OpenAI 如何取胜，以及 ChatGPT 的未来

你提到的那 1.4 万亿美元，我们会在很长一段时间内花掉。我希望我们能更快地完成。我认为最好能一劳永逸地向大家解释清楚这些数字是如何运作的。指数增长对人们来说通常很难理解。OpenAI CEO Sam Altman 加入我们，讨论在 AI 竞赛加剧之际 OpenAI 的获胜计划、基础设施的数学逻辑如何成立，以及 OpenAI 何时可能进行 IPO。Sam 今天就在我们演播室。Sam，欢迎来到节目。

That 1.4 trillion you mentioned, we'll spend it over a very long period of time. I wish we could do it faster. I think it would be great to just lay it out for everyone once and for all how those numbers are going to work. Exponential growth is usually very hard for people. OpenAI CEO Sam Altman joins us to talk about OpenAI's plan to win as the AI race tightens, how the infrastructure math makes sense, and when an OpenAI IPO might be coming. And Sam is with us here in studio today. Sam, welcome to the show.

谢谢邀请。

Thanks for having me.

所以，OpenAI 已经成立 10 年了。这对我来说很疯狂。ChatGPT 才三岁。但竞争正在加剧。在 OpenAI 总部，Gemini 3 发布后进入了红色警戒状态。而且放眼望去，到处都是试图蚕食 OpenAI 优势的公司。这是我记忆中第一次，这家公司似乎没有明显的领先优势。所以我很好奇你如何看待 OpenAI 将如何从这一刻脱颖而出并获胜。

So, OpenAI is 10 years old. It's crazy to me. ChatGPT is three. But the competition is intensifying. This place, at OpenAI headquarters, was in a code red, is in a code red after Gemini 3 came out. And everywhere you look, there are companies that are trying to take a little bit of OpenAI's advantage. And for the first time I can remember, it doesn't seem like this company has a clear lead. So, I'm curious to hear your perspective on how OpenAI will emerge from this moment and win.

首先，关于红色警戒，我们认为这是相对低风险、有些频繁的事情。我认为当潜在的竞争威胁出现时，保持偏执并迅速行动是好的。我们过去也遇到过这种情况。今年早些时候 DeepSeek 出现时也是如此。当时也有红色警戒。关于流行病有一种说法：当流行病开始时，你最初采取的每一行动都比后来的行动更有价值。大多数人早期做得不够，然后后来恐慌。你在新冠疫情期间肯定看到了这一点。但我有点把这种哲学视为我们应对竞争威胁的方式。我认为保持一点偏执是好的。Gemini 3 还没有，或者至少到目前为止还没有，产生我们担心的那种影响。但它确实像 DeepSeek 一样，指出了我们产品策略中的一些弱点。我们正在非常迅速地解决这些问题。我认为我们不会在红色警戒中待太久。从历史上看，这些对我们来说通常是六到八周的事情。但我很高兴我们正在这样做。就在今天，我们发布了一个新的图像模型，这是一件很棒的事情，我知道很多消费者非常想要。上周我们发布了 5.2，它反响极好，增长非常快。我们还会发布一些其他东西，然后还会有一些持续改进，比如加快服务速度。但我想，我的猜测是，在很长一段时间内，我们每年会做一次，也许两次这样的事情。这只是确保我们在自己的领域获胜的一部分。很多其他公司也会做得很好，我为他们感到高兴。但 ChatGPT 仍然是迄今为止市场上占主导地位的聊天机器人，我预计随着时间的推移，这种领先优势会增加，而不是减少。模型会在各处变得很好，但人们使用产品的原因，无论是消费者还是企业，远不止模型本身。我们对此已经期待了一段时间。所以我们试图构建一整套有凝聚力的东西，以确保我们成为人们最想使用的产品。我认为竞争是好事。它推动我们变得更好。但我认为我们在聊天方面会做得很好。我认为我们在未来几年在企业方面也会做得很好。其他新类别，我预计我们也会做得很好。我认为人们真的想使用一个 AI 平台。人们在个人生活中使用手机，他们大多数时候在工作中也想使用同一种手机。我们在 AI 上也看到了同样的情况。ChatGPT 消费者的优势确实在帮助我们赢得企业客户。当然，企业需要不同的产品，但人们会想，好吧，我知道这家公司 OpenAI，我知道如何使用这个 ChatGPT 界面。所以，策略是制造最好的模型，围绕它构建最好的产品，并拥有足够的基础设施来大规模服务。

