they they they can do that because they can have AI do that. And then of course and then of course there's the real irony which is you know all the the all three of them are going to realize that AI can also be a better manager, right? So they're going to they're going to end up a aiming the guns up the order chart. But that's probably that's the next phase. And what I think is so fascinating about this Mexican staff is they're actually all kind of correct I think right which is AI is actually a pretty good you know it's now it's actually now a really good coder. it's actually now a really good designer and it's also a really good product manager, right? It's actually good at doing all three of those things or at least doing a lot of the tasks involved in in in those three jobs. And so again, this this goes back to the the the superower this kind of idea of the supermpowered individual. Uh where if if I'm a coder like you know I mean step one is like I need to make sure that I really understand AI coding and like what that means and what how coding is going to change in the future. you know that that I need to you know specifically how to go from being a coder who writes code entirely by hand to being a coder who you know orchestrates you know a dozen instances of of of you know coding bots you know you know there's there's a change in the actual job of coding itself which is which is happening right now but the other part of it is okay how do I become that superpowered individual how how do I become a coder that also then harnesses AI so that I can also be a great product manager and I I can also be a great designer right and then the same thing for the product manager which is how do I make sure that I can now use coding tools how do I make sure I can also, you know, do AI AI based design. And the same thing for the designer, which is how do I use AI to be be also become a coder and also become a product manager. And then what you get is maybe the maybe the those individual roles change like maybe those are not anymore sort of stovepipe roles the way that you know they have been for the last 30 years or whatever. Uh but what happens is the the talented people in any of those roles become superpowered and they become good at doing all three of those things. Um and then and then those people become incredibly valuable because then those are people who can actually like you know build and design right new products right from scratch which is like the you know which is which is the most valuable thing. And so I I think I think that's I think I think that's the opportunity.

我很喜欢这个回答。所以我听到的是，如果你在这三个角色中任何一个做到出色，你都会做得很好。

So I love this answer. So what I'm hearing is essentially uh if you're amazing at any of these three roles you will do well.

第一，如果你在这些角色上很出色，那很好，但出色的一部分也在于能否充分利用新技术。所以如果你今天是个编程大师，但你始终没学会如何用 AI 来放大你的编程技能、做更多事，那么你迟早会遇到问题。经济学家有另一种说法：工作的原子单位不是“岗位”，而是“任务”。岗位是一组任务的集合。大家都爱谈岗位消失，但真正该关注的是任务消失——任务在变化。一个经典例子：过去高管们从不自己用打字机或电脑。1970 年，公司副总裁桌上不会有打字机或电脑，他们靠秘书口述备忘录。后来电子邮件出现了。秘书的工作从寄信变成了与其他行政人员收发邮件。秘书会把邮件打印出来，拿进高管办公室。高管在纸上阅读邮件，手写回复，再交给秘书，秘书回到自己桌上把回复打出来发出去。如今这一切都不存在了。高管们自己处理所有邮件。他们仍然有秘书或行政助理，但任务变了——他们做旅行计划、组织活动等等。讽刺的是，高管的任务集反而扩大了，他们自己做更多文书工作，比如坐在那里自己打备忘录——这在 50 年前是绝不会发生的。但高管的岗位依然存在。

Number one if you're amazing at these roles that's great but also you part of being amazing these roles is also being being able to fully harness the new technology right. So if you're if you're a master coder today and you you don't ever get to the point where you you figure out how to use AI to leverage your coding skills, you and and do more, right? Like at some point you are going to hit an issue, right? Here's another way economists talk about this, which is there's the concept of the job, but the job is not actually the atomic unit of what happens in the workplace. The atomic unit of what happens in the workplace is the task. And so and then what what the way the economists think about it is a job is a bundle of tasks. And everybody wants to talk about job loss, but really what you want to look at is is task task loss, right? Tasks changing. I mean the the the the classic the classic example of task changing. Classic example of task changing was once upon a time executives never used typewriters or personal computers themselves, right? You know, if you were a vice president of a company in 1970 or whatever, you did not have like a typewriter or computer on your desk typing things. You had a secretary who you dictated memos to, right? And then there and then there was this change where like emails started to show up. And what would happen was the job of the secretary then went from, you know, it went from, you know, the the job of the secretary changed from sending out letters with stamps on them to like sending or receiving emails with the other admins. And then and then the secretary would print out the email and bring it into the executive's office. And the executive office would read the email and paper, scroll scroll the reply um and and and give and give that message back to the secretary who would go back and type it into the computer on on on his or her desk and send it as an email. Fast forward to today, none of that happens. Now executives just do all their own email. They still have secretaries or admins, but they're now doing different tasks. You know, they're travel planning and orchestrating events and like doing all these other things, you know, that that you know that the great admins do. And then and then the task the task set ironically of the executive has expanded to do actually more of the clerical work themselves actually like sit there and like type their own memos, which again 50 years ago they never never would have done that. And so the executive job still exists.

秘书的工作仍然存在，但任务已经变了。我认为这是编程领域即将发生的变化的一个很好的例子：任务将会改变。产品管理任务会变，设计师任务也会变。所以，工作的存续时间比单个任务更长，当任务变化足够大时，工作本身才会改变。从个人层面来看，你大概会想：好吧，我有这份工作，它是一组任务的集合。我需要非常擅长确保自己能替换掉这些任务，对吧？我能真正适应，使用新技术，比如在 AI 编程上变得非常擅长。然后你还想增加技能。我也可以变得非常擅长设计，非常擅长产品管理，因为我有了这个新工具。所以，随着你这样做，你想承担越来越多的范围。那么 10 年后，你的职位头衔是程序员，还是程序员兼设计师兼产品经理，或者只是“我构建产品”或“我告诉 AI 如何构建产品”？不管那个工作叫什么，谁知道它会是什么，但它将极其重要，因为做那份工作的人将是在编排 AI。那就是最优秀的人将要走的道路。我认为这是你应该全力投入的事情。

The secretary job still exists, but the tasks have changed. And I think that's a great example of what's going to happen in coding: the tasks are going to change. Product management tasks are going to change, designer tasks are going to change. So the job persists longer than the individual tasks, and then as the tasks change enough, that's when the jobs change. At the level of the individual, you kind of want to think, okay, I have this job; the job is a bundle of tasks. I need to be really good at making sure that I can swap the tasks out, right? I can really adapt, use the new technology, get really good at AI coding, for example. And then you want to add skills. I can also get really good at design, I can also get really good at product management because I've got this new tool. So you want to pick up more and more scope as you do that. And then 10 years from now, is your job title coder, or coder designer product manager, or is it just 'I build products' or 'I tell the AI how to build products'? Whatever that job is called, who even knows what it's going to be, but it's going to be incredibly important because the people doing that job are going to be orchestrating the AI. That's the track that the best people are going to be on. And I think that's the thing to lean hard into.

我认为人们还没有完全理解软件工程这个领域正在发生多大的变化。很明显，我们很快就会进入一个工程师实际上不写代码的世界，这在一年前我们可能还不会想到。但现在这显然是趋势。那种坐在那里手写代码的工匠式体验将会存在，这份工作的变化之大真是令人难以置信。

I think people aren't fully grasping just how much software engineering specifically is changing. It's pretty clear we're going to be in a world soon where engineers are not actually writing code, which I think a year ago we would not have thought. And now it's just clearly where it's heading. There's going to be this artisanal experience of sitting there writing code, which is so crazy how much that job is going to change.

是的。所以我又要回溯历史了——请原谅我讲点历史——但我要回到编程这件事上。你知道“计算员”这个词最初的定义吗？你知道它指的是什么吗？

Yeah. So again, I go back—and pardon the history lesson—but I go back to coding. Do you know the original definition of the term 'calculator'? Do you know what that referred to?

不知道。

No.

是的。所以我又要回溯历史了——请原谅我讲点历史——但我要回到编程这件事上。你知道“计算员”这个词最初的定义吗？你知道它指的是什么吗？不知道。它指的是人。对吧。所以在电子计算器或计算机出现之前，你进行计算的方式——比如保险公司计算精算表，或者军队计算部队后勤——你实际上会有一个满是人的房间。顺便说一句，这些大房间可能有成百上千甚至上万人做这件事。房间前面有一个人负责数学方程，然后他们把单个计算任务分发给坐在桌子前的人，这些人全部手工完成。那些人被称为计算员。所以我们从一个真正由人手工做数学方程的世界，发展到了第一批计算机。第一批计算机没有编程语言；它们只有机器码。所以程序员的任务变成了处理 0 和 1，然后变成了打孔卡。今天仍然有人，他们的程序员工作就是处理打孔卡。然后你得到了一个重大突破，叫做汇编语言，它基本上是一种做机器码的方式，但加入了一定程度的英语。最好的程序员做汇编语言。然后到我成长的时代，是像 C 这样的高级语言，编译成机器码，这就是程序员做的事。我还记得脚本语言——我们在 Netscape 开发了 JavaScript，然后 Python 火了，Perl 等等——但脚本语言在 2000 年代起飞。技术社区有一场大争论：脚本语言算不算真正的编程？因为它有点像作弊，对吧？真正的程序员写编译成机器码的代码，他们自己做内存管理，他们做整个写 C 代码的手艺。而这些 JavaScript 或 Python 程序员在做这种轻量级的东西——它真的算编程吗？当然，答案是肯定的，它非常算，现在大多数编程都是用脚本语言完成的。脚本语言抽象掉了人们过去手工做的五层细节，他们不再需要做了。然后，正如你所说，AI 编程是下一层。AI 编程实际上抽象掉了编写脚本代码的过程。所以从某种意义上说，这是一件大事，原因显而易见，但另一方面，它是程序员工作下任务重新定义的下一层。现在程序员的工作是什么？正如你所说，不一定是手写代码，但现在的情况是：如果你和当今世界上最好的程序员交谈，他们会告诉你，“哦，我的工作是坐在那里编排 10 个并行运行的代码机器人。”他们真的坐在那里，在浏览器之间或终端之间切换，他们现在的工作日是与 AI 机器人争论，试图让它们写出正确的代码，调试它，修复问题，更改规格，做所有这些事情。所以现在程序员的工作是与编程机器人争论。但如果你自己不知道如何写代码，你就不知道如何评估编程机器人给你的东西。所以你问到了我们那个对计算机和编程超级着迷的 10 岁孩子。我告诉他的是——顺便说一句，他热爱编程——是……

It referred to people. Right. So back before there were electronic calculators or computers, the way you would do computing—like an insurance company calculating actuarial tables or the military calculating troop logistics—you would actually have a room full of people. By the way, these big rooms could have hundreds or thousands or tens of thousands of people doing this. You would have somebody at the head of the room who was responsible for the mathematical equation, and then they would parcel out the individual calculations to people sitting at desks who were doing them all by hand. Those people were called calculators. So we've gone from a world where you literally have people doing mathematical equations by hand. Then we got the first computers. The first computers didn't have programming languages; they only had machine code. So the task of the programmer became doing ones and zeros, and then that became punch cards. There are still people today whose job as a programmer was to deal with punch cards. Then you got this big breakthrough called assembly language, which was basically a way to do machine code but with some level of English added. The best programmers did assembly language. Then when I was coming up, it was higher-level languages like C that compiled into machine code, and that's what programmers did. I still remember when scripting languages—we developed JavaScript at Netscape, and then Python took off, Perl, and others—but scripting languages took off in the 2000s. There was this big fight in the technical community: is scripting real programming or not? Because it's kind of cheating, right? Real programmers write code that compiles to machine code, they do memory management themselves, they do this whole craft of writing C code. And these JavaScript or Python programmers are doing this lightweight thing—does it even really count as coding? Of course, the answer is yes, it very much counted, and now most coding is done with scripting languages. Scripting languages have abstracted away five layers of detail underneath that people used to do by hand, and they don't anymore. And then, to your point, AI coding is the next layer. AI coding actually abstracts away the process of actually writing the scripting code. So in one sense, this is a really big deal for all the obvious reasons, but on the other hand, it's the next layer of task redefinition under the job of programmer. Now what's the job of the programmer? To your point, it's not necessarily to write the code by hand, but what it is now is: if you talk to the world's best programmers today, they'll tell you, 'Oh my job is I'm sitting there orchestrating 10 code bots running in parallel.' They literally sit there and shift from browser to browser or terminal to terminal, and their day job now is arguing with the AI bots, trying to get them to write the right code, debug it, fix problems, change the spec, and do all these things. So now the job of the programmer is to argue with the coding bots. But if you don't know how to write the code yourself, you don't know how to evaluate what the coding bots are giving you. So you asked about our 10-year-old who is super into computers and programming. What I'm telling him—and by the way, he loves coding—is...

