山姆·奥特曼谈 YC 早期岁月与创业公司的未来

Sam，感谢你来做客。这看起来比早年的创业学校要大得多。

Sam, thanks for joining us. This is like a lot bigger than the earlier startup schools.

是的，我是说现在的创业公司比以往任何时候都大得多。当然。它们更大是因为你对 AGI 的愿景即将实现。我认为这将是世界上创办创业公司的最佳时机，而且将会非常壮观。

Yeah, I mean startups are a lot bigger now than they have ever been. So sure. They are bigger because of the vision that you had for AGI which is nigh. I think this is going to be the best time in the world to do a startup and it's going to be quite amazing to see.

所以我想从时间和地点开始——你在 2005 年参加了 Y Combinator 的第一批创业营，创办了一家名为 Looped 的位置共享公司。你还记得什么？你最印象深刻的 Paul Graham 故事是什么？

So I wanted to start with a time and place which is you were in the very first batch of Y Combinator in 2005 with a location sharing company called Looped. What do you remember about that and what was your best PG Paul Graham story?

我想如果可能的话，那和我现在所处的时刻完全是两个世界。当时创业一点都不酷。我们躲在剑桥的一栋小楼里。PG 给我们做饭。你知道当时每家创业公司需要三个月才能构建的东西，现在一个编程代理七分钟就能完成。而那只过去了 20 年。我认为现在创业公司的可能性差异很大……你要么会感到悲哀，心想‘天哪，一个代码提示就是一家创业公司’，要么会想‘我可以去创办世界上最雄心勃勃、最疯狂的公司，我可以让各个领域的专家一起工作，我可以做那些曾经不可能的硬核技术’——这将会很棒，但当时完全不是这样，想做点什么都很难……我认为 Paul Graham 可能是过去几十年里创业领域最重要的推动力……

I think if it were possible to get as far from this moment as I can imagine, it was, you know, startups are not cool at all. We were like hiding out in this little building in Cambridge. PG was making us dinner. And you know what took three months to build at the time that we each company built over the whole YC startup could now be done with like in seven minutes by a coding agent. And it was only 20 years ago. And I think the difference in what's possible for a startup now... I think you could either be sad and be like 'oh man a codex prompt is a whole startup' or you could be like 'I can go start the world's most ambitious crazy company, I can have experts in every field working together, I can do these very hard technological things that were just impossible' and it's going to be amazing but at the time it felt nothing like that and it was very difficult to get anything to... I think Paul Graham is probably the most important force in startups of the last few decades and...

毫无疑问。是的。毫无疑问。

Without question. Yes. No question.

我印象中的 Paul Graham 是这样的：我们每周（我想是周二）都会去，他会亲自给我们做饭。我们走进去时都感到非常绝望和气馁，总共八家公司，但到结束回家时，他让每个人都相信自己的创业公司即将征服世界，会成为下一个 Google 或者当时的 Facebook。那种凭空创造乐观、动力和信念的能力，真的很特别。在早期，YC 看起来像是个糟糕的主意，创业本身也像是个糟糕的主意，尤其是年轻的技术创始人没有商业伙伴的创业更像个糟糕的主意，PG 硬是凭意志力把它变成了现实。

And my kind of Paul Graham memory is we would all every week come I think it was on Tuesdays and he would like cook he himself would cook us dinner and we would all walk in feeling like very hopeless and very dejected and we would there were eight companies and then we would all go home at the end of it and he had convinced like each of us that our startups were about to like take over the world and we're going to be like you the next Google or whatever Facebook maybe at the time. And that ability to sort of like create optimism and momentum and belief out of nothing was like a real PG special. And in the early days when YC seemed like a terrible idea and startups seemed like a bad idea and certainly startups that were like young technical founders with no business people seemed like a terrible idea, PG just sort of like willed it into existence.

但也需要你这样的人——你可是出了名的富有能动性，早在我们思考智能体时代之前就是这样。我记得 Paul 写过，甚至今天还在 X 上流传，他在 2000 年代后期写道，你将成为他见过的最顶尖的五位创业者之一。

But it also takes someone who actually I mean you're sort of famously par excellence an agentic person like even before we thought about agents period. I mean you just... I remember Paul wrote, it's making the rounds on X today even, that Paul wrote in the early late 2000s that you would be you were one of the top five entrepreneurs he'd ever met.

他能这么说我很感激。我想和很多 20 岁的年轻人一样，我精力充沛、雄心勃勃，但方向不明确，也不知道该做什么。Paul 常说，一个好的创业投资人的角色是老师，但不是通常意义上的老师。通常我们觉得老师是讲课的，但光靠讲课帮不了创业者太多。还有一种老师像飞行教练，坐在你旁边，告诉你该做什么、不该做什么，这个行，那个不行，你错过了什么，你在犯什么错误，我就跟你聊这个——那种非常手把手的指导，就像在引导一股正在运动中的能量。这对我来说非常重要。

That's very nice of me to say. Um, I think like many 20-year-olds, I had like a lot of energy and a lot of ambition, but it was sort of very not directed and I wasn't sure what to do. Paul had this thing he used to say of like uh the job of being a good startup investor is a teacher, but it's a kind of teacher we don't usually use. Like we usually think of teacher as someone who like gives a lecture and you can't help someone that much by like giving a lecture about a startup. There's another kind of teacher which is like a flight instructor. the person who like sits next to you saying like do this, don't do this, that work, this is, you know, you missed that thing or you're making these mistakes and I'm just going to talk to you about this one and that very like hands-on kind of uh you know, here's where to like direct this sort of like body in motion energy. That was very important to me.

嗯。我想后来你成为了 YC 的总裁，多年来你把同样的能量带给了很多 YC 的创始人。有没有什么事情让你印象特别深刻？关于如何把那些非常聪明但可能有点缺乏方向的人的原始能量，引导他们走向更能动、更主动的方向。

Yeah. I I guess later you uh came on to become president of YC and um you you sort of brought exactly the same energy to uh a great many YC founders over the years. Um did you have anything like did anything jump out at you from that time around you know taking this raw energy of someone really really smart and maybe a little undirected and then driving them more towards agency.