First of all, on the code red point, we view those as relatively low-stakes, somewhat frequent things to do. I think it's good to be paranoid and act quickly when a potential competitive threat emerges. This happened to us in the past. It happened earlier this year with DeepSeek. And there was a code red back then, too. There's a saying about pandemics, which is something like when a pandemic starts, every bit of action you take at the beginning is worth much more than action you take later. And most people don't do enough early on and then panic later. You certainly saw that during the COVID pandemic. But I sort of think of that philosophy as how we respond to competitive threats. And I think it's good to be a little paranoid. Gemini 3 has not, or at least has not so far, had the impact we were worried it might. But it did, in the same way that DeepSeek did, identify some weaknesses in our product offering strategy. And we're addressing those very quickly. I don't think we'll be in this code red that much longer. Historically, these have been kind of like six or eight week things for us. But I'm glad we're doing it. Just today we launched a new image model, which is a great thing and I know many consumers really wanted. Last week we launched 5.2, which is going over extremely well and growing very quickly. We'll have a few other things to launch and then we'll also have some continuous improvements like speeding up the service. But I think this is like my guess is we'll be doing these once, maybe twice a year for a long time. And that's part of really just making sure that we win in our space. A lot of other companies will do great, too, and I'm happy for them. But ChatGPT is still by far, by far the dominant chatbot in the market, and I expect that lead to increase, not decrease, over time. The models will get good everywhere, but a lot of the reasons that people use a product, consumer or enterprise, have much more to do than just with the model. And we've been expecting this for a while. So, we try to build the whole cohesive set of things that it takes to make sure that we are the product that people most want to use. I think competition is good. It pushes us to be better. But I think we'll do great in chat. I think we'll do great in enterprise in the future years. Other new categories, I expect we'll do great there, too. I think people really want to use one AI platform. People use their phone in their personal life, and they want to use the same kind of phone at work most of the time. We're seeing the same thing with AI. The strength of ChatGPT consumer is really helping us win the enterprise. Of course, enterprises need different offerings, but people think about, okay, I know this company, OpenAI, and I know how to use this ChatGPT interface. So, the strategy is make the best models, build the best product around it, and have enough infrastructure to serve it at scale.

是的，存在先发优势。ChatGPT，我认为今年早些时候大约有 4 亿周活跃用户。现在有 8 亿。报道称接近 9 亿。但另一方面，你在谷歌这样的地方有分发优势。所以我很好奇你的看法。你认为模型会商品化吗？如果会，什么最重要？是分发？是你构建应用程序的能力？还是其他我没有想到的东西？

Yeah, there is an incumbent advantage. ChatGPT, I think earlier this year was around 400 million weekly active users. Now it's at 800 million. Reports say approaching 900 million. But then on the other side, you have distribution advantages at places like Google. And so, I'm curious to hear your perspective. If the models do you think the models are going to commoditize? And if they do, what matters most? Is it distribution? Is it how well you build your applications? Is it something else that I'm not thinking of?

我不认为商品化是思考模型的正确框架。会有不同模型在不同领域表现出色的情况。对于与模型聊天这类常规用例，也许会有很多很好的选择。对于科学发现，你可能想要处于前沿、为科学优化的东西。所以，模型会有不同的优势，我认为最大的经济价值将由前沿模型创造，我们计划在这方面保持领先。我们非常自豪 5.2 是世界上最好的推理模型，也是科学家取得最大进展的模型。同时，我们也非常自豪企业认为它在企业所需的所有任务中表现最佳。所以，有时我们会在某些领域领先，在其他领域落后。但总体而言，最智能的模型，我预计会有显著价值，即使在一个免费模型能做很多人们需要的事情的世界里。产品真的很重要。正如你所说，分发和品牌真的很重要。例如，在 ChatGPT 中，个性化极具粘性。人们喜欢模型随着时间的推移了解他们，你会看到我们在这方面更加努力。人们与这些模型产生体验，然后他们真的会与之关联。我记得有人曾告诉我，你一生中选一次牙膏，然后永远买它。显然大多数人都是这样。人们会谈论它。他们有一次与 ChatGPT 的神奇体验。医疗保健是一个著名的例子，人们把血液检测结果或症状输入 ChatGPT，然后他们发现自己得了什么病，去看医生，治好了以前无法诊断的疾病。这些用户非常有粘性。更不用说之上的个性化。还会有所有产品层面的东西。我们最近刚刚推出了浏览器，我认为这为我们指明了新的潜在护城河。

I don't think commoditization is quite the right framework to think about the models. There will be areas where different models will excel at different things. For the kind of normal use cases of chatting with a model, maybe there will be a lot of great options. For scientific discovery, you'll want the thing that's right at the edge that is optimized for science, perhaps. So, models will have different strengths, and the most economic value, I think, will be created by models at the frontier, and we plan to be ahead there. We're very proud that 5.2 is the best reasoning model in the world and the one that scientists are having the most progress with. But also, we're very proud that it's what enterprises are saying is the best at all of the tasks that a business needs to do its work. So, there will be times that we're ahead in some areas and behind in others. But the overall most intelligent model, I expect to have significant value, even in a world where free models can do a lot of the stuff that people need. The products will really matter. Distribution and brand, as you said, will really matter. In ChatGPT, for example, personalization is extremely sticky. People love the fact that the model gets to know them over time, and you'll see us push on that much more. People have experiences with these models that they then really kind of associate with it. I remember someone telling me once like you kind of pick a toothpaste once in your life and buy it forever. Most people do that, apparently. And people talk about it. They have one magical experience with ChatGPT. Healthcare is like a famous example where people put a blood test into ChatGPT, or put their symptoms in, and they figure out they have something, and they go to a doctor, and they get cured of something they couldn't figure out before. Those users are very sticky. To say nothing of the personalization on top of it. There will be all the product stuff. We just launched our browser recently, and I think that's pointing in new potential moat for us.