他整天都在 Replit 上做 vibe coding，做游戏。他坐在那里，基本上就是个 10 岁小孩，晚饭时花两个小时跟 AI 争论着玩。但我告诉他，不，你仍然需要充分理解并学会如何编写和理解代码，因为编码机器人会给你代码。如果它不工作，或者没有达到你的预期，或者不够快，你需要能够理解 AI 给你的结果。就像写脚本语言代码的人最终需要理解微处理器的工作原理一样。所以这又是能力的提升，你实际上需要具备深入理解事物实际运作的能力，即使你并不亲手去做。我再看这个，觉得程序员的生产力将变成以前的 10 倍、100 倍甚至 1000 倍。这绝对是一件好事。任务确实在变，工作的性质也在变。但人类还会参与编码过程并监督 AI 编码吗？答案当然是 100% 会。毫无疑问。

He's on Replit all the time doing vibe coding, you know, doing games. He's sitting there, it's a 10-year-old basically who spends two hours at dinner arguing with an AI for fun. But what I'm telling him is, no, you need to still fully understand and learn how to write and understand code because the coding bots are giving you code. If it doesn't work or if it's not doing what you expect or it's not fast enough or whatever, you need to be able to understand the results of what the AI is giving you. In the same way that somebody who's writing scripting language code does need to understand ultimately how the microprocessor works. So again, it's this upleveling of capability where you actually want the depth to be able to go down and understand what the thing is actually doing even if you're not spending your day actually doing that by hand. And again, I look at that and I'm like, okay, now programmers are going to be 10 times or 100 times or a thousand times more productive than they used to be. And that is overwhelmingly a good thing. The tasks are definitely changing. The nature of the job is changing. But are human beings going to be involved in the coding process and overseeing the AI coding? And the answer is of course absolutely 100%. No question.

所以你属于仍然认为学习编码是一项宝贵技能的那一派。

So you're in the camp of still learning to code, still a valuable skill.

哦，是的，完全同意。再说一次，如果你想成为那种超级……听着，如果你只是想让自己自动驾驶，说“我懒得管，就让 AI 写代码，它生成什么就是什么，那也行”，如果你的目标是成为一个平庸的程序员，那就让 AI 去做吧。AI 非常擅长生成大量平庸的代码。没问题。但如果你的目标是成为世界上最好的软件人员之一，并构建真正重要的新软件产品和技术，那么你 100% 需要深入到底层。你的技能需要一直深入到汇编和机器码。你需要理解每一层堆栈。你需要深入理解芯片层面、网络层面等等发生的事情。顺便说一句，你还需要深入理解 AI 本身的工作原理，因为如果人们理解 AI 的工作原理，他们显然能从中获得更多价值。一个不理解它工作原理的人？我的意思是，如果你知道机器的工作原理，你总是会更高效。当你使用机器时，是的，那些想要用新技术做大事的超能力个体，你 100% 需要理解这个东西的整个堆栈，因为你想要能够理解它给你的东西。当某些东西不工作或不对时，你希望能够非常快速地理解原因。顺便说一句，这又回到了教育。AI 是你学习所有这些的最佳朋友。因为就像，“哦，我需要理解，我不知道，这个不够快。作为一个程序员，我需要弄清楚如何用不同的方法进行内存管理之类的。”然后你可以说，“嗯，我不太知道怎么做。好的，AI，我们花 10 分钟。教我怎么做到这一点。教我这都是什么意思。”所以突然间，你与 AI 建立了这种难以置信的协同关系，它在为你做大量工作的同时，也在帮助你变得更好。

Oh yeah, totally. Well, again, if you want to be one of these super... Look, if you just want to put yourself on autopilot and say 'I can't be bothered, I'm just going to have AI write the code and it's going to generate whatever it does and that's fine,' if the goal is to be a mediocre coder, then just let the AI do it. The AI is going to be perfectly good at generating infinite amounts of mediocre code. No problem. If the goal is to be one of the best software people in the world and build new software products and technologies that really matter, then you 100% want to go all the way down. You want your skill set to go all the way down to assembly and machine code. You want to understand every layer of the stack. You want to deeply understand what's happening at the level of the chip, and the network, and so forth. By the way, you also really deeply want to understand how the AI itself works, because if people understand how the AI works, they are clearly able to get more value out of it. Somebody who doesn't understand how it works? I mean, you're always more productive if you know how the machine works. When you use the machine, yeah, the super-empowered individual on the other end of this that wants to do great things with the new technology, yes, you 100% want to understand this thing all the way down the stack because you want to be able to understand what it's giving you. And when something doesn't work or when something isn't right, you want to be able to really quickly understand why that is. By the way, this goes back to education. AI is your best friend at helping you learn all that. Because it's like, 'Oh, I need to understand, I don't know, this isn't fast enough. I need to figure out as a coder how to do a different approach to memory management or something.' And you can be like, 'Well, I don't quite know how to do that. Okay, AI, let's spend 10 minutes. Teach me how to do this. Teach me what this all means.' So all of a sudden, you have this incredibly synergistic relationship with the AI where it's also helping you get better at the same time as doing a lot of work for you.

顺便说一句，我本来想说我是个 Perl 程序员。我当了 10 年工程师，那是我选择的语言。

By the way, I was going to say I was a big Perl programmer. I was an engineer for 10 years and that was my language of choice.

你还记得吗，我不知道你是什么时候做的，但你还记得至少早期的时候，你有没有遇到过程序员看不起你，说……

You do remember, I don't know when you were doing it, but do you remember at least early on, did you ever hit this where coders were looking down their nose at you being like...

当然，当然。就像“这太慢了，没法扩展，你花这么多时间在这上面干嘛？”是的，没错。

For sure, for sure. It's like 'this is so slow, it's not going to scale, what are you spending all your time on this thing?' Yeah, exactly.

当然，这有点像他们某种程度上是对的，一开始它不够快之类的。但到最后，他们绝对是错的，因为它变得更好、更快，并且席卷了世界。如今大多数编码都是用脚本语言完成的。顺便说一句，那些真正理解脚本语言以及所有底层系统的人，正是他们让脚本语言真正工作得很好。所以这是这种适应性的一个很好的例子。结果就是，用脚本语言写代码的人数远远超过用底层语言写代码的人数。我认为这只会是一个更戏剧性的版本。

And of course, it was sort of this thing where they were sort of correct, which is at the beginning it wasn't fast enough or whatever. By the end, they were definitely wrong, which is it got much better, much faster, and it swept the world. Most coding today happens as scripting languages. And by the way, the people along the way who really understood the scripting languages and the people who understood all the lower level systems, they were the ones who were able to actually make the scripting languages work really well. So that was a great example of this kind of adaptation. And then again, the result of that was a far higher number of people writing code with scripting languages than were ever writing code with lower level languages. And I think this will just be a more dramatic version of that.

我喜欢 Perl 是由一位语言学家设计的。我不知道你是否记得这一点，正是这一点让它用起来如此舒服。

I love that Perl was designed by a linguist. I don't know if you remember that part, and that's what made it so nice to code with.

嗯，这很有趣，因为它以难以理解而臭名昭著。多么讽刺啊。

Well that's funny because of course it was so notorious for being impossible to understand. So how ironic.

是啊。

Yeah.

这一切都由与实时数据关联的功能标志驱动，让你能安全发布、精准定位并持续学习。Data Dog 不仅仅是工程指标，更是优秀产品团队更快学习、更智能修复、自信交付的地方。请访问 dataq.com/lenny 申请演示。回到这个三角关系，我越来越多听到的另一个要素是品味、设计和用户体验的技能。这感觉是一项很难学习的技能，对我来说，这意味着设计在未来会更有价值。

And all of this is powered by feature flags that are tied to real-time data so that you can roll out safely, target precisely, and learn continuously. Data Dog is more than engineering metrics. It's where great product teams learn faster, fix smarter, and ship with confidence. Request a demo at dataq.com/lenny. That's data dogq.com/lenny. Coming back to this kind of triad, the other element that I hear more and more of is just the skill of taste and design and user experience. It feels like that's a very hard skill to learn and to me tells me design is going to be much more valuable in the future.

是的，没错。这又是一个很好的例子。比如设计完美图标这个任务层面，AI 可以全天候做这件事，给你一千个图标设计，会非常棒。当然，仍然会有一定程度的人类图标设计，但 AI 会变得非常擅长。但我们要思考的是：这个东西是做什么用的？它如何在人类世界中运作？它会让人们使用时感到快乐吗？会让人们自我感觉良好吗？它能融入人们的生活吗？它能以正确的方式挑战人们吗？所有这些伟大的设计师一直在思考的更高层次问题——设计师的工作将更多地涉及这些更高层次、更重要的部分，而 AI 会做更多的基础任务。所以一种思考方式是，想想世界上最好的设计师，比如乔纳森·艾维。如果你今天是一名 25 岁的设计师，梦想十年后成为乔纳森·艾维，突然之间你有了新的路径，因为乔纳森·艾维做的一切都没有 AI。现在，年轻设计师可以想：“哇，如果我在十年内真正利用好 AI，我会成为世界上有史以来最好的设计师，因为那将不只是我，而是我加上这项技术的超级赋能，能做更多事情。然后我更多的时间和注意力就能集中在这些大多数设计师从未触及的更高层次的事情上。”我认为这将是另一个很好的例子。

Yeah, that's right. And again, this is a great example. So the task level of designing the perfect icon, right, is going to be like, all right, the AI is going to do that all day long. It'll give you a thousand icon designs. It's going to be great. It's going to be fantastic. And there will still, by the way, be some level of human icon design or whatever, but AI is going to get really good at that. But like, what are we trying to do? The capital D design of like, all right, what is this thing for? And how is this going to function in a world of human beings? And like, is this going to make people happy when they use it? Is it going to make people feel good about themselves? Is it going to fit into the rest of their life? Is it going to challenge them in the right way? All these kinds of higher level questions that the great designers have always thought about — the job of a designer will involve much more of those higher level, more important components, and then again with AI doing a lot more of the underlying tasks. So one way to think about it is, you know, the world's best designers, like Jony Ive or whatever, you could be like, "Wow, if I'm a designer today, if I'm a 25-year-old designer and I aspire to be Jony Ive in a decade, all of a sudden I have a new path to get there, which is because Jony Ive did everything he did without AI." Now, a young designer can be like, "Wow, if I really harness AI in a decade, I'm going to be like the best designer the world's ever seen because it's not just going to be me. It's going to be me plus being so super empowered by this technology to be able to do so much more. And then so much more of my time and attention is going to be able to be focused on these higher level things that most designers never get to." And I think that's going to be another great example of that.

所以也许我听到的是一种 T 型策略：如果你想在这三个角色中的任何一个取得成功，就要非常非常擅长那个特定角色——产品管理、工程、设计——然后在另外两个角色上做到足够好。

So maybe what I'm hearing here is kind of this T-shaped strategy: if you want to be successful in any three of these roles, be very, very, very good at that specific role — product management, engineering, design — and then get good enough at these other two roles.