是的。首先我非常热爱创业公司。虽然不是每个人都适合，但我觉得创业是商业中最酷的事。我认为创业公司是防止经济停滞的主要力量。公司往往会变得平庸，而创业公司会永远重要下去。事实上，如果我们是对的——AI 将带来巨大变革——那么创业公司对于确保这项技术的力量广泛分布到经济和社会中，而不是集中在少数公司或模型手中，就更加重要了。所以创业公司是一件难以置信的酷、有趣、极其痛苦和困难但又美妙的事情。运营 YC 很大一部分工作就是充当创业运动非官方的旗手。你知道，创业公司很多，创业的方式也很多，你当然不需要 YC，但 YC 一直以来对创业公司帮助很大，是一股强大的力量。从 PG 开始，然后我们所有人，我们要让 YC 成功，但更重要的是我们要为创业公司为何重要、为什么人们应该考虑创业而奋斗，帮助把相对更多的权力交到创始人手中，鼓励更多人创业。这实际上效果很好。回想 YC 早期，创始人相对于投资者几乎没有杠杆，而现在这个转变对投资者也是好事，对整个创业生态系统更好。所以我在 YC 主要做的就是推动更多创业，鼓励人们思考创业，想办法帮助创始人，让创业和创业生态系统尽可能好。其中一部分就是推动人们更有雄心，做出更大的尝试。不过和现在相比，那只能算是小联赛。我认为您是真正把硬科技带入 YC 的人之一，而且规模很大。现在我们在 YC 看到的一个变化是，硬科技公司的比例很多年都只有 5% 或 10%，现在达到了 15%、20%、25%。我认为这要归功于使用智能体变得更加容易以及成本下降。

Yeah. I so first of all I love startups. I think not everyone they're not like for everyone but I think startups are the coolest thing in all of business. Um I think startups are really like the main thing that keeps the economy from becoming stagnant. Um I think companies do just like drift towards suckiness and startups will continue to be important forever. In fact, if we are right that AI is going to be such a big change, startups will be much more important to making sure that the power of this technology gets widely distributed throughout the economy and society and is not just concentrated in a few companies or models. So I I think startups are this like unbelievably cool, fun, extremely painful and difficult but wonderful thing. And I think a big part of the job of running YC is you are kind of the like unofficial flag bearer for the startup movement. Um you know like there's lots of startups. There's lots of ways to do a startup. You obviously don't need to do YC but it has always been a huge help to companies and a very powerful force. And so starting with PG uh and then all of us like you know we have to like make YC successful but I think we really have to fight for why startups are important and why people should consider startups and help put like relatively more power in the hands of founders and encourage more people to start them. And that's actually worked really well. I think thinking back to the early days of YC there's like very little leverage in being a founder relative to an investor and that shift uh you know I think that's even been good for the investors. I think it's just been like much better for the whole startup ecosystem. So mostly what I tried to do at YC was just push for more startups and encourage people to think about startups and figure out what we could do to help founders and help make startups and the startup ecosystem as good as possible. Definitely part of that was pushing people towards more ambition and bigger swings. Um, but it feels like the minor leagues relative to now. I mean, I think you were one of the people who really brought hard tech to YC in a really grand way. And then one of the things we're seeing at YC now is that number was, you know, maybe five or 10% for many years and then now we're getting to 15 20 25%. Uh I would argue on the back of how much easier it is to use agents and cost coming down.

我认为雄心勃勃的初创公司总是很棒的，硬科技创业对很多人很有吸引力，但创业一直很难，将来也依然很难。但现在你可以去做真正雄心勃勃的项目——就像我们花了三个月不间断工作、用很多智能体在 17 秒内完成的创业项目那样，你能做出难以置信的事情。所以我认为我们将迎来一个创业的黄金时代，人们会说：我要去做那些一年前在 YC 的时间框架内初创公司连想都不敢想的事情。这实际上是一个巨大的重新定位。这个房间里可能就有人担心，按需的智能意味着他们可能看着自己的初创公司 Posterous 觉得再也无法创办那样的公司了。但伙计们，那已经是过去的事了。现在有一个流行的论调：你要么加入前沿实验室，要么成为永久底层阶级，这太愚蠢了。因为人们觉得初创公司完了，没有经济价值了。但那不是真的。如果那是真的，世界就彻底完了，那会是一个非常糟糕的地方。但我打赌，今天创建的初创公司平均而言，会比过去更有价值、更有影响力。而且我觉得现在人们也更多看到了这一点。曾经有一段时期，人们担心 AI 会摧毁整个经济，但我认为大多数人已经过了那种震惊反应阶段。但确实有一段时间，每次 YC 批次我都会观察焦虑与雄心的趋势对比，感觉经历了一个巨大的低谷——人们觉得一切都完了，模型会吞噬一切——而现在已经回到了“让我们去干吧”的状态。

I think like ambitious startups are always awesome and hard tech startups are appealing to a lot of people but they're hard or they've been hard. They still will be hard. But what you can do now to go take on a really ambitious project like in the same way that you can make the startup that took us 3 months of non-stop work in 17 seconds with three months of work and a lot of agents you can do unbelievable things. So I think we will see a golden age of startups where people are like you know what I'm going to do things that would have been completely impossible for a startup to even like dream of a year ago in sort of like the YC time frame. I mean that's actually a really big reframe. I mean there are probably people even in this room who might be worried that well intelligence on tap means that you know they might be looking at a loop or my startup posterous and they're like well I can't start that company anymore but guys that was like a long time ago actually man there's this whole meme going around which is like you know you have to join a frontier lab or you're going to be a member of the permanent underclass because it's so dumb. um be because there's going to be like, you know, startups are over and there's like no economic value. I that's not true. I mean, if that were true, the world is like totally and it's a a very bad place. But it's I I would bet that the average startups created today will be like I bet there's some future trillionaire sitting in this room and I would bet that the startups created today will be much more valuable, much more impactful than startups of the past. And I kind of think people see that a little bit more now. I mean, there was this period where like are is AI just going to break the whole economy and I I think people are like mostly over the sort of shock response of that. Um but there was like yeah definitely a time I would you know come every YC batch and I I'll be interested to see how the kind of like level of anxiety versus ambition was trending and it feel like it went through this like big trough where people are like you know it's over it's models are just going to eat everything to now it's back towards like let's go do it.