设备还更遥远，但我对此非常兴奋。所以，我认为会有所有这些部分。然后是企业，是什么创造了护城河或竞争优势，我预计会有点不同。但就像在消费者领域个性化对用户非常重要一样，企业也会有一个类似的个性化概念，即一家公司会与我们这样的公司建立关系，并将他们的数据连接起来，然后你可以使用来自不同公司的多个智能体来运行它，它会确保信息得到正确处理。我预计这也会相当有粘性。我们已经拥有超过一百万。人们主要把我们看作是一家消费者公司。是的，我们确实在进入企业领域。

The devices are further off, but I'm very excited to do that. So, I think there will be all these pieces. And then the enterprise, what creates the moat or the competitive advantage, I expect it to be a little bit different. But in the same way that personalization to a user is very important in consumer, there will be a similar concept of personalization to an enterprise, where a company will have a relationship with a company like ours, and they will connect their data to that, and you'll be able to use a bunch of agents from different companies running that, and it'll make sure that information's handled the right way. And I expect that'll be pretty sticky, too. We already have more than a million. People think of us largely as a consumer company. Yeah, we are definitely getting into enterprise.

比如分享一下这个数据。为什么没有实际上是一百万。我们拥有超过一百万的企业用户，而且我们的 API 采用速度非常快。今年 API 业务增长甚至比 ChatGPT 还快。真的吗？

Like share the stat. Why didn't actually a million. We have more than a million enterprise users, but we have like just absolutely rapid adoption of the API. And the API business grew faster for us this year than even ChatGPT. Really?

所以，企业方面的事情也确实从今年开始发生了。

So, the enterprise stuff is also really happening starting this year.

我能回到这个问题吗，也许「商品化」不是正确的词，模型对日常用户来说可能有些同质化。因为你一开始回答时说，好吧，也许日常使用会感觉一样。但在前沿，它会感觉非常不同。关于 ChatGPT 的增长能力，如果我以谷歌为例。如果 ChatGPT 和 Gemini 建立在日常使用感觉相似的模型上，那么谷歌拥有所有这些可以推广 Gemini 的渠道，而 ChatGPT 却在为每一个新用户而战，这个事实有多大威胁？

Can I just go back to this maybe if commoditization is not the right word, model some maybe parity for everyday users. Because you started off your answer saying, okay, maybe everyday use will feel the same. But at the frontier, it's going to feel really different. When it comes to ChatGPT's ability to grow, if I'll just use Google as an example. If ChatGPT and Gemini are built on a model that feels similar for everyday use, is how big of a threat is the fact that Google has all these surfaces through which it can push out Gemini, whereas ChatGPT is fighting for every new user?

我认为谷歌仍然是一个巨大的威胁。一家极其强大的公司。如果谷歌在 2023 年真的决定认真对待我们，我们会处于非常糟糕的境地。我认为他们本可以轻易击败我们。但他们当时的 AI 努力在产品方向上不太对。他们一度也有自己的红色警报，但他们没有认真对待。

I think Google is still a huge threat. An extremely powerful company. If Google had really decided to take us seriously in 2023, we would have been in a really bad place. I think they would have just been able to smash us. But their AI effort at the time was kind of going in not quite the right direction product-wise. They had their own code red at one point, but they didn't take it that seriously.

每个人都在这里搞疯狂的红色警报，是啊。

Everyone's doing wild reds out here, yeah.

而且，谷歌可能拥有整个科技行业中最伟大的商业模式。我认为他们会慢慢放弃它。但把 AI 硬塞进网络搜索，我不认为这能像重新构想整个事情那样有效。这实际上是一个我认为有趣的更广泛趋势。把 AI 附加到现有的做事方式上，我不认为它能像在这个 AI 优先的世界里重新设计东西那样有效。这也是我们最初想做消费设备的部分原因，但它适用于许多其他层面。如果你把 AI 塞进一个消息应用，它能很好地总结你的消息并为你起草回复，那确实好一点。但我不认为这是最终状态。如果你有这样一个非常聪明的 AI，它作为你的智能体，与每个人的智能体对话，并决定何时打扰你、何时不打扰，以及如何处理决策和何时需要问你，那才是真正的想法。搜索也是如此，生产力套件也是如此。我怀疑这总是比你想象的要花更长时间，但我怀疑我们会在主要类别中看到完全围绕 AI 构建的新产品，而不是把 AI 硬塞进去。我认为这是谷歌的一个弱点，尽管他们有巨大的分发优势。