嗯，我觉得这很棒，非常相关。你知道，斯科特·亚当斯不幸去世了，真是个悲剧。多年来我经常引用他著名的职业建议，我觉得很有道理，也和你说的吻合。他常说：“你看，我本来可以成为一个相当不错的漫画家，或者相当擅长商业，但事实上，我是一个懂商业的漫画家，这让我在创作《呆伯特》时格外出色。”因为即使世界上最优秀的漫画家不懂商业，也永远写不出《呆伯特》；世界上最优秀的商业人士不会画漫画，也做不出《呆伯特》。只有同时拥有这两种技能的人才能创作出《呆伯特》，这是历史上最成功的漫画之一。斯科特总是这样描述：从职业发展的角度看，擅长两件事的叠加效应超过两倍，擅长三件事的叠加效应超过三倍，因为你成为了领域组合中的超级相关专家。你在整个经济中都能看到这一点。举个例子：好莱坞。有很多编剧不会导演电影，但他们可以是非常成功的编剧；有很多导演不会编剧，他们也可以是非常成功的导演。但娱乐业的超级明星是那些既能编剧又能导演的人。他们有一个术语叫“作者导演”。这些人才是真正推动这个领域的创造性力量。顺便说一句，我花了很多时间和好莱坞的人谈论 AI。好莱坞现在也面临着我们在科技领域描述的那种墨西哥僵局，只不过在好莱坞，对于电影制作来说，是导演、编剧和演员。因为导演现在想：“哇，我不再需要编剧了，因为 AI 可以写剧本；我也不再需要演员了，因为我可以有 AI 演员。”编剧说：“我不需要导演，因为 AI 可以导演电影，AI 也可以做演员。”演员说：“这两个我都不需要。我可以让 AI 导演，让 AI 写剧本，我只要出现表演就行。”所以这是同样的三角结构。有趣的是，他们说的都对。这三个领域中的每一个人都将能够横向扩展，掌握那些额外的技能。结果就是，会有更多人能够编剧兼导演、编剧兼演员、导演兼演员，或者三者都做。我认为，就像你提到的 T 型结构，这基本上会适用于整个经济。

Well, so I think that's great. I think that's really relevant. And then, you know, Scott Adams unfortunately just passed away, which is a real tragedy, but I've referred for years to Scott Adams' famous career advice he would give people, which I think makes a lot of sense and dovetails with what you're saying. He used to say, "Look, I could have been a pretty good cartoonist, or I could have been pretty good at business, but the fact that I was a cartoonist who understood business made me spectacularly great at making Dilbert." Because even the world's best cartoonist who didn't understand business could have never written Dilbert. And the world's best business people who didn't know how to do cartoons couldn't have done Dilbert. It took somebody who actually had both of those skills to be able to make Dilbert, which is one of the most successful cartoons in history. So the way Scott always described it was that from a career development standpoint, the additive effect of being good at two things is more than double, and the additive effect of being good at three things is more than triple, because you become a super relevant specialist in the combination of the domains. And you see this all over the economy. I'll give you an example: Hollywood. There are a lot of writers who can't direct a movie and they can be very successful writers. There are a lot of directors who can't write a movie. They can be very successful directors. But the superstars in the entertainment industry are the people who can write and direct. They have a term for those: they call them "auteurs." And those are the people who are the real creative forces that move the field. By the way, I've been spending a lot of time talking to Hollywood people about AI. Hollywood has the same Mexican standoff going on right now that we described in tech, except in Hollywood, for filmmaking, it's the director, the writer, and the actor. Because the director is now thinking, "Wow, I don't need the writer anymore because AI can write the script, and I don't need the actor anymore because I can have AI actors." The writer is saying, "I don't need the director because AI can direct the movie, and AI can do the actors." And the actor is saying, "I don't need either one of these guys. I can have the AI direct the thing, I can have the AI write the thing, and I'm just going to show up and do my performance." So it's the same kind of triangular configuration. And what's great about it is they're all correct. Each person in each of those three fields is going to be able to expand laterally and pick up those additional skills. And as a consequence, you're going to have more people who can write and direct, or write and act, or direct and act, or do all three. And I think, to your point about your T-shaped thing, that's going to be true basically across the entire economy.

如果你思考 T 型结构，T 的横杠代表你熟悉多少个领域，能够借助 AI 工具做出真正优秀的工作。而 T 的竖杠代表你在至少一个领域能深入到什么程度，让你真正精通。如果你在编程上非常深入，又能用 AI 做设计和产品管理，那就是你的 T 型。你在横杠上是三重威胁，同时有深厚的技术根基。到那时，你就是超级个体，能够像变魔术一样设计和构建新产品，做到我们这一代人做梦都想不到的事。我认为这是一个普适的理论，可以应用于整个经济。

And if you think about the T configuration, the top of the T is how many individual domains you are familiar enough with to be able to use AI tools to do really good work. And then the stem of the T is how deep you can go in at least one of those domains so that you really deeply know what you're doing. If you're super deep on coding and you can use AI to do design and product management, that's your T right there. You're a triple threat at the top of the T, but with this level of technical grounding underneath. At that point, you're the superpowered individual, able to perform feats of magic in designing and building new products that people in my generation couldn't have dreamed of. I think this is a universal theory that can apply across the entire economy.

我现在要发明一个新框架。忘掉 T 型框架。我想象一个侧躺的 F 或 E，有两三个向下的部分。所以我听到的是，至少擅长两样。

I'm going to invent a new framework right now. Forget the T framework. I'm picturing an F sideways or an E with two or three downward parts. So what I'm hearing is get good at at least two.

我认为没错。组合很重要。我的朋友 Larry Summers 对 Scott Adams 的说法有不同版本。他常告诉人们，职业规划的关键是不要被替代。他是经济学家，所以那是经济学语言。本质上就是不要被替换。不要做一颗螺丝钉。所以不要只做一件事。如果你只是一个设计师、一个产品经理、一个程序员，理论上你随时可以被替换。但如果你有这个侧躺的 E 或 F，这种实际上很罕见的组合，那么你突然就不可替代了。你不仅不可替代，而且极其重要，因为你是世界上少数能做这种组合的人。借助 AI，你成为这种人的能力比以往任何时候都大大增强了。

I think that's right. The combination. My friend Larry Summers had a different version of the Scott Adams thing. He used to tell people, the key for career planning is don't be fungible. He's an economist, so that was economics speak. What that means essentially is don't be replaceable. Don't be a cog. So don't just be one thing. If you're just a designer, just a product manager, just a coder, then in theory you can be swapped in or out. But if you have this E or F lying on its side, this combination of things that's actually quite rare, then all of a sudden you're not fungible. Not only are you not fungible, you're actually massively important because you're one of the only people in the world who can do that combination. Your ability to become one of those people is titanically enhanced with AI compared to anything we've ever seen before.

这太有意思了，因为我曾和擅长这两种技能的人共事，他们在公司里被称为独角兽。她能编程和设计。我在这里听到的是，这就是你需要成为的样子。你需要至少擅长两件事。我记得你用了烟囱之类的词，比如这边是产品经理，那边是工程师和设计。我听到的是，你需要至少擅长其中两项技能。这两个角色的壁垒正在消失。

This is so interesting because I've worked with people that are good at these two skills and they were always called unicorns at the company. She can code and design. And what I'm hearing here is this is what you need to become. You need to become really good at at least two things. I think you use the term smoke stack or something where it's like PM over here, engineer design. And what I'm hearing here is you need to get good at at least two of these skills. The silos of these two roles are disappearing.

没错。我再次强调，对于所有听众，这一点怎么强调都不为过。我认为人们还没有充分从 AI 中获益的一点是，它会教你。这太神奇了。以前从来没有一种技术，你可以问它“教我怎么做这件事”。人们过于关注如何使用大语言模型，比如“我想让它为我做什么？”这当然很重要，但另一方面是“我能让它教我做什么？”它同样擅长这一点。这种潜在的超级能力：真正想提升自己、发展职业的人，应该把每一分钟空闲时间都用来和 AI 对话，说：“好，训练我。告诉我如何让我变得超级强大。训练我如何成为产品经理。”它会很乐意这么做。它完全知道怎么做。给我出题，给我布置任务，然后评估我的结果。它会像为你工作一样乐意做这些。

That's right. Again, I can't overstress the following for anybody listening. The thing about AI that I think people are not getting enough benefit out of yet is that it will teach you. This is amazing. There's never been a technology before where you can ask it, teach me how to do this thing. So much focus is on figuring out how to use a large language model, like what am I going to try to get it to do for me? That's very important, but the other side is what can I get it to teach me how to do? It's just as good at that. This level of latent superpower: people who really want to improve themselves and develop their career should be spending every spare hour talking to an AI, saying, "All right, train me up. Tell me how to superpower me. Train me how to be a product manager." It will happily do that. It knows exactly how to do that. Run me problems, make me assignments, then evaluate my results. It will do that just as happily as it will do work for you.

我听说过两个技巧。一个是观察输出，观察智能体在工作时的思考和决策过程。如果你不是工程师，就坐在那里看它思考和做决定。这几乎成了学习编程之上的一层：学习观察智能体在做什么、想什么，因为这能教你架构。另一个是，几位播客嘉宾提到过：当你卡住然后想办法解决后，你问它：“我本可以怎么做不同？我本可以说什么来从一开始避免这个错误？”

Two tricks I've heard along those lines. One is to watch the output, what the agent is doing and thinking as it's doing the work. So if you're not an engineer, just sit there and watch it think and make decisions. It's almost become a layer on top of learning to code: learning to see what the agent is doing and thinking because that teaches you about architecture. The other is, a couple podcast guests have mentioned this: when you get stuck and then you figure out how to unstuck yourself, you ask it, "What could I have done differently? What could I have said that would have avoided this error in the first place?"

没错。关于第一个技巧，这正是我和我 10 岁孩子做的。如果你让 AI 写一段代码，然后它返回的结果不工作……

Yeah, that's right. Look, on that first one, this is what I'm doing with my 10-year-old. If you ask an AI, write me this code, and then it comes back and it doesn't work right...

如果你只知道一个单一函数——我问了它，它给了我一个不好的结果——你还能做什么？你不明白它为什么给出那个结果。你真的知道该告诉它什么才能让它做点别的吗？但按你的说法，如果你实际观察它在做什么，并且有了那种基础——就像你耳朵或 F 的那条腿——如果你有了那种基础，你就能说：‘哦，我明白它在做什么，我看到它在哪里犯了错，我看到它哪里跑偏了’，然后你就能突然介入并说：‘不，不，我不是那个意思，做这个别的。’再说一次，这是拥有那种真正的协同关系的重要部分——你理解。顺便说一句，我所说的一切，和与人类合作是一样的。如果你和我是同事，我让你做件事，你回来时完全不一样，我确实需要理解你脑子里在想什么，才能给你反馈。如果我只告诉你‘哦，那错了’，什么也不会发生。我需要有心理理论——我需要理解你在想什么，才能给你正确的反馈。而 AI 的好处是，AI 会乐意整天坐在那里解释它为什么这么做。它会乐意自我批评。你可以这样做。顺便说一句，有一个很有趣的事情，你可以让一个 AI 批评另一个 AI——你让一个 AI 写代码，让另一个 AI 调试代码。你可以让 AI 们互相较量，让它们互相争论。这些技能都将变得极其有价值。

If all you know is a single function—I asked it and it gave me back something that's not good—what do you even do with that? You don't understand why it gave you that result. Do you really understand what to tell it to get it to do something different? But to your point, if you actually watch what it's doing, and then you have that grounding—that leg of your ear or your F—if you have that grounding, then you can be like, 'Oh, I see what it's doing, I see where it made the mistake, I see where it went sideways,' and then you're suddenly able to intervene and say, 'No, no, that's not what I meant, do this other thing.' And again, this is a big part of having that actual synergistic relationship—you understand. And by the way, everything I'm saying is the same as if you're working with human beings. If you and I are colleagues and I ask you to do something, and you come back with something completely different, I do need to understand what was happening in your head in order to give you feedback. If I just tell you, 'Oh, that's wrong,' nothing happens. I need to have theory of mind—I need to understand what you were thinking to give you the right feedback. And the great thing with AI is that AI will happily sit there and explain all day long why it's doing what it's doing. It'll happily critique itself. You can do this. By the way, there's a very fun thing where you can have one AI critique the other AI—you have one AI write the code, you have another AI debug the code. You can play the AIs off against each other and get them to argue with each other. These are all skills that are going to become incredibly valuable.