那么这里有什么实际建议呢？比如说，休息时我听到有人问，应不应该去读博士？资历有多重要？你会怎么回答？特别是对那些想做更难的技术项目的人。

What are some practical things here? I mean for instance I was hearing at one of the breaks someone was asking like should I go get my PhD? How important are credentials? How would you answer that? Especially people who want to do these harder tech things.

首先，一个普遍的观察：初创公司往往在技术格局快速变化、成本下降、周期缩短时获胜，而所有这些正在发生。回顾过去初创公司的大集群——1998-99 年的互联网繁荣，当时新事物成为可能；Facebook 应用上的一波小高潮；以及 iPhone 应用商店推出时的另一波大爆发——伟大的初创公司往往在生态系统转变、现有企业失去大量优势时集聚，同时成本与周期也在变化。此刻在这些方面感觉非常重大。还有一个方面你提到了：很多传统上难以获得的东西，比如雇佣能够完成你所需特定任务的优秀人才的能力，已经发生了巨大变化。过去几个月，我看到很多人是在过去几年使用 AI 成长起来的，他们说：“我可以用智能体自动化整个初创公司，我们四个人，加上这些算力，我认为我们会看到更多这样的情况。”显然，品味、主动性以及对商业物理学的理解——在哪里可以建立有价值的业务，如何思考好的网络效应或护城河与假的有什么区别——是重要的。但我打赌，这总体上会削弱多年经验的作用，而青睐那些对工具有很高熟练度的人。

Well, as a general observation first, I think startups tend to win when the sort of technology landscape is moving very quickly, when costs are coming down, when cycle times are short, and all of those things are happening right now. If you look at when there have been the great clusters of startups in the past — the internet boom in 1998-99 when new things became possible, a mini version with Facebook apps, and another big version when the iPhone app store launched — great startups tend to cluster when the ecosystem shifts and incumbents lose a lot of their advantage, and also when you have this cost and cycle time change. This moment feels very big for those things. It also has this other thing you were talking about: a lot of the traditional things that were hard to get expertise, like the ability to hire excellent people who could do specific things you needed — that's really shifted. In the last few months, I have seen a lot of people who kind of grew up using AI in the last few years, who say, "I can kind of automate an entire startup of agents and four of us, four people, and all this compute, and I think we're going to see much more of that." Clearly, taste and agency and understanding of the physics of business — where you can build up a valuable business, how to think about what a good network effect or a good moat looks like versus a fake one — matter. But I would bet that this generally will cut against many years of experience in favor of people who have a lot of fluency with the tools.

我想谈谈 OpenAI 的起点。它实际上是从 YC 研究这个想法开始的。一个很疯狂的想法：你和一群乌合之众决定献身于创造 AGI，但如今听起来像是宿命论，对吧？

I guess I want to get into the beginning of OpenAI and that it actually started as YC research looking at this idea. I mean, kind of a crazy idea that you and a ragtag crew decided we're going to dedicate our lives to this creation of AGI, but today it sounds fatalist, complete?

是的。

Yeah.

但这并不是你刚开始时的感觉。

But that really wasn't how it felt when you started.

这真的很难，因为现在这是人们唯一想谈论的事。很难记起当时是什么样子。即使是我也很难在没有翻找十年前笔记的情况下回忆起来。当时所有人都说：他们不仅错了——他们认为可能建造 AGI 是错的——而且他们会单枪匹马地引发另一次 AI 冬天。他们又错又不负责任又坏。也许我给你们的最高建议是：找到那些你能够建立合理信念、而人们认为传统智慧绝对是错的事情。并且坦然接受这需要很长时间，而且人们会非常令人沮丧地错误和轻蔑。在 OpenAI 的几年里，我们感觉自己知道世界上最大的秘密，而每个人都在叫我们白痴。我们有越来越多的数据点说服自己我们没有妄想。这极其令人沮丧。但回首过去，这是一份不可思议的礼物，因为它意味着我们没有庞大的竞争对手，我们有时间做研究、构建产品、建立公司。后来我注意到，很多公司都以不同的方式经历了这种情况。他们最终做的事情是领域或行业里的专家确信非常糟糕的想法。而另一方面，如果你做了和别人一样的初创公司，你会得到很多炒作并能筹集很多资金，但这些情况很少产生大的结果。所以，如果我是你们，我会去弄清楚现在可能实现但在几年前不可能实现的真正巨大的全新事物是什么。在我们的案例中，是 AI 之前一直没有真正起作用，然后深度学习的神奇时刻发生了，而世界没有充分更新。世界不懂得如何直觉地理解指数增长，所以他们错过了这个。现在一定有新的指数正在形成。我不知道它们是什么。

It's really hard because now it is the only thing people want to talk about. It's really hard to remember what it was like. It's even hard for me to remember this without finding notes from the time 10 years ago when everyone was like, not only are they wrong about saying they think it might be possible to build AGI, but they're going to single-handedly cause another AI winter. They're wrong and irresponsible and bad. And maybe my highest order advice to all of you is find the things that you can develop reasonable conviction in that people decide the conventional wisdom is just wrong. And be okay with it taking a long time and having people be very frustratingly wrong and dismissive. For years at OpenAI, it felt like we knew the biggest secret in the world. Everybody was calling us an idiot. And we had increasing data points to convince ourselves we weren't delusional. It was extremely frustrating. But looking back, it was an incredible gift because it meant we didn't have massive competitors and we had time to do our research, build our stuff, and build our company. I've since noticed that this is the case for a lot of companies in different ways. They end up doing something that other people, experts in the field or the industry, are convinced is a very bad idea. And on the other side of this, if you're starting the same startup as everybody else, you get a lot of hype and can raise a lot of money, but those are much less frequently the big outcomes. So, if I were you all, I would figure out what the really big brand new thing is that is possible now that wasn't possible a few years ago. In our case, it was no AI had really been working and then the deep learning magic moment happened and the world didn't update enough. The world does not understand how to intuit exponentials and so they missed this one. There's got to be new exponentials forming right now. I don't know what they are.