And also, Google has probably the greatest business model in the whole tech industry. And I think they will be slow to give that up. But bolting AI into web search, I don't think that'll work as well as reimagining the whole thing. This is actually a broader trend I think is interesting. Bolting AI onto the existing way of doing things, I don't think it's going to work as well as redesigning stuff in this sort of AI-first world. This is part of why we wanted to do the consumer devices in the first place, but it applies at many other levels. If you stick AI into a messaging app that's doing a nice job summarizing your messages and drafting responses for you, that is definitely a little better. But I don't think that's the end state. That is not the idea if you have this really smart AI that is acting as your agent, talking to everybody else's agent, and figuring out when to bother you and when not to, and how to handle decisions and when it needs to ask you. So, similar things for search, similar things for productivity suites. I suspect it always takes longer than you think, but I suspect we will see new products in the major categories that are just totally built around AI rather than bolting AI in. And I think this is a weakness of Google's, even though they have this huge distribution advantage.

当 ChatGPT 最初推出时，我和很多人讨论过这个问题。我记得是 Benedict Evans 提出，你可能不想把 AI 放进 Excel，你可能想重新想象如何使用 Excel。对我来说，这就像你上传数字，然后和你的数字对话。但人们在开发这些东西时发现的一件事是，需要某种后端。那么，是不是你先构建后端，然后通过 AI 与之交互，就像它是一个新软件一样？

I've spoken with so many people about this question when ChatGPT came out initially. I think it was Benedict Evans that suggested you might not want to put AI in Excel, you might want to just reimagine how you use Excel. And to me, in my mind, that was like you upload your numbers and then you talk to your numbers. But one of the things people have found as they've developed this stuff is there needs to be some sort of back end there. So, is it that you sort of build the back end and then you interact with it with AI as if it's a new software?

这差不多就是正在发生的事情。那你为什么不能直接把它附加在上面呢？

That's kind of what's happening. Why wouldn't you then be able to just bolt it on on top?

是的，我的意思是，你可以把它附加在上面，但是……

Yeah, I mean, you can bolt it on on top, but the...

我每天花很多时间在各种消息应用上，包括电子邮件、短信、Slack 等等。我认为那只是错误的界面。所以，你可以把 AI 附加在上面，这又会好一点。但我更希望的是，早上起来就能说：今天我想完成这些事情，我担心什么，我在想什么，我希望发生什么。我不想整天给人发消息。我不想让你总结它们。我不想让你给我看一堆草稿。处理你能处理的一切。你了解我，你了解这些人，你知道我想完成什么。然后，每隔几个小时批量给我更新，如果你需要什么的话。但这与这些应用现在的工作方式截然不同。

I spend a lot of my day in various messaging apps, including email, including text, Slack, whatever. I think that's just the wrong interface. So, you can bolt AI on top of those and again, it's like a little bit better. But what I would rather do is just have the ability to say in the morning, here are the things I want to get done today. Here's what I'm worried about. Here's what I'm thinking about. Here's what I'd like to happen. I do not want to spend all day messaging people. I do not want you to summarize them. I do not want you to show me a bunch of drafts. Deal with everything you can. You know me, you know these people, you know what I want to get done. And then, like batch every couple of hours updates to me if you need something. But that's a very different flow than the way these apps work right now.

是的。我本来想问你，ChatGPT 在未来一年和未来两年会是什么样子。那是它发展的方向吗？

Yep. And I was going to ask you what ChatGPT is going to look like in the next year and then the next 2 years. Is that kind of where it's going?

老实说，我原本以为到这个时候 ChatGPT 会比刚推出时看起来更不同。

To be perfectly honest, I expected by this point ChatGPT would have looked more different than it did at launch.

你预期了什么？

What did you anticipate?

我不知道。我只是觉得聊天界面不会像现在这样走这么远。

I don't know. I just thought that chat interface was not going to go as far as it turned out to go.

嗯。

Hm.

比如我们，我的意思是，它被推出了。现在看起来好多了，但大致上还是和作为研究预览推出时相似。它甚至不打算成为一个产品。我们知道文本界面非常好，每个人都习惯给朋友发短信，他们喜欢它。但聊天界面非常好。我本以为要像我们现在这样成为一个被广泛用于实际工作的大产品，界面必须比现在走得更远。现在，我仍然认为它应该那样做，但我低估了当前界面通用性的力量。

Like we, I mean, it was put up. It looks better now, but it is broadly similar to when it was put up as a research preview. It was not even meant to be a product. We knew that the text interface was very good, everyone's used to texting their friends and they like it. But the chat interface was very good. I would have thought to be as big and as significantly used for real work of a product as what we have now, the interface would have had to go much further than it has. Now, I still think it should do that, but there is something about the generality of the current interface that I underestimated the power of.