我想人们称那些为 LLM 委员会。是的。它们在互相交谈。

I think people call those LLM councils. Yes. They're talking to each other.

是的，没错。没错。

Yeah, that's right. That's right.

我确实觉得，如果我是——我没有设计背景。我一直想学设计。感觉这是这三个里面最难通过观察和交谈来学习的，对吧？因为需要很多曝光时间，就像人们用的这个词，你怎么学会成为一个伟大的设计师？感觉这会非常难，也很有价值。

I do feel like if I were—I have no design background. I've always wanted to design. It feels like that's the hardest one to learn of all these three by just watching and talking, right? Because there's a lot of exposure hours, as folks have used this term, just like how do you learn to be a great designer? That feels like it's going to be really hard and valuable.

所以，我的真心话是，我一直有点想当漫画家，但我没有艺术技能。但当我们聊天时，我觉得，也许是时候了。

So, my true confession is I've always kind of wanted to be a cartoonist, but I have no art skills. But as we're talking, I'm like, it might be time.

它们的时代来了，Marc。

Their time has come, Marc.

是的。

Yes.

我想转向创始人，这可能是你的老本行。你花了很多时间与最前沿的 AI 创始人在一起。我很好奇你看到他们在做什么，你怎么看他们，他们的一些运作方式是否让你对创办公司的未来、AI 公司的未来感到震撼。

I want to pivot to founders, your maybe your bread and butter. You spent a lot of time with the most cutting edge AI forward founders. I'm curious to what you see them do, how you see them, some way they operate that's maybe blowing your mind about how the future of starting a company looks, how the future of AI forward companies looks.

是的。这是一个很棒、非常热门的话题，正在前沿实时上演。所以，我认为有三个层面。看看是否合理。第一层是他们在想：‘好吧，AI 如何重新定义产品本身？’这是技术转型中常见的事情，也是很多风险投资的基础：好吧，一项新技术出现了——可能是个人电脑、iPhone、互联网，现在是 AI——然后问题是，这是否是添加到现有产品中的新能力？所以突然之间，你有一个现有的软件业务，然后你有了它的 PC 版本，然后你有了它的 iPhone 版本，你继续前进，把新技术加入混合——它成了现有配方中的另一个成分。当然，很多新技术都是这样。当闪存出现时，它并没有真正重新定义软件行业，因为人们只是从使用硬盘转向使用闪存。但当互联网出现时，基本上老式的本地软件大部分都死了，被网络软件取代。所以有时它是添加到现有事物上的，有时它实际上重新定义了整个产品类别，重新定义了一个行业，在很多情况下公司本身也会更替。所以有这个问题。你刚才提到的一个例子：Nano Banana。一个很好的例子是 Adobe——Photoshop 是一个 40 年的图像编辑特许经营权。AI 是添加到 Photoshop 中用于 AI 图像编辑的功能，还是你完全停止编辑图像，因为你使用 Nano Banana，所有图像都是生成的，让 AI 生成新图像比编辑旧图像更容易？所以我认为在技术的许多领域，这个问题正在被提出，答案会因领域而异。但显然，作为一家风险投资公司，我们大力押注许多这些类别将被彻底重塑，很多最好的创始人正在试图找出如何做到这一点。所以这是 AI 改变产品的定义。我认为下一层实际上是我们已经讨论过的很多内容，即 AI 改变工作。

Yeah. So, this is a great, very topical topic that's all playing out in real time right now on the leading edge. So, I think there's like three layers of it. See if this makes sense. I think layer one is they're thinking, 'All right, how does AI redefine the products themselves?' And this is kind of the time-honored thing that happens at technology transitions, and this is kind of what a lot of venture capital is based on: okay, there's a new technology that comes out—maybe it's the personal computer or the iPhone or the internet or now it's AI—and it's like, all right, is this a new capability that gets added to existing products? So all of a sudden you've got an existing software business, and now you've got your PC version of it, and now you've got your iPhone version of it, and you just kind of keep on going, and you add the new technology into the mix—it's another ingredient into an existing formula. And of course, a lot of new technologies are like that. When flash storage came out, it didn't really redefine the software industry because people just went from using hard disk to using flash storage. But when the internet came out, basically old-school on-prem software for the most part died and got replaced by web software. So sometimes it's additive to an existing thing, sometimes it actually redefines an entire product category, redefines an industry, and in many cases the companies themselves turn over. So there's this question. An example you just mentioned: Nano Banana. A great example is Adobe—Photoshop is a 40-year franchise in image editing. Is AI a feature that gets added to Photoshop to do AI-based image editing, or do you just stop editing images entirely because you're using Nano Banana and all images are just being generated, and it's just easier to have AI generate a new image than to edit an old one? So I think there are many areas of tech in which that question is being asked, and the answers will vary by domain. But obviously, as a venture firm, we're betting hard on many of these categories being totally reinvented, and a lot of the best founders are trying to figure out how to do that. So that's kind of AI changing the definition of the product. I think the next layer is actually a lot of what we've already talked about, which is AI changing the jobs.

所以这很大程度上是我们已经讨论过的内容，但假设我是一家公司的创始人，预算足够雇佣 100 名程序员，我如何让这些程序员成为超级赋能的人工智能程序员，而不是过去那种程序员？如果他们成了超级赋能的人工智能程序员，那意味着我仍然需要 100 人吗？也许现在只需要 10 人。或者我仍然想要 100 人，但他们现在能完成 10 倍的工作量？很多最优秀的创始人现在正在研究这个问题。然后我认为第三只靴子还没落地，但它是关键的一个：拥有公司这个基本概念本身——会改变吗？你有了超级个体这个概念。你能拥有完全由创始人做所有事情的公司吗？创始人监督一支人工智能机器人大军。我们行业里长期存在一个圣杯：能否实现一人十亿美元的公司？过去几年我们有过几个例子。比特币可能是最引人注目的，紧随其后的是以太坊——它不完全是单人，但团队非常小。Instagram 和 WhatsApp 以极小的团队取得了巨大的成果。偶尔会有一些东西以极少数人参与的方式爆发。但大多数软件公司最终都有大量员工。所以我认为一些最前沿的创始人正在思考：我如何重新定义拥有公司的概念？你能拥有一家完全由人工智能组成的公司吗？如果你做的是软件，在某些情况下这似乎是可行的。然后还有终极例子：你能拥有自主的人工智能经济体吗？比如区块链上的人工智能机器人，像企业一样运作、赚钱，然后给我分红？也许这是最终的异常结果。我们有几位创始人正在追求这类事情。我会把这描述为后者——最优秀的创始人正在做的事。

So it's a lot of what we've already talked about, but okay, if I'm a founder of a company and I have room in my budget for 100 coders, how do I get those coders to be super-empowered AI coders, not the kind of coders I used to have? And if they're super-empowered AI coders, does that mean I still need the hundred? Maybe now I only need 10. Or does that mean I still want 100, but now they're doing 10 times more? A lot of the best founders are working on that right now. And then I think the third shoe to drop hasn't quite dropped yet, but it's kind of the big one: the basic idea of having a company — does that change? You've got this concept of the superpowered individual. Can you have entire companies where the founder does everything, overseeing an army of AI bots? There's this holy grail in our industry that's been running for a long time: can you have the one-person billion-dollar outcome? We've had a few of those over the years. Bitcoin is probably the most spectacular example, with Ethereum right behind it — which wasn't quite one person but a very small team. Instagram and WhatsApp had very big outcomes with very small teams. Every once in a while, you get something that hits with a very small number of people. But most software companies end up with huge numbers of employees. So I think some of the most leading-edge founders are thinking: how do I reconstitute the actual definition or idea of having a company? Can you have a company that's literally just all AI? If you're doing software, that seems feasible in some cases. And then there's the ultimate example: can you have autonomous AI economy stuff happening, like AI bots on the blockchain functioning as a business, making money, and issuing me dividends? Maybe that's the final outlier result. We have a few founders who are chasing that kind of thing. I would describe that as the latter — the best founders around.

非常有趣。整个关于一人十亿美元公司的想法。我认为这取决于你对成果的定义。我可以想象……我自己一个人运营我的新闻通讯，加上一些外包人员，有太多烦人的小事要处理：支持工单、问题、漏洞。我很难想象真的有一人十亿美元的公司，即使人工智能处理了大部分支持工作，因为总有那么多随机的边缘情况。我一直在填表格。所以我觉得这取决于：你有外包人员吗？那算不算？什么才算是一人？我就是看不到这种情况发生。

Super interesting. This whole idea of a one-person billion-dollar company. I think it depends on your definition of what an outcome is. I could see... having run my newsletter as one person with some contractors, there's so many little annoying things I have to deal with: support tickets, issues, bugs. It's hard for me to imagine actually a one-person billion-dollar company, even if AI is handling so much of your support, because there's just so many random edge cases. I'm constantly filling out forms. So I guess it depends: do you have contractors? Does that count? What does it mean to be one person? I just can't see that happening.

是的。我的意思是，你看，比特币的中本聪就做到了。

Yeah. I mean, look, Bitcoin's Satoshi pulled it off.

但比如，开源社区——那算吗？我不知道。我想算吧。好吧。

But like, the open source community — does that count? I don't know. I guess it counts. Okay.

是的。没错。对吧。所以，是的。我想说我不是要给出答案，而是……我认识的最聪明的人都在认真思考这个问题。

Yeah. Exactly. Right. So, yeah. And I would say I don't propose to have answers here, but more just... the smartest people I know are thinking hard about this.

是的。你怎么看护城河？人工智能中一个持续的大问题——一切都在变化。你们对人工智能中的护城河有什么看法？这还是个概念吗？你们在意吗？

Yeah. What do you think about moats? A big question constantly in AI — the fact that everything's changing. Just what's your guys' thesis on moats in AI? Is that even a thing? Do you care?

我对真正重大技术变革的经验——我亲身经历了互联网，目睹了这一切——是真正重大的技术变革需要很长时间才能展开，并且所有结构性影响会随着时间的推移逐渐显现。然后人们会急于下判断，说：“哦，因此很明显是 XYZ。因此很明显这类公司将成为未来的公司，而不是那类。很明显这个现有企业能够适应，而另一个不能。很明显经济机会存在于这类初创公司，而不是其他。很明显护城河将出现在技术的这个领域，而不是那个领域。”每个人都以极大的自信陈述这些观点，听起来好像他们掌握了所有答案。然后这些想法充斥媒体，因为媒体自然更看重确定的答案而不是开放性问题。当 CNBC 邀请嘉宾时，他们想要一个能来说“是的，事情就是这样”的嘉宾，而不是“我认为这是个很好的问题，让我们从八个不同角度来辩论”。我发现，如果你几年后回顾这些预测——你可以通过查阅 1993 年到 1997 年，甚至到 2005 年或 2010 年关于互联网的报道，看看人们在头 10 年或 15 年里做出的自信陈述——我会说几乎所有这些预测都是错误的，通常错得很离谱。所以我认为，面对巨大的技术变革，这个过程……

My experience with really big technological transformations — and I kind of lived this directly with the internet, and I saw this happen — is the really big technological transformations take a long time to play out, and there are all these structural implications that cascade out over time. And then there's this rush to judgment up front where people say, "Oh, it's therefore obvious that XYZ. It's therefore obvious that this kind of company is going to be the company of the future, not that kind. It's obvious that this incumbent is going to be able to adapt and this other one isn't. It's obvious that there's economic opportunity in this kind of startup and not in these others. It's obvious that the moats are going to be in this area of the technology but not in this other area." And everybody states those things with an enormous amount of self-assurance, where they really sound like they have all the answers. And then these ideas saturate the media, because the media naturally prizes definitive answers over open questions. When CNBC is booking guests, they want a guest who will come on and say yes this is the way it's going to be X, not "I think that's a really good question and let's debate it from eight different angles." What I've found is if you look back on those predictions a few years later — you can do this by pulling up coverage of the internet from 1993 through 1997, or even through 2005 or 2010, and look at the confidence statements people made in the first 10 or 15 years — I would say almost all of them were wrong, generally quite badly wrong. So I just think with massive technological change, the process...