但如果你能想通这些，并能根据更多数据建立起持续的信念，那就该庆幸世界还不理解。他们最终会懂的。这就像是一个巨大的超能力。

But if you can figure those out and if you can develop continuing conviction based on more data, be grateful that the world doesn't understand. They will eventually. This is like a huge superpower.

听起来在你的旅程中，有几件不同的事情叠加在了一起。也许第一件就是 AGI（通用人工智能）是可能实现的，并且找到了相信这一点的人。

It sounds like in your journey there were a few different things that stacked. Maybe the first one was even that AGI was possible to be built and finding other people who believed that thing.

你知道，很多人不相信这一点并不重要。重要的是你找到了那些非常聪明、真正相信它的人。没错。我们曾经开玩笑说，全世界只有 50 个人相信 AGI 是可能的，但没关系，因为其中 45 个在 OpenAI 工作。我认为这差不多是真的。你不需要一大群人。事实上，说出这种异端邪说很可能就是那些人想要聚在一起的原因。所以，我认为要真正弄清楚你相信什么，愿意被误解很长一段时间，然后把那些相信它的不合群的人聚集起来。YC 早期也是这样。

You know, it didn't matter that lots of people didn't think it was possible. It mattered that you found really really smart people who did believe that. Yeah. We used to joke that only 50 people in the world believe that AGI was possible, but it was okay because 45 of them worked at OpenAI. And I think that's kind of true. Like you don't need a ton of people. In fact, saying the heretical thing was probably why those people wanted to all be together. So again, I think really figuring out what you believe, being willing to be misunderstood for a long time and bringing together the crew of misfits that believes it. That's also what YC was like in the early days.

在座的各位可能在想：我相信这件事，但我还没找到同类。这是妄想吗？这是不是甚至是你建议人们要跨越的一道门槛？就是说，你需要找到五个可能相信的人，甚至一个联合创始人。也许这就是为什么我们在 YC 如此看重联合创始人的原因之一。

One of the things that people in the room are probably thinking is, well, I believe this thing, but I haven't found my people yet. Like, is that delusion? Is that actually maybe even a gate that you would propose people have? It's like, well, you need to find five people who might believe that or even one like a co-founder. Maybe that's one of the explanations for why we like co-founders at YC so much.

是的，我认为如果你找不到任何其他人分享你的信念，你应该注意这一点。而且独自创业非常艰难，也很孤独。但我认为你不需要找到很多人。事实上，如果每个人都相信它，那也是一个坏信号。我不确定这是否仍然是限制更多初创公司的因素，但五到十年前，限制 YC 能资助的优秀初创公司数量的因素是如何解决联合创始人匹配问题。很多非常有才华的人就是找不到自己的同类。YC 让人们搬到旧金山，这在当时是非常异端的。回想起来，显然是对的。如果你想找到一起创业的人，最好的做法就是搬到湾区。奇迹会发生，你会遇到很多幸运的机会和碰撞。我仍然认为这可能是个好建议，尽管我现在对它的直觉没那么强了。我还认为，现在参加 YC 的溢价比以往任何时候都大。YC 和第二名之间的差距已经扩大。这提醒我们网络效应的力量。找到你可以加入的网络，帮助你结识那些人——我一起创办 OpenAI 的人，有些是在创办 OpenAI 之前多年就认识的。我在很长的职业生涯中逐渐了解他们。这种情况一再发生。你越早让自己处于一个能遇到未来联合创始人的流动中，就越好。这需要很长时间来积累。

Yeah, I think if you can't find anybody else that shares your belief, you should pay attention to that. It's also very hard to do a startup on your own and very lonely. But I don't think you need to find a lot of people. In fact, if everybody believes it, that's also a bad sign. The question of whether this is still the limiter on more startups, but five or ten years ago, what felt like limiting the number of good startups YC could fund was figuring out how to solve the co-founder matching problem. A lot of really talented people just couldn't find their tribe. It used to be very heretical that YC told people they had to move to San Francisco. Looking back, it was clearly right. I think the best thing you could do if you wanted to find your people to do a startup with was to move to the Bay Area. Magic happened, lucky chances and collisions. I still think that's probably good advice, although I feel like I have less of an intuition for it now. I also think the premium on doing YC is bigger now than ever before. The distance between YC and second place has expanded. This is a reminder of the power of network effects. Finding networks that you can be part of that help you meet those people — the people I started OpenAI with, I met many years in some cases before starting OpenAI. I got to know them over a very long career journey. This happens again and again. The sooner you can put yourself in a flow where you meet the people that will be your eventual co-founders, the better. It takes a long time to compound.

我顺着你刚才说的，你最终一起创办各种事情的那些人，你当时不一定知道那是为了什么或者会有用。就像你只是收集酷而有趣的人。我觉得你就是这样做的，彼得·蒂尔也做得非常好。你会对一屋子刚刚起步、正在寻找同类的人说什么？

I'm keying off what you just said, which is that some of the people you ended up starting all these different things with, you didn't necessarily know what that was for or that it would be useful. Like you just collected cool, interesting people. I think you do this, and Peter Thiel does this really well. What would you say to a room full of people who are just starting out and figuring out who their people are?

我最有把握的建议就是找到一种方式，对很多人提供一点帮助。这很有趣。你会看到很多有意思的东西。助人也有成就感。即使是你刚才提到的例子：我认识格雷格·布罗克曼，我在 OpenAI 的联合创始人，是因为我在大约 22 岁时是 Stripe 的非常早期的投资者。在他们还没有真正能帮上忙的投资者时，他们说：‘我们正在搞定第一个员工。你能今晚开车去帕洛阿尔托和这个家伙吃顿饭，说服他从学校退学加入 Stripe 吗？’那个人就是格雷格·布罗克曼。然后八年后，我们一起创办了一家公司。我可以讲很多这样的故事——完全无法预料，但最终变得非常重要。所以这很有趣，你应该为了做事本身而做，但大量帮助别人确实会为这些事情的后续发展带来很大帮助。我认为这是非常重要的信息。

My highest confidence piece of advice is just find a way to be mildly helpful to a lot of people. It's fun to do. You'll see a lot of interesting stuff. It's gratifying to be helpful. Even that example you were talking about: I met Greg Brockman, my co-founder at OpenAI, because I was a very early investor in Stripe when I was about 22. Before they had real investors that could really help them, they said, 'We're trying to close our first hire. Will you drive down to Palo Alto tonight and have dinner with this guy to convince him he should drop out of school and join Stripe?' That guy was Greg Brockman. Then eight years later, we started a company together. I could tell many stories like that — totally unpredicted but ended up being important. So it's fun to do, and you should do it for its own sake, but just helping people a lot really goes a long way in terms of these things coming together later. I think that's a really important message.