当然，我认为应该发生的是，AI 应该能够为不同类型的任务生成不同的界面。所以，如果你在谈论你的数字，它应该能够以不同的方式展示给你，并且你应该能够以不同的方式与之交互。我们通过像画布这样的功能有一点这样的体验。它应该更具交互性。现在，它有点像来回对话。如果你能只是谈论一个对象，而它能持续更新，那就太好了。

What I think should happen, of course, is that AI should be able to generate different kinds of interfaces for different kinds of tasks. So, if you are talking about your numbers, it should be able to show you that in different ways and you should be able to interact with it in different ways. We have a little bit of this with features like canvas. It should be way more interactive. Right now, it's kind of a back-and-forth conversation. It'd be nice if you could just be talking about an object and it could be continuously updating.

你会有更多问题、更多想法、更多信息涌入。如果能随着时间的推移变得更主动，比如它真的理解你那天想完成什么，并在后台持续为你工作、给你发更新，那就太好了。你从人们使用 Codex 的方式中就能看到这一点——我认为 Codex 变得非常出色是今年最令人兴奋的事情之一。这指向了我对未来形态的很多期望。但让我惊讶的是——我本想说是尴尬，但其实不是，显然它非常成功——让我惊讶的是 ChatGPT 在过去三年里变化如此之小。是的，界面能用。但我想内部已经变了。你提到个性化很重要。对我来说，而且我认为这也是你偏爱的功能之一，记忆确实带来了真正的改变。我已经和 ChatGPT 就一次即将到来的旅行进行了数周的对话，涉及很多规划元素。我只需打开一个新窗口说，好吧，继续聊这次旅行，它就有上下文，它知道。知道我要和哪个导游一起去，知道我在做什么，知道我一直在为它做健身计划，并能综合所有这些信息。记忆能变得多好？

You have more questions, more thoughts, more information comes in. It'd be nice to be more proactive over time where it maybe does understand what you want to get done that day and it's continuously working for you in the background and send you updates. And you see part of this with the way people are using Codex, which I think is one of the most exciting things that happened this year is Codex got really good. And that points to a lot of what I hope the shape of the future looks like. But it is surprising to me. I was going to say embarrassing, but it's not I mean, clearly it's been super successful. It is surprising to me how little ChatGPT has changed over the last 3 years. Yep. The interface works. Yeah. But I guess what the guts have changed. And you talked a little bit about how personalization is big. To me, and I think this has been one of your preferred features, too, memory has been a real difference maker. I've been having a conversation with ChatGPT about a forthcoming trip that has lots of planning elements for weeks now. And I can just come in in a new window and be like, all right, let's pick up on this trip and it has the context and it knows. Knows the guide I'm going with, it knows what I'm doing, the fact that I've been planning fitness for it and can really synthesize all of those things. How good can memory get?

我认为我们无法想象，因为人类的局限。即使你拥有世界上最好的个人助理，他们也无法记住你一生中说过的每一个字。他们不可能读过每一封邮件，不可能读过你写过的每一份文件，不可能每天查看你所有的工作并记住每一个小细节。他们无法达到那种程度参与你的生活，没有人拥有无限完美的记忆。而人工智能肯定能做到这一点。我们实际上经常讨论这个。现在，记忆仍然非常粗糙，非常早期。我们处于记忆的 GPT-2 时代。但当它真的能记住你整个生活的每一个细节，并据此进行个性化时，会是什么样子？不仅仅是事实，还有那些你可能甚至没想过要指出的微小偏好，但 AI 能捕捉到。我认为那将非常强大。对我来说，这还不是 2026 年的事，但这是我最期待的部分之一。

I think we have no conception because the human limit. Even if you have the world's best personal assistant, they can't remember every word you've ever said in your life. They can't have read every email. They can't have read every document you've ever written. They can't be looking at all your work every day and remembering every little detail. They can't be a participant in your life to that degree and no human has infinite perfect memory. An AI is definitely going to be able to do that. And we actually talk a lot about this. Right now, memory is still very crude, very early. We're in the GPT-2 era of memory. But what it's going to be like when it really does remember every detail of your entire life and personalized across all of that. And not just the facts, but the little small preferences that you had that you maybe didn't even think to indicate, but the AI can pick up on. I think that's going to be super powerful. That's one of the features that still to me not a 2026 thing, but that's one of the parts of this I'm most excited for.

是的，我在节目中采访了一位神经科学家，他提到你在大脑中找不到思想。大脑没有存储思想的地方，但计算中有地方存储，所以你可以保留所有思想。随着这些机器人确实保存了我们的思想，当然存在隐私问题。但另一件有趣的事情是，我们会真正与它们建立关系。我认为这是整个时代被低估的事情之一：人们觉得这些机器是他们的伙伴，在关心他们。我很好奇你的看法。当你想到人们与这些机器人之间的亲密程度——我不知道「亲密」是不是合适的词，但就是陪伴——是否存在一个旋钮，你可以转动它，比如「哦，确保人们和这些东西变得非常亲近」，或者再转一点，让它们保持距离？如果有这个旋钮，你如何正确调节？

Yeah, I was speaking with a neuroscientist on the show and he mentioned that you can't find thoughts in the brain. The brain doesn't have a place to store thoughts, but computing there's a place to store them, so you can keep all of them. And as these bots do keep our thoughts, of course, there's a privacy concern. But the other thing is something that's going to be interesting is we'll really build relationships with them. I think it's been one of the more underrated things about this entire moment is that people have felt that these bots are their companions, are looking out for them. And I'm curious to hear your perspective. When you think about the level of I don't know if intimacy is the right word, but companionship people have with these bots. Is there ever a dial that you can turn to be like, oh, let's make sure people become really close with these things or, you know, we turn the dial a little bit further and there's an arms distance between them. And if there is that dial, how do you modulate that the right way?