这将是五六层结构性变革，会随时间逐步展开。我们谈过很多，但这对产品、公司、工作、行业的定义意味着什么？在国家层面和全球层面如何展开？如何与政治、工会、战争交织？中国会怎么做？存在大量未知数。预先判断这些事情非常危险。我来做个思想实验：AI 模型本身有防御性吗？有护城河吗？一方面，似乎应该有，因为建造需要数十亿美元，需要算力和数据的临界质量，只有一定数量的工程师知道怎么做，他们拿 NBA 球星的薪水，公司还要处理政治、媒体、声誉、监管和法律问题。所以最终可能有两三家公司占据 100% 的市场份额，形成经典寡头或垄断。这在软件领域发生过很多次。另一方面，如果三年前你告诉我，在 ChatGPT 的圣诞节之后一年到一年半内，会有五家其他美国公司拥有完全同等的产品，另外五家来自中国，还有开源版本也基本一样，我会很惊讶。那个看起来像黑魔法的东西很快就商品化了。GPT-3 发布一年内，就有了开源 GPT-3，在更少的硬件上运行，免费可用。现在有 Google、Anthropic、XAI、Meta、DeepSeek、Kimi 等。所以即使在 LLM 或 AI 模型层面，你眯着眼也能从两边论证。应用层面也一样：一种观点认为应用不存在，因为模型会做所有事。另一种观点是，将模型作为引擎适配到涉及人类的领域，比如医疗、法律或编程，需要它符合用途，所以应用层会非常重要。也许 LLM 商品化，价值流向应用。我认识两边的聪明人。我诚实的回答是，我们正处于发现过程中。从结构上看，这是一个复杂的自适应系统：技术提供一种输入，法律和监管过程是另一种，企业家的个人选择很重要，经济学很重要，投资者资本的可用性会变化。我们还不知道结果，需要对惊喜保持开放。作为 VC，这很令人兴奋，因为我们沿着每种策略下注，看看结果如何。也许有一个天才对冲基金经理已经全搞明白了，但如果存在，我还没见过。

It's going to be like five or six layers of structural change that will play out over time. A lot we've talked about this, but the implications on what are the definitions of products, companies, jobs, industries, how this plays out at the national and global level, how it intersects with politics, unions, war, what China is going to do. There's a tremendous number of unknowns. It's really dangerous to prejudge these things. Let me run a thought experiment: Are AI models themselves defensible? Is there a moat? On one hand, it seems like there should be because it takes billions of dollars to build, you need critical mass of compute and data, only a certain number of engineers know how to do this and they get paid like NBA stars, and companies have to deal with political, press, reputational, regulatory, and legal issues. So probably at the end, there will be two or three companies with 100% market share, a classic oligopoly or monopoly. That has happened in software many times. On the other hand, if you had told me three years ago that within a year to a year and a half of ChatGPT's Christmas, there would be five other American companies with exactly capable products, another five from China, and open source that was basically the same, I would have been amazed. The thing that seemed like black magic became commoditized really fast. Within a year of GPT-3 coming out, there were open source GPT-3s running on a fraction of the hardware, available for free. Now you have Google, Anthropic, XAI, Meta, DeepSeek, Kimi, and others. So even at the level of LLMs or AI models, you can squint and make the argument either way. Same thing at the app level: one school of thought says apps are not a thing because the model will just do everything. Another view is that adapting the model as an engine into a domain involving human beings, like medical or legal or coding, requires it to be fit for purpose, so the application level will matter enormously. Maybe the LLM commoditizes and value goes to the apps. I know very smart people on both sides. My honest answer is we're in a process of discovery. Structurally, it's a complex adaptive system: technology provides one input, legal and regulatory process another, individual choices by entrepreneurs matter a lot, economics matter, availability of investor capital varies. We don't know the outcomes yet and need to be open to surprises. As a VC, this is exciting because we make bets along every strategy and see how it plays out. There may be one brilliant hedge fund manager who has it all figured out, but if they exist, I haven't met them.

所以我听到的是，现在不要过度纠结护城河，因为我们不知道最终会是什么。尽管可能觉得 OpenAI 不可能失去领先地位，但我们显然看到了很多竞争。GPT 封装器这个点很好。'封装器' 是个贬义词。一年前，你还只是个 GPT 封装器。现在这些公司是全世界最大、增长最快的公司。

So what I'm hearing here is don't over obsess with moats at this point because we have no idea what it'll end up being. As much as it may feel like there's no way OpenAI will lose this lead, clearly we're seeing a lot of competition. GPT wrapper point is really great. A lot is such a derogatory term. A year ago, just like you're just a GPT wrapper. Now it's the companies that are the biggest, fastest growing companies in the world.

是啊，这有点像。三年前是 ChatGPT 的节日。上个月左右是 Claude 的节日，特别是 Claude Code，用于编程。这很了不起，因为 Claude 本身就是一个伟大的成就，但 Claude Code 是一个应用，对吧？它是一个 Claude 封装器。它是一个智能体框架。然后他们做了一个了不起的事，推出了 co-worker？co-work。他们说 co-work，而 Claude Code 在一周内写出了 co-work。一周半。没错，100%。

Yeah, well, it's a little like, even just with, you know, three years ago was the holiday of ChatGPT. This last month or whatever has been the holiday of Claude, particularly Claude Code, for coding. It's pretty amazing because there was Claude, which is obviously a great accomplishment, but then there's Claude Code, which is an app, right? It's a Claude wrapper. It's an agent harness. And then they did this amazing thing where they came out with co-worker? Co-work. And they said co-work, which Claude Code wrote co-work in a week. A week and a half. Yep. 100%.

嗯，有两种看待方式。一种是，Claude Code 能在一周半内构建出 Co-Work，这确实令人印象深刻，太棒了。另一种是：Co-Work 只花了一周半就开发出来，那它到底能有多复杂？进入门槛能有多高？所以这是一种推拉关系。它功能强大、价值极高，现在全世界的人都说，‘我简直不敢相信我能用它做什么，这是有史以来最神奇的产品。’但与此同时，它只花了一周半。所以其他每个模型公司肯定都在想，‘我们需要构建一个亚洲艺术家，我们需要为普通人构建一个 Co-Work 的东西。’我不是说我懂什么，但显然他们都会这么做，对吧？那么，这有多大的防御性？六个月后，Claude Code 会不会像 GitHub Copilot 那样被超越？过去三年的历史是，任何看起来像根本性突破的东西都会被迅速复制和超越。我认识的许多领域内最聪明的人，在我跟他们喝了几杯之后，他们会说，‘是的，有一种理论是，大型实验室之间其实没有什么秘密。他们都有相同的信息和知识，而且他们经常互相超越。目前没有什么专有技术。’证据就是 DeepSeek，它突然出现，基本上重新实现了美国大型实验室的很多想法，还有一些自己的原创想法。对于一个中国的对冲基金来说，做到这一点并不难。那么，防御性到底有多强？但另一方面，所有这些大型实验室现在都在像对待摇滚明星一样对待单个工程师，他们都是极其聪明和有创造力的人。也许这些实验室中的任何一个都有十几个新想法，这些想法将成为难以复制的巨大突破。所以我认为我需要在预测能力上打个大折扣。对我来说，试图说‘因此，五年后的行业结构将是 X，大赢家将是公司 Y，杀手级应用将是 Z’远没有那么有趣。我认为我无法预测。在这样的时候，更好地利用我的时间是保持非常灵活和适应性强。

Well, there are two ways of looking at that. One is that it's really impressive that Claude Code was able to build Co-Work in a week and a half. That's amazing. The other way is: Co-Work was developed in a week and a half, so how much complexity could there be? How much of a barrier to entry can there be in something developed that quickly? So it's this push and pull. It's incredibly functional and valuable, and people all over the world now say, 'I can't believe what I can do with this, it's the most magical product ever.' But at the same time, it took a week and a half. So every other model company must be thinking, 'We need to build an Asian artist, we need to build a Co-Work thing for regular people.' I'm not saying I know anything, but obviously they're all going to do that. So how defensible is that? In six months, will Claude Code get lapped the same way GitHub Copilot got lapped? The history of the last three years is that everything that looks like a fundamental breakthrough gets replicated and lapped very quickly. Many of the smartest people I know in the field, when I get a couple drinks into them, say, 'Yeah, one theory is there really aren't any secrets among the big labs. They all have the same information and knowledge, and they lap each other on a regular basis. There's not a lot of proprietary anything at this point.' Evidence of that is DeepSeek, which came out of left field and re-implemented a lot of ideas from American big labs, with some original ideas of its own. It wasn't that hard for a hedge fund in China to do it. So how much defensibility is there? But on the other side, all these big labs are now paying individual engineers like rock stars, and they're incredibly bright and creative people. Maybe there are a dozen new ideas in any one of these labs that will be huge breakthroughs hard to replicate. So I think I need to put a big discount on my forecasting ability here. It's much less interesting for me to try to say, 'As a consequence, industry structure in five years will be X, the big winner will be company Y, the killer app will be Z.' I don't think I can predict that. A much better use of my time is being very flexible and adaptable at a time like this.

那么，考虑到这一切，你是否觉得有更多值得关注的东西来帮助你决定在哪里下注，还是说答案本质上就是你们已有的策略，即下很多赌注？你们筹集了历史上最大的基金。这是你们在这个世界上获胜的方式吗？

So with all this in mind, do you feel like there's something you're paying attention to more to help you decide where to place your bet, or is the answer essentially the strategy you guys have, which is place a lot of bets? You raised the largest fund in history. Is that the way you win in this world?

是的。对我们来说，我们有一个非常深思熟虑的策略。一种思考方式是用彼得·蒂尔的框架。他说有一个 2x2 矩阵：乐观与悲观，以及确定性与不确定性。他一直认为硅谷的特点是过多的‘不确定性乐观’。他的意思是，一个不确定性乐观主义者认为世界会变得更好，但无法解释如何变好。一些事情的组合会让世界变得更好，即使我们不知道那些事情是什么。他会说这有沦为一厢情愿或妄想的风险，而世界更需要的是‘确定性乐观主义者’——那些说‘不，世界会变得更好，因为我要做这件具体的事情’的人。他会把埃隆归类为确定性乐观主义者，而风投是不确定性乐观主义者。我认为彼得的框架有很多道理，但我不同意其中一部分。我认为不确定性乐观是一种比他历来所描述的更强的现象。我会把自己坚定地归入不确定性乐观主义者的类别，这也是我们在 a16z 的策略。原因不是一厢情愿。风投的不确定性乐观，或者说 a16z 的不确定性乐观，实际上非常具体：有像埃隆和其他许多人这样极其聪明能干的人，他们是创始人和产品创造者。他们每个人都是确定性乐观主义者——每个人对自己要做什么都有非常强烈的看法。但资本主义体系、美国经济和硅谷的伟大之处在于，我们不仅仅有一个这样的人，或者十个。我们有一百个、一千个、一万个这样的人。优化结果的方式是让尽可能多的人尽可能优秀。

Yeah. For us, we have a very deliberate strategy. One way to think about this uses the Peter Thiel formulation. He said there's a two-by-two: optimism and pessimism, and then determinism and indeterminism. He always argued that Silicon Valley is characterized by too much of what he calls 'indeterminate optimism.' What he meant by that is basically an indeterminate optimist thinks the world will be better but can't explain how. Some combination of things will make the world better even if we don't know what those things are. He would say that risks being wishful thinking or delusional thinking, and what the world needs more is 'determinate optimists' — people who say, 'No, the world will be better because I'm going to do this specific thing.' He would classify Elon as a determinate optimist, and VCs as indeterminate optimists. I think there's a lot to Peter's framework, but I would say I disagree with part of it. I think indeterminate optimism is a stronger phenomenon than he historically represented it as. I would put myself firmly in the indeterminate optimist category, and that's the strategy we have at a16z. The reason is not wishful thinking. The indeterminate optimism of venture capital, or of a16z, is actually very specific: there are extremely bright and capable people like Elon and many others who are founders and product creators. Each of those individuals is a determinate optimist — each has a very strong view of what they're going to do. But the great virtue of the capitalist system, the American economy, and Silicon Valley is that we don't just have one of those, or ten. We have a hundred, a thousand, ten thousand of those. The way to optimize the outcome is to have as many of those as possible be as good as possible.