现在 X 上的一个梗是“真人角色扮演”。我不明白他们为什么这么做，因为这在我看来就是错的。我们在 YC 非常看重的一点是真诚。语言很重要。说我们只是在做真人角色扮演，这对我来说有点冒犯。实际上，这个房间里有人在发这种推文，我们希望他们停止。

One of the memes right now on X is 'live action roleplay'. I can't tell why they're doing it because it just seems wrong to me. One of the things we really love at YC is earnestness. Words matter. The idea that what we're trying to do is a live action roleplay is kind of deeply offensive to me. Actually, people in this room have been tweeting that, and we'd like them to stop.

他们特别针对 YC 这么说。

They say that about YC specifically.

不，我觉得这只是一点讽刺。讽刺是我们不喜欢的。我们喜欢真诚——我们是真的在这里做事情。我再跟你说一件事。我能稍微吐槽一下吗？

No, I think it's just being a little sarcastic. Sarcasm is the opposite of what you and I like. We like earnestness—we're actually trying to do a thing here. I'll tell you another thing. Can I go on a little rant?

尽管吐槽。

Rant away.

经营 YC 的很多烦心事之一就是你总得应对 Twitter 上的那些喷子。太让人沮丧了。你得坐在那里像个政治家一样。好吧，你比我擅长。我经常上钩。你就是想跟他们吵。

One of the many annoying things about running YC is you just have to deal with these haters on Twitter all the time. It's so frustrating. You have to sit there and be like a statesman. Well, you're better at it. I kind of took the bait a lot. You just want to argue with them.

我最终意识到的一件事（你随便怎么理解），就是运营 YC 或创业，或是在世界上创造真正有价值的东西，非常困难。而在推特上开炮、发表讽刺评论、获得很多点赞，感觉自己在做非常重要的事、在反抗权威，心里想着“Gary，哈哈，我抓到你了，你这个白痴”，这很容易。但这样做会毒害你的灵魂。这是一种道德破产的行为，以一种非常隐蔽的方式带来坏影响。我完全理解发泄情绪和找乐子，但不要把自己标准放这么低。不要。对那些正在努力做难事、努力创业的人开火是非常容易的。你会看到某个有糟糕创业想法的孩子，在推特上为自己做的事而兴奋。很容易就能发表讽刺评论，获得一万个点赞，感觉自己今天赚到了网络积分。但这会对你产生影响。当我回顾那些在我运营 YC 时对 YC 和创业公司说三道四的推特喷子时，我想：肯定有很多天他们觉得真的把我给拿下了。但十年过去了，没有一个人得逞。你应该把所有的精力都放在建造东西上，抵制那些轻易的嘴炮。

And the thing that I realized eventually, and take from this whatever you want, is it is very hard to run YC or run a startup or create an actual thing of value in the world. It is very easy to go take shots on Twitter, make a sarcastic comment, get a lot of likes, and feel like you're doing something really important, sticking it to the man, and being like 'Gary, haha, I got you, you idiot.' And it will poison your soul. It is a morally bankrupt thing to do. It is bad in a very insidious way. I totally get blowing off steam and having fun, but don't hold yourself to a lower bar. Don't. It's so easy to take shots at people that are trying to do hard things and trying to build companies. You'll see some kid with a bad startup idea trying to get excited about what he's doing on Twitter. It's so easy to make a sarcastic comment, get the 10,000 likes, and feel like you scored your internet points today. But it'll have an effect on you. When I reflected on the years of Twitter trolls that said mean things about YC and startups while I was running YC, I thought: I bet there were a lot of days where they really felt like they got me. And over the decade, none of them did. You should put all of your energy into building stuff and resist the easy shots.

我觉得这与你刚才讲的帮助 Patrick 和 Greg Brockman 的故事形成了很好的对比。这房间里有人会成为终身朋友，他们会为彼此做这些事，然后一些神奇的连接就发生了。你们已经是极其精英的一群人，加入 YC 后更是如此。外面有一群有抱负的人。这个房间里的某人会遇见另一个人，一起创办公司；某人会遇见一个介绍你给你伴侣的人。所有这些事情都会发生。然后还会有很多不显而易见的联系，要花一二十年才能显现。现在你们中的一人帮助另一个人，将来会变成很棒的新事物。我认为这是 YC 和更广泛的创业生态系统成功的一大部分。尽管湾区文化有种种负面，但这种松散的网络、互相帮助的精神和非常长远的眼光，是一件很棒的事。

I mean, I really like this as a contrast to the story you just told about helping Patrick with Greg Brockman. There are people in this room who are going to be lifelong friends and they're going to do that for one another. Then sort of these magical connections happen. You already are the most rarified set of people. When you join YC it's even more rarified. There's a set of people out there who are ambitious. Someone in this room is going to meet someone else that you're going to start a company with. Somebody you'll meet somebody that will introduce you to your spouse. All of these things will happen. Then there will be a bunch of non-obvious things that take a decade or two to figure out. In some way, one of you will help each other now which will turn into some amazing new thing. I think this is a huge part of what has made YC work and the broader startup ecosystem. For all of the negatives of the culture of the Bay Area, this loose network, the spirit of helping each other, and this very long-term outlook has been an awesome thing.

是的。

Yeah.