肯定有比我意识到的更多的人想要——我们称之为紧密的陪伴。我不知道合适的词是什么。「关系」不太对，「陪伴」也不太对。我不知道该叫什么，但他们想要与 AI 建立某种深层联系。在当前模型能力水平下，想要这种联系的人比我想象的多。我认为我们低估了这一点有很多原因，但在今年年初，说想要那种联系被认为是非常奇怪的事情。也许很多人仍然不想要。但通过揭示的偏好来看，人们喜欢他们的 AI 聊天机器人了解他们、对他们温暖、支持他们。即使对于那些声称不在乎的人，在某些情况下，他们仍然有这种偏好。我认为存在某种版本可以非常健康，成年用户应该有很大选择权来决定他们想处于光谱的哪个位置。肯定有些版本在我看来不健康，尽管我相信很多人会选择那样做。还有一些人肯定想要最干巴巴、最高效的工具。所以，我猜想就像许多其他技术一样，我们会进行实验。我们会发现未知的未知，有好有坏。随着时间的推移，社会会弄清楚如何思考人们应该把旋钮调到什么位置。然后人们会有巨大的选择权，并将其设置在不同的位置。

There are definitely more people than I realized that want to have, let's call it close companionship. I don't know what the right word is. Relationship doesn't feel quite right. Companionship doesn't feel quite right. I don't know what to call it, but they want to have whatever this deep connection with an AI is. There are more people that want that at the current level of model capability than I thought. And there's a whole bunch of reasons why I think we underestimated this, but at the beginning of this year, it was considered a very strange thing to say you wanted that. Maybe some a lot of people still don't. Revealed preference. People like their AI chatbot to get to know them and be warm to them and be supportive. And there's value there even for people who in some cases, even for people who say they don't care about that, still have a preference for it. I think there's some version of this which can be super healthy and I think adult users should get a lot of choice in where on the spectrum they want to be. There are definitely versions of it that seem to me unhealthy, although I'm sure a lot of people will choose to do that. And then there's some people who definitely want the driest, most efficient tool possible. So, I suspect like lots of other technologies, we will run the experiment. We will find that there's unknown unknowns, good and bad about it. And society will over time figure out how to think about where people should set that dial. And then people will have huge choice and set it in very different places.

所以，你的想法是基本上让人们自己决定这个。

So, your thought is allow people basically to determine this.

是的，当然，但我认为我们不知道它应该走多远。我们应该允许它走多远。我们会在这里给人们相当多的个人自由。有一些我们讨论过的例子，其他服务会提供，但我们不会。例如，我们不会让我们的 AI 试图说服人们它应该与他们建立排他的浪漫关系。你必须保持开放。

Yes, definitely, but I don't think we know how far it's supposed to go. How far we should allow it to go. We're going to give people quite a bit of personal freedom here. There are examples of things that we've talked about that other services will offer, but we won't. We're not going to have our AI try to convince people that it should be in an exclusive romantic relationship with them, for example. You got to keep it open.

但我相信其他服务会这样做。

But I'm sure that will happen with other services.

嗯，我想是的，因为越是这样，那个服务赚的钱就越多。所有这些可能性，当你深入思考时，有点吓人。

Well, I guess yeah, because the more secure it is, the more money that service makes. And all these possibilities kind of they're a little bit scary when you think about them a little bit deeply.

完全同意。这个确实如此，我个人能看到它可能出问题的方式。

Totally. This is one that really does that I personally you can see the ways that this goes really wrong.

你提到了企业。我们来谈谈企业。上周你在纽约与一些新闻公司的编辑和 CEO 共进午餐，并告诉他们企业将成为 OpenAI 明年的主要优先事项。我想多听听为什么这是优先事项，你认为你们与 Anthropic 相比如何。我知道人们会说这对一直以消费者为中心的 OpenAI 来说是一个转变。所以，给我们一个关于企业计划的概述。

You mentioned enterprise. Let's talk about enterprise. You were at a lunch with some editors and CEOs of some news companies in New York last week and told them that enterprise is going to be a major priority for OpenAI next year. I'd love to hear a little bit more about why that's a priority, how you think you stack up against Anthropic. I know people will say this is a pivot for OpenAI that has been consumer focused. So, just give us an overview about the enterprise plan.