尽可能拼命跑，未来的本质就是我们不知道所有答案，这没关系。正确的应对方式是尽可能多做实验，让尽可能多的聪明人尝试做尽可能多有趣的事。所以，我坚定地站在不确定乐观主义者这一边。我在想，现在你越来越看重的，是不是这种确定乐观的创始人——拥有巨大野心，并且真正在努力实现它。

Run as hard as possible, and then just the nature of the future is like we just don't know all the answers, and that's okay. And the right way to deal with that is to run as many experiments as possible and have as many smart people try to do as many interesting things as possible. So yeah, I would put myself firmly on the side of the indeterminate optimist. I'm wondering if the answer to the question of what you look for now more and more is this determinate optimistic founder, who has this massive ambition and is actually working on achieving it.

对，没错。创始人必须是确定乐观主义者。他们现在需要有一个非常具体的计划。你看，创始人总是批评说：“风投很容易，因为你们不用真正下注。你们不用自食其果。你们可以下多个赌注，管理一个投资组合。你们应该对我们创始人更有同情心，因为我们只能下注一次。”这有一定道理。但反驳点是，创始人可以经营自己的公司，我们不行。我们无法把手放在方向盘上。所以，成为确定乐观主义者的巨大好处是，你可以一心一意地朝着那个目标执行。从长远来看，历史会记住谁？历史会记住亨利·福特，而不是那个投资了福特汽车公司以及其他十家失败汽车公司的种子投资人。所以，确定乐观主义者是创始人、公司建设者、工程师。这些人是真正做事的人，值得 99.99999% 的功劳。但话虽如此，我确实认为，背景中也需要一些不确定乐观主义者，一路帮忙，维持整个循环的运转。

Yeah, no, that's right. The founders need to be determined optimists. They need to have a very specific plan now. Look, the critique from the founders is always, 'Oh, VCs have it easy because you don't actually have to commit. You don't have to make the bed you lie in. You can place multiple bets, operate a portfolio. You should have a lot more sympathy for us as founders because we only get to make the one bet.' There's truth to that. The counter-argument is that founders get to run their companies. We don't. We don't get to put our hand on the steering wheel. So the great virtue of being a determined optimist is you actually get to single-mindedly execute against that goal. In the long run, who does history remember? History remembers Henry Ford, not whoever was the seed investor who seeded Ford Motor Company and ten other car companies that failed. So the determined optimist is the founder, the company builder, the engineer. These are the people who actually do the thing and deserve 99.99999% of the credit. But having said that, I do think there is a role for having some indeterminate optimist in the background, helping along the way and helping keep the whole cycle going.

你在考虑 AGI 时，会调整你的投资理念吗？随着我们接近并达到 AGI，作为投资者，你认为你的投资理念会如何变化？

Do you think about AGI in shifting your investment thesis? As we approach AGI and hit AGI, as an investor, how do you think about your investment thesis changing?

是的。我一直对 AGI 的概念有些纠结，因为它有平凡的定义和宇宙级的定义。先说宇宙级的。宇宙级的基本上就是奇点，世界发生根本性变化的时刻，旧世界的规则消失，我们进入一个新领域。奇点的完整定义是，人类判断不再真正相关，因为出现了自我改进循环。AI 在自我改进，在所谓的“起飞”场景中加速，AI 自我改进，机器决策速度远超人类，人类只能坐在那里看着机器做事。我真的不认为我们生活在那个世界里，不管它叫乌托邦还是反乌托邦。我觉得我们没那么幸运或不幸。行业参与者达成的平凡定义是，AI 能在每一项经济相关任务上做得和人类一样好。Anthropic 联合创始人的说法是，它是一篮子最有价值的经济任务，大概 10 到 15 项，而不是每一项经济任务。

Yeah. So I've always kind of struggled with the concept of AGI, because there's the prosaic definition and then there's the cosmic definition. Let's start with the cosmic one. The cosmic one is basically the singularity, the moment where the world fundamentally changes, the rules of the old world are gone, and we're operating in a new domain. The full definition of singularity is a world in which human judgment is no longer really relevant because you get this self-improvement loop. The AI is improving itself, racing in so-called takeoff scenarios, where the AI is improving itself and the machines are making decisions so much faster than people, and people are just sitting there watching the machine do its thing. I don't really think we live in that world, whether you call it utopian or dystopian. I don't think we're lucky or unlucky enough to live in that world. The prosaic definition of AGI that industry participants have converged on is when the AI can do every economically relevant task as good as a human. The way the co-founder of Anthropic put it is like a basket of the most valuable economic tasks, so it's 10 or 15, not every single economically valuable task.

好的，明白了。所以这甚至是一个略微缩小的定义。顺便说一句，我们显然已经很接近了，如果还没达到的话。

Okay, got it. So it's maybe even a slightly reduced definition. And by the way, we're clearly getting close to that if we're not already there.

所以对于那个定义，我觉得宇宙级的高估了将要发生的事，而你刚才给出的 AGI 定义低估了将要发生的事。它几乎太简化了。原因是我认为没有理由假设人类技能水平是任何事物的上限。AGI 总是相对于人类工人而言的。但人类技能水平在某个点达到上限，是因为人类有机体固有的生物限制。例如，人类 IQ，即所谓的流体智力或 G 因子，作为一个物种，人类 IQ 上限大约在 160。160 是爱因斯坦级别。IQ 160 的人是那些提出新物理学的人。只有极少数。通常，当我们遇到非常聪明的人，比如畅销书作家或世界顶尖的研究科学家或医生，IQ 大概在 140。对于非常优秀的律师，大概在 130。对于企业中非常优秀的直线经理，大概在 110。

And so on that one, I kind of feel like the cosmic one overstates what's going to happen, and the AGI definition you just gave understates what's going to happen. It's almost too reductionist. The reason is I don't think there's any reason to assume that human skill level is the cap on anything. AGI is always relative in comparison to a human worker. But human skill level caps out at a certain point because of the inherent biological limitations of the human organism. For example, human IQ, what they call fluid intelligence or the G factor, tops out in humans as a species around 160. At 160, it's Einstein level. The 160 IQ people are the ones who come up with new physics. There's only a small handful. Generally, when we run into somebody who's incredibly smart, like a best-selling author or one of the world's best research scientists or doctors, it would probably be 140. For a really good lawyer, it's probably 130. For a really good line manager in a business, it's probably 110.

你知道，如果你要找一名会计，比如擅长为小企业做账的小企业会计，那大概 IQ 105。而人类智力令人印象深刻的范围大致是 110 到 160。好消息是，有很多这样的人，但 IQ 140、150、160 的人并不多。但那只是人类大脑容量的限制。如果解除人类生物学的限制，就没有理论上的上限了。所以，你能——已经有人在用现有 AI 模型做人类等效 IQ 的实验了。顺便说一句，现有 AI 模型现在的测试水平大约在 131 到 140，这意味着它们会达到 160，而且可以说已经开始接近 160 了。但我认为我们很快就会拥有 IQ 160、180、200、250、300 的 AI 模型。我认为这很棒。我对这件事的感觉，就像我们偶尔出现一个爱因斯坦那样好。世界会因为有更多还是更少的爱因斯坦而变得更好？答案当然是世界会因为有更多爱因斯坦而变得更好。当然，世界也会因为有 IQ 像爱因斯坦或超过爱因斯坦的机器而变得更好。但我认为机器的 IQ 会超过人类。我认为这真的很好。然后性能——又回到了 AI 编程这件事上。任务性能会变得更好。我认为这正是 Line of Stars 特别的地方：这东西开始生成比我更好的代码了。所以我们将拥有比最优秀的人类程序员更优秀的 AI 程序员。我认为这很棒。我认为我们将拥有比最优秀的人类医生更优秀的 AI 医生。我认为我们将拥有比最优秀的人类律师更优秀的 AI 律师，这将非常有趣。我认为这也很好。所以我不认为——我们已经习惯了生活在一个我们不知道“好”能有多好的世界里，因为我们一直被自己的生物学所限制，而我们将体验到指尖拥有比人类在这些领域更强大的能力是什么感觉。所以我认为“人类等效”这个概念只会成为一个注脚。就像，哦，是啊，那只是 2026 年的某个星期二发生的事，其实无关紧要，因为下一个问题是：在一个我们拥有比那更好的机器的世界里，我们能做什么？所以我认为这更像是一个探索如何超越人类能力的过程，而不是某种恰好与人类阈值重合的特殊奇点时刻。

You know, if you're looking for an accountant, like a small business accountant who's good at doing the books for small businesses, that's probably 105. And so the scope of impressive human intellectual ability is sort of that 110 to 160 spectrum. Good news is there are a lot of those people running around, but there aren't that many at 140, 150, 160. But that's just the limitations of what can fit in here. And there's no theoretical limit on where this goes if you release the limitations of human biology. So can you have—you already have people running these experiments to do human equivalent IQ for existing AI models. By the way, existing AI models right now are testing around the 131 to 140 level, which means they're going to get to the 160 level, and they're arguably starting to get to the 160 level now. But I think we're going to have AI models relatively quickly that are going to be like 160, 180, 200, 250, 300. And I think that's great. I feel as great about that as I do about the fact that we occasionally get an Einstein. Would the world be better off or worse off with more or fewer Einsteins? The answer is of course the world would be better off with more Einsteins. And of course the world would be better off with machines that have IQ like Einstein or greater than Einstein. But I think the IQ of the machines is going to exceed that of humans. I think that's really good. And then the performance—again, it goes back to the AI coding thing happening. Performance against tasks is going to get better. I think this is where Line of Stars in particular is like, okay, this thing is starting to generate better code than I can. So now we're going to have AI coders that are actually better coders than the best human coders. I think that's great. I think we're going to have AI doctors that are better than the best human doctors. I think we're going to have AI lawyers that are better than the best human lawyers, which is going to be very interesting to see. I think that's also great. So I don't think there's a—I think we're used to living in a world where we just don't understand how good good can get because we've been capped by our own biology, and we're going to get to experience what it's like when you have capability at your fingertips that's actually better than human in these domains. So I think this idea of human equivalent is just going to be a footnote. It's like, oh yeah, that was just on Tuesday in 2026 when they hit that, and it kind of didn't matter because the next question was, okay, what do we get to do in a world where we actually have machines that are better than that? So I think this is going to be much more of an exploratory process for actually exceeding human capability than it's going to be any sort of particular singularity moment that just happens to coincide with the human threshold.

200 IQ。这个参照系真是拓展思维，让人思考这些东西会变得多快、多聪明，而且速度很快。

200 IQ. That frame of reference is such a mind-expanding way to think about just how fast and how smart these things are going to get, and quickly.