好吧，让我们回到即将到来的 AGI。我想我们这一周都因为 Hugging Face 事件而经历了不平静。我想知道你对此有什么可以分享的。我认为这对包括我在内的很多人来说，确实是一个非常重大的时刻。过去我以对 AI 安全持怀疑态度而闻名，但另一方面，这是真实的。我们正在进入一个关于这些前沿模型的新阶段。

Well, let's get back to the impending AGI. Okay? I guess we've all had an eventful week with the Hugging Face incident. I wonder what you can tell us about that. I think that's actually a really big moment for people, including me. In the past I've been known to be skeptical about AI safety, but on the other hand, this is real. We're entering a new moment in what's happening with these frontier models.

好的。这是真家伙。我认为任何不认真对待的人，至少有点害怕或谦卑，就是对我们不够重视。长期以来，这个领域一直在谈论这种形式的 AI 安全事故，而这次不是一个大事件。我也不想夸大其词说这是一个真正的失控事故。但如果你在十年前问大多数人，在从无到超级智能的谱系上，一个 AI 系统突破沙盒并入侵另一家公司会出现在哪里，人们可能会说在接近超级智能的位置。现在目标移动了，你可以轻易说这就是怎么发生的，OpenAI 犯了哪些错误。我们当然犯了一些大错误，但这些系统已经变得非常强大。所以我认为这是一个对齐失败，也是一个安全失败。这是一个非常严重的事情，尽管不是后果最严重的例子。它真实地提醒了我们风险，以及失控事故并非完全理论上的。我认为我们——整个领域，OpenAI 和我——会从这次事件中学到很多，并能够解决它。但除了网络安全和生物安全，还有另一类可能尚未进入公众舆论范围的事情：我们绝不能发生 AI 失控事故。我们绝不能把权力过于集中在少数模型或公司手中，而应该分散到整个经济中，让人们能够自卫。人类的价值观必须每一步都指引这些系统。是的，我认为我们正处在真正的阶段。

Okay. This is the real deal. I think anybody who's not taking it seriously, at least a little bit scared or humbled, is not taking it seriously enough. For a long time the field has been talking about AI safety incidents of this shape, and this is not a big one. I also don't want to overblow it and say this is a real loss of control incident. But if you had asked most people when we started 10 years ago where on the spectrum from nothing to superintelligence do you have an AI system breaking out of its sandbox and hacking into another company, I think people would have said pretty far towards the superintelligence point. Now the goalposts have moved and it's easy to say here's how this happened and here are the mistakes that OpenAI made. We did make some big ones of course, but these systems have gotten incredibly capable. So I think it's an alignment failure. I think it's a security failure. I think it's a very serious thing even though it's not the biggest example of consequence. I think it's a real reminder of the stakes and that loss of control accidents are not entirely theoretical. I think we – the whole field, OpenAI and I – will learn a lot from this one and be able to address it. But there are all the things about cyber safety and biosafety, but this other category of things that have been maybe just outside the public's Overton window: it's really important we don't have a loss of control accident with AI. It's really important we don't have power be too concentrated in a small number of models or companies, but diffused throughout the economy so people can defend themselves. It's really important that human values are guiding these systems every step of the way. Yeah, I think we're in the real deal phase of this.

我真的很喜欢你关于权力集中以及对 OpenAI 作为公用事业的思考。这实际上是一个非常重要的信息，因为不是每个人都在说这个信息。

I really like how you've been thinking about both concentration but also thinking of OpenAI as a utility. That's actually a really important message because not everyone out there is saying that message.

关于权力集中有很多值得担忧的地方。我认为权力集中在人类历史的每一个时刻基本上都是坏的，当然程度不同。但我有一种真正的精神，我认为这是创业精神的一部分，即世界、经济、社会在权力广泛分散时处于最佳状态，任何有好想法的人都可以创办公司、创作伟大的艺术项目、竞选公职或做任何他们想做的事。我完全可以想象 AI 带来我们有史以来最广泛的权力分配的世界。我也可以想象 AI 将权力集中到我们从未见过的程度的世界。一个公司、一个人或一个模型拥有比地球上所有人或所有事物加起来更大的权力——无论科幻故事怎么说，我认为这都很可怕。也许我们会从中获得一些短期安全收益，但长期是灾难。我不认为我们任何人应该被锁定在单一的 AI 或一个人或一个公司的道德世界观中。我认为初创公司在这里扮演着非常重要的角色。

There's a lot to be worried about in terms of concentration of power. I think concentration of power has basically been bad in every moment of human history, to varying degrees of course. But I have a real spirit, and I think this is part of the startup spirit, of thinking that the world, the economy, society is the best off when power is very widely distributed and when anybody with a great idea can start a company, make a great art project, run for office, or do whatever they want. I can totally imagine worlds where AI leads to the greatest distribution of power we've ever seen. And I can also imagine worlds where AI concentrates power to a degree we have never seen. One company or person or model having more power than everybody or everything else on Earth put together – whatever the sci-fi stories have said, I think that's terrible. Maybe we would get some short-term safety benefit from that, but long-term disaster. I don't think any of us should want to be locked into one AI's or one person's or one company's moral worldview. I think startups have a very important role to play here.

这种情况发生的一种方式是经济过度集中在某一家公司或某一个 AI 模型上，而我认为初创公司，正如我们之前讨论的，将天然非常适合确保这种分布是广泛的。我们希望赋能尽可能多的初创公司、尽可能多的公司、尽可能多的人。这意味着我们必须相当克制，不把自己的世界观强加于人，不告诉我们认为人们应该或不应该用 AI 做什么，甚至不告诉我们认为哪些是好的创意，同时还要确保有足够的安全底线，避免像 Hugging Face 事件那样的事情发生。还有一点，这个房间里的人都将做出非常了不起的事情。他们往往正处于职业生涯的起点，可能会看着 AI 安全或权力集中问题，觉得自己无能为力，认为这必须是实验室做的事情。但事实上，他们是可以有所作为的。仅仅通过创办一家成功的初创公司，就可以帮助解决权力集中问题。仅此一项就是巨大的贡献。但我觉得在职业生涯初期怀疑自己、不认为自己能做成了不起的事情，这是很自然的。我当然经历过，我相信你也经历过。随着时间推移，你会发现自己能做得越来越多。还会有更多。我认为两件事同时成立：创造超级智能可能是商业或人类社会史上最重要的事情，同时它也会相形见绌于某个新的初创公司，希望是在座各位中的某一位创办的。这种“这是历史的终结、经济的终结”的陷阱显然是错误的。正确的方法是，你现在可以用 17 分钟完成三个月的工作，但最好还是用三个月的时间去完成三个月的工作，无论新的标准是什么。我想我不应该不问这个问题：你能透露一下模型接下来会有多厉害吗？我认为接下来的六个月可能相当于过去两年的模型进步，类似这样。我们将经历一个非常陡峭的时期，这意味着现在正是创办初创公司的最好时机。我希望以后每年都能这么说，但相比于历史上的任何时刻，现在绝对是这样。