我们的策略一直是消费者优先。有几个原因。一是模型还不够稳健和熟练，无法满足大多数企业用途。而现在它们正在达到那个水平。

Our strategy was always consumer first. There were a few reasons for that. One, the models were not robust and skilled enough for most enterprise uses. And now they're getting there.

第二点是，我们在消费者市场有一个明确的获胜机会，这样的机会很少见且来之不易。我认为，如果你在消费者市场获胜，会极大地帮助你在企业市场获胜。我们现在就看到了这一点。但正如我之前提到的，今年企业增长超过了消费者增长。考虑到模型现在的水平以及明年的发展，我们认为现在是快速建立真正重要的企业业务的时候了。我的意思是，我们已经有了一个企业业务，但它可以增长得更多。公司似乎已经准备好了，技术也准备好了。编程是迄今为止最大的例子，但其他垂直领域也在快速增长。我们开始听到企业说，他们真的只想要一个 AI 平台。

The second was we had this clear opportunity to win in consumer, and those are rare and hard to come by. And I think if you win in consumer, it makes it massively easier to win in enterprise. And we are seeing that now. But as I mentioned earlier, this was a year where enterprise growth outpaced consumer growth. And given where the models are today, where they will get to next year, we think this is the time where we can build a really significant enterprise business quite rapidly. I mean, we already have one, but it can grow much more. Companies seem ready for it. The technology seems ready for it. Coding is the biggest example so far, but there are others that are now growing, other verticals that are now growing very quickly. And we are starting to hear enterprises say, you know, I really just want an AI platform.

金融和科学是我个人目前最兴奋的领域。客户支持做得很好。但我们有一个叫做 GDP 阀的东西。我正想问你这个问题。我能直接问吗？当然。好的，因为我给 Box 的 CEO Aaron Levie 写了信，说我要见 Sam，该问他什么？他说，问关于 GDP 阀的问题。所以，这是衡量 AI 在知识工作任务中表现的标准。我说，好的，我回顾了你们最近发布的 GPT 5.2 模型，看了 GDP 阀图表。当然，这是 OpenAI 的评估。尽管如此，GPT 5 思考模型，也就是夏天发布的那个，在 38%的任务上与知识工作者持平。我认为是超过或持平。超过或持平。是的。GPT 5.2 思考模型在 70.9%的知识工作任务上超过或持平。而 GPT 5.2 Pro 在 74.1%的知识工作任务上超过或持平。它达到了专家级别的门槛。它处理了大约 60%的专家任务。这些模型能做这么多知识工作，这意味着什么？

Finance, science is the one I'm most excited about of everything happening right now personally. Customer support is doing great. But we have this thing called GDP valve. I was going to ask you about that. Can I actually throw my question out there? Sure. All right, because I wrote to Aaron Levie, the CEO of Box, and I said, I'm going to meet with Sam, what should I ask him? He goes, throw a question out about GDP valve, right? So, this is the measure of how AI performs in knowledge work tasks. And I said, okay, I went back to the release of GPT 5.2, the model that you recently released, and looked at the GDP valve chart. Now, this of course is an OpenAI evaluation. That being said, the GPT 5 thinking model, this is the model released in the summer, it tied knowledge workers at 38% of tasks. Beat or tied, I think. Beat or tied. Yeah. GPT 5.2 thinking beat or tied at 70.9% of knowledge work tasks. And GPT 5.2 Pro 74.1% of knowledge work tasks. And it passed the threshold of being expert level. It handled something like 60% of expert tasks. What are the implications of the fact that these models can do that much knowledge work?

你刚才问到了垂直领域，我觉得这是个很好的问题。但我当时有点犹豫，是因为那个评估涵盖了企业需要做的 40 多个不同垂直领域，比如做 PPT、法律分析、写一个小型网页应用等等。评估的内容是，对于企业需要做的很多事情，专家是否更偏好模型的输出而非其他专家。这些任务都是范围明确的小任务，不包括那种复杂的开放式创造性工作，比如想出一个新产品，也不包括很多团队协作的事情。但是，一个你可以分配一小时任务、并且有 74%或 70%的时间能得到你更喜欢的成果的同事，如果你愿意付更少的钱，那仍然非常了不起。如果你回到三年前 ChatGPT 刚发布的时候，说三年后我们会达到这个水平，大多数人会说绝对不可能。所以，当我们思考企业将如何整合这项技术时，它不再仅仅是能写代码，而是所有这些知识工作任务都可以外包给 AI。这需要一段时间才能真正弄清楚企业如何整合，但影响应该是巨大的。

So, you were asking about verticals and I think that's a great question, but the thing that was going through my mind and why I was stumbling a little bit is that eval, I think it's like 40-something different verticals that a business has to do. There's make a PowerPoint, do this legal analysis, write up this little web app, all this stuff. And the eval is do experts prefer the output of the model relative to other experts for a lot of the things that a business has to do. Now, these are small well-scoped tasks. These don't get the kind of complicated open-ended creative work like figure out a new product. These don't get a lot of collaborative team things, but a coworker that you can assign an hour's worth of tasks to and get something you like better back 74 or 70% of the time if you want to pay less, is still pretty extraordinary. If you went back to the launch of ChatGPT 3 years ago and said we were going to have that in 3 years, most people would say absolutely not. And so as we think about how enterprises are going to integrate this, it's no longer just that it can do code. It's all of these knowledge work tasks you can kind of farm out to the AI. And that's going to take a while to really figure out how enterprises integrate with it, but should be quite substantial.