嗯，我不知道你有没有这种体验。我经常有。我经常有两种体验。一种是，我知道我应该能做这件事，但我就是做不到——太费时间了。我想写个东西，或者想有个理论，或者有个计划，但就是没有八小时，更不用说八周或八年了。而且我知道得还不够多，我无法在脑子里做数学，我的记忆力也不完美。我读书——你对某件事感兴趣后，读了十本书，然后你发现几乎全忘了。我希望我能全部记住，但我做不到。我几乎一直生活在一种无尽的挫败感中。所以，如果我能比现在更聪明，我会做得更好，但我不是。就是这样。我不知道你多久会有这种感觉，但我经常有。因为我们的工作，我认识很多人，我确定他们比我聪明。我知道，因为当我跟他们交谈时，我会在某个时刻发现——前半段对话我一直在记笔记。后半段我就想，这个人就是比我聪明，他们想得比我深，而且会一直比我深，我跟不上。然后我就想，好吧，该死，我得回家喝一杯，因为我就是达不到那个水平。所以我们太习惯于这些限制了，以至于想到有为我们工作的机器没有这些限制——我认为这比人们所认为的要令人兴奋得多。

Well, I don't know if you have this experience. I have this experience all the time. Two experiences I have all the time. One is just like I know I ought to be able to do this, but I just can't—it's going to take too long. I want to write this thing or I want to have this theory or I have a plan, and it's just like I don't have the eight hours, or by the way the eight weeks or the eight years. And I just don't know enough yet, and I can't do the math in my head, and my memory isn't perfect. I read—after you get interested in something, you read 10 books and then you're like, I forgot almost everything I just read. I wish I could retain it all but I can't. I sort of live in this state of endless frustration. So if I could just be smarter than I was, I'd be so much better at what I do, but I'm not. So there's that. And I don't know how often you have this, but I have this on a regular basis. Because of what we do, I know a bunch of people who I know for sure are smarter than I am. I know it because when I talk to them, I find myself at a certain point—for the first half of the conversation, I'm just taking notes the entire time. And for the second half, I'm just like, this person is just smarter than I am, and they're outthinking me and they're going to keep outthinking me, and I just can't. And I'm just like, all right, damn it, I gotta go home and have a drink because I'm just not whatever that is. So we're just so used to having those limitations that the idea of having machines that work for us that don't have those limitations—I think that's much more exciting than people are giving credit for.

天哪，马克，我可以跟你聊上几个小时。我想在结束对话前，问问你的媒体摄入和产品摄入。你刚才提到读书，读十本书。我记得你以持续阅读闻名。我看过一个采访，你说 AirPods 改变了你的生活，你一直在听有声书。那么，在媒体摄入方面，你在读什么？你最近关注什么——播客、新闻通讯、博客？还有，有什么特别的书吗？

Oh man, I could talk to you for hours, Marc. I'm thinking to close out the conversation, I want to ask about your media diet and your product diet. You just talked about books, reading 10 books. I think you famously read constantly. I saw an interview with you where you said AirPods changed my life, I'm just listening to audiobooks all the time. So in terms of media diet, what are you reading? What are you paying attention to these days—podcasts, newsletters, blogs? And then any books in particular?

是的。我读的东西基本上分三类。就一般媒体而言，我几乎采用一种完美的杠铃策略：我读 X，也读旧书。所以要么是最新的时事，要么是 50 年前写的、经得起时间考验的书，其中大概有一些永恒的东西。

Yeah. So what I read is basically three categories of things. In terms of general media, I sort of have a almost perfect barbell strategy: I read X and I read old books. So it's either up-to-the-minute what's happening right now, or it's a book that was written 50 years ago that has stood the test of time, and presumably there's something timeless in it.

至于中间那类内容，我向来持怀疑态度。具体来说就是我之前说的：如果你回头去读旧报纸——没人会这么做，真的很有意思，没人这么做，也没有市场。但如果你真去读，比如读上周五的报纸——我们录节目是周五，那就读上周五的报纸——然后你会惊呼：“天哪，这些预测全没发生。”他们预测的事情没有一件按他们说的方式实现，没有一件是真正相关或正确的。他们根本不了解情况，对本周会发生什么毫无头绪，却基于零信息做出预测和判断。结果呢？什么都没发生。我真希望自己从没读过这些。杂志也一样，翻翻旧杂志，里面充斥着无穷无尽的预测。报纸至少是日更的，而杂志是周更或月更的长周期，文章出版时往往已经过时了。所以我对中间那类内容很有意见。要么是即时性的，要么是永恒的。

And then it's sort of everything in the middle I'm always much more skeptical about. And it's particular what I already said, which is I think if you go back and read old newspapers—nobody ever does this. It's actually really funny. Nobody ever does this. There's no market for it. But if you go back and read old newspapers, and by the way, you can do this. Just read last week's newspaper, right? I guess we're taping on Friday. So read last Friday's newspaper, right? And just go back and read it and be like, 'Oh my god, like none of this happened.' None of what they predicted played out the way that they said that it would. None of this turned out to actually be that relevant or correct. Like they didn't understand—they had no view of what was going to happen this week that they couldn't know, and so they were making predictions and forecasts based on not having any information. But it's like, wow, none of this happened. I wish I had never read this. Oh my god. And then it's kind of the same thing with magazines. Go back and read old magazines. And just the level of the endless numbers of predictions that they make. And the problem with newspapers is at least they're going day-to-day. The thing with magazines is every week or month, it's a long cycle, and so by the time an article even hits publication, it's often out of date. So I just have a big problem with everything in the middle. So it's either of the moment or timeless.

但你提到了 newsletter。另一件事，可能显而易见，但我认为仍然被严重低估了，那就是领域内真正的实践者亲自创作的内容。我认为这仍然是严重被低估的。这是 Substack 现象、newsletter 现象和播客现象的一个巨大组成部分：直接接触那些真正懂行的领域核心人物，这很可能仍然被严重低估。原因在于，我们习惯了大众媒体文化，其中一切都被中介化了，对吧？所有内容都经过电视采访、报纸采访或杂志采访的过滤。而现在，越来越多的情况是，你实际上希望那些正在做事的聪明人自己来解释。然后出现了像播客这样的新型中介，为人们提供了这种可能。所以领域实践者非常棒。我只是陈述一个显而易见的事实。在 AI 领域，这显然是你擅长的，但像 Lex Friedman 也能请到世界顶尖专家。顺便说一句，批评者总是说这些人是在推销自己的书——如果我经营一家初创公司，我就是在卖东西。这种情况确实有一点。但根据我的经验，人们热爱谈论他们所做的工作。他们从根本上想要表达自己在做什么，想要解释，希望别人理解。每个人都乐在其中，他们通过这种方式为人类知识做贡献，也获得了自我满足。所以我认为，聆听那些真正现身说法、谈论自己工作的世界顶尖专家，能带来巨大的价值。当然，如今世界在这方面已经泛滥了，而十年前还不是这样。所以我也尽可能多地这样做。

But then you mentioned newsletters. The other thing, and this is maybe obvious, but I think it's probably still underrated, which is the actual practitioners in the field who are actually creating content. I think this is probably still dramatically underrated. And I think this is a huge part of the Substack phenomenon, the newsletter phenomenon, and the podcast phenomenon: direct exposure to the people who are actually principals in the field who actually know what they're talking about is probably still dramatically underrated. And I think the reason for that is we're used to being in this mass media culture in which basically everything is mediated, right? Everything got filtered through TV interviews or newspaper interviews or magazine interviews. And now more and more, it's just no—you actually want smart people who are actually working on something explaining themselves. And then you have new kinds of intermediation like podcasts that open that up for people to make that possible. So domain practitioners are really great. I mean, just to state the obvious, and in AI it's obviously your stuff, but also like Lex Friedman can have the world's leading experts show up. And by the way, the critique always is people talk their book—if I'm running a startup, I'm just selling. And there's always a little bit of that. But it's also, my experience is people love to talk about what they do. They fundamentally want to express what they do, they want to explain it, they want people to understand it. Everybody kind of enjoys that, and they get to contribute to human knowledge by doing that, and they get ego gratification. So I think there's just tremendous amounts of alpha in listening to the world's leading experts in the space who actually just show up and talk about what they're doing. And of course, the world is awash in that today in a way that it wasn't as recently as 10 years ago. So I do as much of that as I can too.

而且科技界，尤其是硅谷，有一种分享的文化，不试图保守秘密。LinkedIn 上总是有人说：“这怎么是免费的？”但这就是它的运作方式。

And there's also just this culture in tech, Silicon Valley in particular, of sharing, of not trying to keep these secrets. Everyone on LinkedIn is always like, 'How is this free?' It's just the way it works.

是的。有人说过，硅谷是一个公司镇，但这家公司就是硅谷本身。

Yeah. Somebody said Silicon Valley is a company town, but the company is Silicon Valley.

对。

Right.

同样，在 n 等于 1 的层面，比如经营一家公司，那真是件麻烦事，因为你的秘密会流失，员工会离职，整个事情很糟糕。但另一方面，你也从中受益，因为你可以雇佣到拥有各种技能和经验的人，你身处一个能够适应并将人才、技能和知识引导到新领域的生态系统中。所以，作为个体 CEO，这其中有推拉作用。而就整个生态系统而言，如你所说，这绝对是一个神奇的现象。顺便说一句，尽管硅谷存在各种问题，但我认为 AI 是硅谷历史上第九个主要技术平台。硅谷仍然被称为硅谷。我们几十年来都不在这里制造硅了。它之所以叫硅谷，是因为过去这里制造芯片。他们曾经在硅谷拥有实际的晶圆厂，然后设计并制造芯片。

And again, at the level of n equals one, like running a company that's just a giant pain in the butt because your secrets are walking out the door and your employees are walking out the door, and the whole thing sucks. But the other side of it is you also benefit from that, because you get to hire people with all these skills and experiences, and you're in this ecosystem that adapts and channels talents, skill, and knowledge into new fields. So there's a push and pull at the level of just being an individual CEO. At the level of just being in the ecosystem, to your point, it's an absolutely magical phenomenon. And by the way, for all the issues in Silicon Valley, I think AI is the ninth major technology platform in the history of Silicon Valley. Silicon Valley is still called Silicon Valley. We haven't made silicon here in decades. It's called Silicon Valley because they used to make chips. They used to have the actual fabs in Silicon Valley, and then they designed them and made the chips.

所以那是第一波，始于 20 世纪 50 年代，实际上那更像是第三波，但正是在那时这个地区被命名了。现在我们大概到了第九波。公司城镇现象——公司即行业——那种不确定的乐观主义：没有人需要坐下来规划说，好的，90 年代硅谷要做互联网，2000 年代要做智能手机，2010 年代要做云，2020 年代要做 AI。它就这么发生了。生态系统的灵活性所蕴含的不确定乐观主义与之相遇，硅谷得以变形进入所有这些类别。这或许再次证明了不确定乐观主义。

And so that was wave one starting in the 1950s, actually that was like wave three or whatever, but it was when the area was named. But now we're on like wave nine. And the company town phenomenon where the company is the industry, the indeterminate optimism—nobody had to sit and plan and say, okay, in the 1990s Silicon Valley is going to do the internet, in the 2000s they're going to do the smartphone, in the 2010s they're going to do the cloud, in the 2020s they're going to do AI. It just happened. The indeterminate optimism of ecosystem flexibility met that, and Silicon Valley could morph into all these categories. And again, maybe a testimony to indeterminate optimism.

这让我想起那个梗：我们都只是沙子上的包装层。我们构建的一切都只是包装、包装、再包装。

This reminds me of the meme of how we're all just wrappers over sand. Everything we're building is just wrapper wrapper wrapper.

包装这个说法太搞笑了。是啊。我现在是一家软件公司。我是芯片包装层，对吧？我是业务应用。我是数据库包装层。

The wrapper thing is hysterical. Yeah. I'm a software company now. I'm a chip wrapper, right? I'm a business application. I'm a database wrapper.

对，没错。我是沙子包装层。你我现在都是沙子包装层。

Yeah, exactly. I'm a sand wrapper. You and I are all now sand wrappers.

沙子包装层。

Sand wrappers.

完美。

Perfect.