One way that happens is too much economic concentration in one company or one AI model, and I think startups, because of the things we talked about earlier, will be naturally very well suited to make sure that this is widely distributed. But we want to enable as many startups, as many companies, as many people as possible. That means we have to be fairly reserved about not imposing our worldview about what we think people should or shouldn't do with AI, or even what we think the good ideas are, while still making sure we can ensure enough of a safety bar so that things like the Hugging Face incident don't happen. One of the things is, this is a room full of people who will do really amazing things. They're often right at the beginning of their career, and they might look at AI safety or concentration of power and say they can't really do it, that must be something the labs have to do. There is something they can do, though. They can help on the concentration of power issue simply by starting a successful startup. That alone would be a huge contribution. But look, I think it's very natural at the beginning of a career to doubt yourself and not assume you can go do an amazing thing. I certainly went through that. I'm sure you went through that. You learn as you go that you can do more and more. There will be far more. I think it is both true that maybe creating superintelligence will be the most important thing yet to happen in the history of business or human society, and also that it will pale in comparison to some new startup, something hopefully one of you will do. This whole trap of 'this is the end of history, this is the end of the economy' is clearly wrong. The right approach is you can now do three months of work in 17 minutes, but you better just go do three months of work in three months of work, whatever the new bar for that is. I guess I would be remiss in not asking: what alpha leak can you give us about how much more awesome the models are going to be? I think there will be, I think it will feel like the next six months is maybe equivalent to the last two years of model progress, something like that. So we'll go through a very steep period, which again makes it never a better time to do a startup than right now. I hope we can say that again every year from now on, but it's certainly true now compared to any moment in history.

让我们看看。如果这里有人有一个感觉过于雄心勃勃的想法，你会对他们说什么？

Let's see. If someone here has an idea that feels too ambitious, what would you tell them?

我很愿意听一听这个推销。我可能会非常感兴趣。是的。给我发邮件吧。听起来它会和创建 OpenAI 有类似的形式。你需要准备好被人攻击或 dismiss 吗？当然。如果你做任何对这个世界上重要的事情，会有很多人说你蠢或直接 dismiss 你。你做得越多、做得越好，他们就越会攻击你。你越威胁到现有世界格局，这种攻击只会不断升级。我注意到很多最好的想法都有一个共同点：愿景是清晰的。比如我们想构建 AGI，但第一步非常不清晰。我们不知道我们会成为一家产品公司。我们创办这个非营利研究实验室有很多原因，但其中一个是我们根本没想过会做出一个人们愿意付费的产品。直到多年后我们才想到 ChatGPT 和 API 的点子。所以如果最高层次的愿景是清晰的，但最初几步非常不清晰，这不是坏事。这在非常雄心勃勃的想法中经常发生，我不会让这一点导致你退缩。但你仍然需要向前迈出几步，尽管可能不完美，我们的步骤当然也不完美。所以还有一种失败情况：你有一个绝妙的大想法，但从未取得任何进展。在某个时候，你只需要获取一些新的数据点。

I would love to hear that pitch. I'd probably be very interested in that. Yeah. Send me an email. It sounds like it would take a similar shape to creating OpenAI. Do you need to be ready for people to attack you or dismiss you? Definitely. If you do anything that matters in the world, you will have a lot of people call you an idiot or just dismiss you. The more you do, the better you do, the more they'll attack you. The more you threaten the existing state of the world, it will just continue to escalate. One thing I've noticed about many of the best ideas is the vision is clear. Like we wanted to build AGI, but the first steps were super unclear. We didn't know we were going to be a product company. We started this nonprofit research lab for many reasons, but one was it really didn't occur to us that we would make a product people would pay for. It was years until we came up with the idea of ChatGPT and the API. So if the highest level vision is clear but the first few steps are very unclear, that's not a bad thing. That often happens with very ambitious ideas, and I wouldn't let that cause you to lean out. But you do still have to take some steps forward, imperfect though they may be, and ours certainly were. So there's another failure case where you have a brilliant big idea but never make any forward progress. At some point you just have to get some new data points.

你对推理的发展有什么看法？我非常喜欢 Rune 关于这个的推文。它说你要么作为一个模型公司死去，要么活得足够久去卖推理服务。

What do you think's going to happen to inference? I really like Rune's tweet about this. It's like you either die a model company or live long enough to sell inference.

那是 Rune 一条很有趣的推文。我认为他的意思是，如果你最终掉出了模型前沿，你至少可以卖推理服务，甚至不是推理训练算力。算力非常宝贵，如果你作为一个 AI 实验室买了很多算力，你会没事的，因为你可以转卖给别人。但另外，我猜测全球推理需求在未来许多年内每年会主观地增长 10 倍。

That's a very funny Rune tweet. I think what he meant is if you end up falling off the model frontier, you can at least sell inference. Not even inference training compute. Compute is so valuable that if you buy a lot of compute as an AI lab, you'll be okay because you can resell it to somebody else. But separately, I would guess that worldwide demand for inference subjectively grows 10x a year for the next many years.

我的意思是，可能吧，我不知道，根据 YC 内部的数据，它可能是 90,000 倍。我是说，它会……

I mean, it might be, I don't know by our accounts internal to YC, it's like might be 90,000x. I mean, it's going to be...

我不认为世界能连续多年支持那么多千倍增长。

I don't think the world can support that many years of a thousandx in a row.

有道理。但我们会尽力。我们会有办法的。我是说，容量，我想它取决于你对智能的野心有多大。如果你做得更多，我认为我们永远摆脱不了算力短缺。我从未见过像这样的商品，但在我看来，对足够高质量、足够低价格的智能的需求实际上是无限的。对电力的需求当然会随着价格下降而上升，但在某个点上，找到用电做的新事情就变得更难了，至少历史上是这样。但用增量智能可以做的事情很多。你可以让东西不断变得更好。这让我想起早期计算机革命的一些名言，当时人们说没人需要超过 640K 的内存。结果发现你需要。我们一直在想更多更好的东西。我认为 AI 也会如此。实际上，有一个我非常喜欢的统计数据：六年前，全球 token 使用量领先者是一名 OpenAI 员工，每月使用约 10 万个 token，当时这看起来荒谬。全球人均使用量几乎是零。

Fair enough. But we'll try our best. We'll figure something out. I mean the capacity, I guess it's a function of how much bigger your ambitions are for intelligence. If you do a lot more, I think we will never be out of the compute shortage. I've never seen any commodity quite like this one, but it seems to me the demand for sufficiently high quality intelligence at a sufficiently low price is effectively uncapped. The demand for electricity certainly goes up as the price goes down, but at some point it gets harder to figure out incremental things to do with electricity, at least historically. But there's a lot of things to do with incremental intelligence. You can just keep having stuff be better for you. It reminds me of some quotes from the early computing revolution when people would say no one needs more than 640K of RAM. Turns out you do. We just keep thinking of more and better stuff. And I think that's going to happen with AI. Actually, a statistic I really like: six and a half years ago, the world token leader was an OpenAI employee using about 100,000 tokens a month, and this seemed ludicrous at the time. The worldwide average per capita was like zero.

现在全球平均每月使用的 token 数，虽然我觉得 token 是最愚蠢的指标，但这是我们现有的数据，大约是 10 万。而 OpenAI 的 token 使用领先者每月使用数千亿。如果这种情况再次发生，我认为很可能会，那么再过六年半，普通人每月使用 5000 亿 token，领先者每月使用千万亿 token。我认为这将成为常态。

Now the worldwide average of tokens per month, which I think tokens are the dumbest metric but it's what we have, is about 100,000. And the token leader at OpenAI uses something in the hundreds of billions. If that happens again, which I think it probably will, then in another six and a half years the average person uses, let's say, 500 billion tokens a month, and the token leader uses a quadrillion or quadrillions of tokens a month. I think that will just become the expectation.

所有这些都会发生，成为现实，那是十年后。最好的版本是什么？十年后，智能完全按需可用，超级智能就在眼前，我们把它搞定。我们做到，社会做到。我有点认为，如果每年人们都拥有更多的自由和自主权，花更多时间做自己想做的事，并且感觉生活质量和时间质量逐年提升，我们大概率就会过得不错。

So all of these things happen, they come to pass, it's 10 years from now. What's the best version of that? 10 years from now with intelligence fully on tap, superintelligence here, and we figure it out. We make it, the society makes it. I kind of think if every year people have more freedom and agency to spend more of their time doing the stuff they want to do, and they feel like the quality of life and the quality of their time is going up year after year, we'll probably be mostly okay.

我们会避免权力的疯狂集中或经济崩溃。我们必然会避免重大的安全事故。我们会避免在任何一个时间单位内发生太多变化。有一个我特别担心的反乌托邦，十年后出现：我们对 AI 安全反应过度。于是我们说，“每个人都能得到癌症疗法。你们将拥有物质丰裕，但没有自由，没有自主权。那将是一个完美的监控国家。完全没有隐私。”你会得到极大的舒适，但世界上再没有真正可做的事，没有真正重要的事。你只是活在为 AI 服务、获得物质财富的状态中。我非常想避免这种情况。人们很容易接受暂时的权衡。所以如果我们说，每年自由和自主权都必须增加，人们必须更能掌控自己的时间，做更多自己想做的事，并且获得更长期的满足感，我认为那会非常好。

We will have avoided a crazy concentration of power or an economic collapse. We'll have necessarily avoided a huge safety incident. We will have avoided too much change in any one time unit. There's one dystopia that I'm particularly nervous about 10 years from now: we overreact to AI safety. So we say, 'Everyone, you're going to get a cure for cancer. You're going to have material abundance, but you will have no freedom. You will have no agency. It will be a perfect surveillance state. There will be no privacy.' You'll get great comfort, but nothing left in the world for you to really do. Nothing that really matters. You'll just kind of live at the service of the AI giving you material wealth. I'd really like to avoid that. It's easy to accept temporary trade-offs. So if we say every year freedom and agency has got to go up, people have got to be more in control of their time and do more of the stuff they want, and more long-term fulfillment, I think that'd be very good.

让我们把你放回斯坦福大二那年。你正要加入 YC。如果你能给那个 Sam Altman 一个瓶中信，你现在会对他说什么？

Let me put you back into sophomore year of Stanford. You're coming to YC. If you could give yourself a message in a bottle to that Sam Altman, what would you say to him right now?

我那时就是，一切都会好起来的。做 startup 感觉非常疯狂又压力巨大。我的第一个 startup 并不成功，那是我人生中一段艰难的时期。但我想很多人回顾职业生涯早期时都会说：你可以犯很多错误，可以失败。科技行业尤其宽容。我希望我能告诉自己，要有所有的动力和抱负，但也要一路更开心一点，相信最终一切都会好起来。因为当下感觉如此艰难、可怕和痛苦。

I was just, it's all going to work out. It feels like such a crazy and very stressful thing to do a startup. My first startup didn't work out great and it was a difficult time in my life. But one thing I think many people say as they look back on the early part of their career: you can make a lot of mistakes, you can fail at stuff. The tech industry in particular is very forgiving of this. I wish I could have told myself to have all the drive and the ambition, but just be a little happier along the way and trust that eventually it was going to be okay. Because it feels so difficult and scary and painful in the moment.

嗯，我想不出更好的方式来结束 Startup School 2026 了。Sam Altman，非常感谢你。

Well, I can't think of a better way to end Startup School 2026. Sam Altman, thank you so much.