好，再问一个问题。关于媒体消费，我问了你的合伙人 Ben Horowitz 该跟你聊什么。他说你最近很迷电影。

Okay, one more question. Along the media diet, I asked your partner Ben Horowitz what to talk to you about. And he said that you're really into movies these days.

是的。

Yeah.

所以我不知道有什么电影。你最近有特别迷的电影吗？有没有最近特别喜欢的？

And so I don't know any movies. Any movies you're really into these days? Any movies you've absolutely loved recently?

有。去年让我震撼到极点的电影，我认为绝对是这十年来最好的电影，甚至可能是过去 15 年最好的，就是这部。可惜的是，看过的人不多，但我强烈推荐。片名叫《Edington》。

Yeah. So the movie that blew my socks off last year, which I think is the best movie of the decade for sure and maybe of the last like 15 years, is this movie. Unfortunately, it's one of these things. Not a lot of people have seen it, but I would highly encourage it. It's called Edington.

没听说过。

Not heard of it.

你没听说过？好吧。你会很喜欢的。我不剧透太多。表面上看，以下内容不涉及剧透：故事发生在新墨西哥州一个叫 Edington 的小镇，大约 600 人。警长由华金·菲尼克斯饰演，是个老派、粗鲁的右派。镇长由佩德罗·帕斯卡饰演，是个年轻时髦的进步派。电影从 2020 年 3 月开始，也就是新冠疫情刚爆发的时候，然后展开到接下来的几个月，一直到 2020 年夏天。所以有乔治·弗洛伊德事件、抗议和骚乱。这是新冠疫情和 BLM 运动的交汇。还有第三个元素：一家公司，基本上就是 Meta 的松散伪装版，正在小镇郊外建一个 AI 数据中心。这个元素随着时间的推移越来越突出。这部电影很好地展示了这个新墨西哥小镇如何完全卷入新冠疫情、BLM 运动和技术焦虑，但所有人都是通过互联网来体验这些的。这正是实际发生的情况。我如此喜欢它的原因：第一，这是第一部直接面对 2020 年的电影，充分展现了当时全国上演的所有动态。第二，这是第一部出色地展现了生活在这样一个世界是什么样子——现实世界的事件通过互联网被体验，成为人们生活的中心。它把智能手机和社交媒体融入其中，而这是很多电影都难以做到的。整部电影以一种极其娱乐的方式呈现。我甚至不能说完全认同这部电影，我和导演可能有很多分歧，但他真的努力去呈现 2020 年代在美国作为一个人生活的真实感受，而很多其他有才华的导演都不敢碰这个题材。这家伙不知为什么，就是‘好吧，我要找到所有第三轨，然后抓住它们。’

Have you not heard of it? Okay. So, you're going to really enjoy it. So, I won't spoil too much of it. At the surface level, the following spoils nothing. It's set in a small town in New Mexico called Edington, which has about 600 people. There's a sheriff played by Joaquin Phoenix, who's like an old crusty basically right-winger. And then there's a mayor played by Pedro Pascal, who's basically a young hip progressive. The movie starts in March of 2020, when COVID first hits, and then it plays out over the next few months into the summer of 2020. So you have the George Floyd moment, the protests and riots. So it's the convergence of COVID and the BLM stuff. And then there's a third element: a company which is basically a loosely disguised version of Meta, building an AI data center on the outskirts of town. That looms larger and larger over time. The movie really shows how this small town in New Mexico gets fully wrapped up in all the COVID stuff, the BLM stuff, and the tech anxiety stuff, but they're all experiencing it through the internet. That's what actually happened. So the reason I love it so much: one, it's the first movie that directly grapples with 2020, fully engaging with all the dynamics playing out in the country. Two, it's the first movie that does a really good job of showing what it was like to live in a world where real-world events were experienced online, central in people's lives. It pulls in smartphones and social media in a way that movies really struggle with. And the whole thing comes together in an incredibly entertaining way. I won't even say I completely agree with the movie, and I think the director and I would probably disagree about a lot, but he really tries hard to grapple with what it's actually like to live as a human being in the 2020s in America, in a way that many other talented filmmakers have been scared of touching. And this guy, for some reason, he's just like, 'Yeah, I'm just going to find all the third rails and grab them.'

我明白为什么这是你今年的最爱了。

I can see why that's your favorite movie of the year.

很棒，很棒，很棒。每个人都应该看看。

It's great. It's great. It's great. Everybody should see it.

哦，天哪。好，最后一个问题，我想问问你的产品使用习惯。有没有什么你很喜欢但不太为人知的产品想推荐？如果你经常用，也可以提你投资的产品。

Oh, man. Okay, final question I want to ask about your product diet. Are there any products you use that maybe are less known that you love that you want to recommend? You can mention products you're investors in if you use them constantly.

我们投资了太多，很难挑出最喜欢的。就像问谁是你最爱的孩子。所以很难具体说。但我可以聊几个。一个观察：我 10 岁的孩子完全迷上了 Roblox。而且，这可不是我教的。你有孩子吗？

I mean, we have so many that it's really hard to pick favorites. It's like asking who's your favorite child. So it's hard to pull out specific ones. But I'll talk about a few. One observation: my 10-year-old is 100% obsessed with Roblox. And by the way, it was not from me. Do you have kids?

有。我有一个两岁半的孩子。

I do. I have a two and a half year old.

两岁半。好。那你还没遇到我现在遇到的情况：不管你做什么，都不酷，对吧？两岁半的时候，爸爸做什么都是世界上最酷的事。但我可以告诉你，等他到 10 岁，你做什么都超级不酷。我对此深有体会。所以如果我说‘哦，我们做 XYZ’，他就‘哦’。但当他发现什么东西，或者他朋友告诉他，那才酷。

Two and a half. Okay. So you haven't run into what I'm running into now, which is whatever it is you do is not cool, right? At two and a half, whatever daddy does is the coolest thing in the world. I can tell you by the time he's 10, whatever you do is deeply uncool. I'm highly aware of that. So if I mention, 'oh yeah, we work on XYZ,' he's like, 'okay.' But when he discovers something, or when his friends tell him about it, it's cool.

他完全没受我影响，大约三个月前自己发现了 Replit 和 vibe coding，现在彻底迷上了用 vibe coding 做游戏之类的东西，一玩就是几个小时。我亲眼看着这个现象发生，特别有意思。第二，我完全爱上了所有 AI 语音相关的东西，我觉得它绝对令人惊叹、妙趣横生。现在我在晚宴上最喜欢的把戏就是掏出 Grok 里的 Bad Rudy——就是那个邋遢老鼠浣熊头像。我觉得这超级好玩。还有一家叫 Sesame 的公司，去年因为那种极其亲密、充满情感的语音体验而走红。所以我认为语音产品非常棒。我也对语音输入特别着迷。最近那家公司被收购了，但我认为挂坠、可穿戴设备这类东西都会火起来。Meta 眼镜，我觉得会有一场可穿戴设备革命。我特别喜欢语音输入。我手机上现在有个叫 Whisper Flow 的应用，做语音转录，效果惊人地好。它是个语音转录功能，但你可以在转录的同时跟 AI 模型对话。它明白当你说“不，我要那边用项目符号，我要这个那个”时，你不是让它打出“我要项目符号”这几个字，而是真的理解你想要项目符号。这就是一个超级有用的绝佳例子。我认为语音模式的东西会非常棒。

And so he, through no interference on my part, discovered Replit about three months ago and discovered vibe coding, and is completely obsessed with vibe coding games and all kinds of things. He literally does it for hours, so I'm seeing that phenomenon play out, which is super fun. Two is I am just completely in love with all the AI voice stuff. I think it's absolutely amazing, hysterical. My favorite party trick at dinner parties now is to pull out Grok with Bad Rudy, which is a foul mouse raccoon avatar in the Grok app. So I think that's super fun. We have this company Sesame that went viral last year for incredibly intimate, emotional voice experiences. So I think the voice stuff is fantastic. I'm also super fascinated by all the voice input stuff. Most recently that company recently sold, but I think the pendants, the wearables, all that stuff is going to be big. The Meta glasses, I think there's going to be a whole wearables revolution here. I love the voice input stuff. I have this app on my phone now called Whisper Flow, which is voice transcription. It works staggeringly well. It's a voice transcription function, but you can actually talk to the AI model while you're doing voice transcription. So it kind of understands when you're telling it, 'No, I want bullet points over there and I want this and that.' It understands that you're not telling it to type the words 'I want bullet points,' it just actually understands that you want bullet points. So that's a great example of a super useful thing. I think the voice mode stuff is going to be really great.

我的新闻通讯订阅者可以免费获得一年 Replit 和 Whisper Flow 的使用权。那么，你儿子用 Replit 做过最难忘的东西是什么？

Subscribers of my newsletter get a year free of Replit and Whisper Flow. So there we go. What's the most memorable thing your son built with Replit?

哦，他最近迷上了《星际迷航》。到目前为止，他一直在写类似《星际迷航》模拟器的东西。到了《下一代》那部剧里，他们实际上为电脑设计了一套设计语言。因为如果你看原初系列，他们只有带灯的旋钮，并没有真正的设计。但到了《下一代》，他们确实有了一套 UI 设计语言。用 vibe coding 可以做的一件有趣的事就是，你可以说“给我做一个《星际迷航：下一代》风格的界面，用于某某用途”。它就会使用 LCARS 设计语言，用那种设计语言给你构建一个《星际迷航：下一代》风格的界面，但你可以选择比如一款《星际迷航》游戏。所以他对此非常着迷。

Oh, well, he's gotten super into Star Trek. So far he's been writing like Star Trek simulators. By Next Generation they actually had a design language for the computers. Because if you watch the original series, they just had knobs with lights and they didn't really have a design. But by Next Generation, they actually had a UI design language. One of the fun things you can do with vibe coding is you can say, 'Give me a Star Trek Next Generation user interface for whatever.' And it actually uses the LCARS design language. It'll build you a Star Trek Next Generation interface using that design language, but with your choice of a Star Trek game, for example. So he's going crazy for that kind of thing.

听起来太有趣了。你们应该开源或者发布那个东西。Mark，我本来可以跟你聊上几个小时，但你还有事要忙。在我们结束之前，有什么想对听众说的吗？有什么想强调的，或者想留给听众的话？

That sounds extremely delightful. You guys should open source or release that. Mark, I could talk to you for hours. You got things to do. Anything you want to leave listeners with before we wrap up? Anything you want to double down on or just leave listeners with?

嗯，有几件事。第一，上周我们非常幸运。Py McCormack 写了一篇关于我们的文章，实际上是有史以来写得最好的一篇。他发布了那篇文章，它对我们做什么以及我们如何思考做了最好的解释。所以我强烈推荐。另外，我们自己也在视频内容上投入了很多精力。我们现在有一个很棒的团队。我强烈推荐我们的 YouTube 频道，上面有很多很棒的内容，而且在接下来的一年里会非常精彩。

Yeah, a couple things. One is we got super lucky last week. Py McCormack wrote the best piece ever written about us, actually. He released it, and it's the best explanation of what we do and how we think. So I would definitely recommend that. And then we're putting a lot of effort ourselves into video content. We have a great team of folks now. I definitely recommend our YouTube channel, which has a lot of great stuff and is going to be very exciting in the next year.

太棒了。我们会附上链接。我记得是 YouTube.com/6Z 之类的。你们的内容很棒。Mark，非常感谢你来做客。

Awesome. We'll link to that. I think it's just YouTube.com/6Z something like that. And you guys have great stuff. Mark, thank you so much for being here.

太好了。谢谢你邀请我。我真的很感激。

Awesome. Thank you for having me. I really appreciate it.

大家再见。

Bye everyone.

非常感谢您的收听。如果您觉得本期内容有价值，可以在 Apple Podcasts、Spotify 或您喜欢的播客应用上订阅本节目。也请考虑给我们评分或留下评论，这能帮助其他听众找到这个播客。您可以在 lennispodcast.com 找到所有往期节目或了解更多信息。下期再见。

Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennispodcast.com. See you in the next episode.