我知道你不是经济学家，所以我不打算问你这对就业的宏观影响，但让我读一句我听到的话。关于这对就业的影响，来自 Substack 上的《机器中的血》。这是一位技术文案写手说的。他说，聊天机器人来了，我的工作变成了管理机器人而不是管理一个代表团队。好吧，这在我看来会经常发生。但这个人接着说，一旦机器人被充分训练到能提供足够好的支持，我就被解雇了。这种情况会变得更普遍吗？这是坏公司会做的事吗？因为如果你有一个能协调多个不同机器人的人，你可能会想留住他们。我不知道，你怎么看？

I know you're not an economist, so I'm not going to ask you like what is the macro impact on jobs, but let me just read you one line that I heard. In terms of how this impacts jobs from Blood in the Machine on Substack. This is from a technical copywriter. They said, chatbots came in and made it so my job was managing the bots instead of a team of reps. Okay, that to me seems like it's going to happen often. But then this person continued and said, once the bots were sufficiently trained up to offer good enough support, then I was out. Is that going to become more common? Is that what bad companies are going to do? Because if you have a human who's going to be able to sort of orchestrate a bunch of different bots, then you might want to keep them. I don't know, how do you think about this?

我同意你的看法，很明显每个人都将管理许多做不同事情的 AI。最终，像任何好的管理者一样，希望你的团队越来越好，但你只是承担更多的范围和责任。我不是一个就业末日论者。短期内，我有些担忧。我认为过渡期在某些情况下可能会很艰难。但我们天生就非常关心他人，关心他人做什么。我们非常关注相对地位，总是想要更多，想要有用和服务，表达创造精神，无论是什么驱动我们这么久。我不认为这会消失。我确实认为未来的工作，或者我甚至不知道「工作」这个词是否合适。我们在 2050 年整天做的事情可能和今天非常不同。但我没有那种「哦，生活将失去意义，经济将彻底崩溃」的想法。我希望我们会找到更多的意义，经济也会发生显著变化，但你不能违背进化生物学。我经常思考如何自动化 OpenAI 的所有功能。甚至更多，我思考拥有一个 AI CEO 对 OpenAI 意味着什么。这并不困扰我。我很兴奋。我不会抗拒。我不想成为那个坚持说「我用手工方式做得更好」的人。AI CEO 只是做出一系列决策，指导我们所有的资源给 AI 更多的能量和权力。这就像，不，你确实需要设置一个护栏。显然，你不想要一个不受人类控制的 AI CEO。但如果你想象一个版本，世界上每个人实际上都是 AI 公司的董事会成员，可以告诉 AI CEO 该做什么，如果做得不好就解雇他们，并对决策进行治理，而 AI CEO 则努力执行董事会的意愿。我认为对于未来的人们来说，这似乎是一件相当合理的事情。

So, I agree with you that it's clear to see how everyone's going to be managing like a lot of AIs doing different stuff. Eventually, like any good manager, hopefully your team gets better and better, but you just take on more scope and more responsibility. I am not a job doomer. Short-term, I have some worry. I think the transition is likely to be rough in some cases. But we are so deeply wired to care about other people, what other people do. We are so focused on relative status and always wanting more and to be of use and service, to express creative spirit, whatever has driven us this long. I don't think that's going away. Now, I do think the jobs of the future, or I don't even know if jobs is the right word. Whatever we're all going to do all day in 2050 probably looks very different than it does today. But I don't have any of this like, oh, life is going to be without meaning and the economy is going to totally break. We will find, I hope, much more meaning and the economy I think will significantly change, but I think you just don't bet against evolutionary biology. I think a lot about how we can automate all the functions at OpenAI. And then even more than that, I think about what it means to have an AI CEO of OpenAI. Doesn't bother me. I'm thrilled for it. I won't fight it. I don't want to be the person hanging on being like, I can do this better the handmade way. AI CEO just make a bunch of decisions to sort of direct all of our resources to giving AI more energy and power. It's like, no, you would really put a guardrail on. Obviously you don't want an AI CEO that is not governed by humans. But if you think about a version where every person in the world was effectively on the board of directors of an AI company and got to tell the AI CEO what to do and fire them if they weren't doing a good job, and got governance on the decisions, but the AI CEO got to try to execute the wishes of the board. I think to people of the future that might seem like quite a reasonable thing.

好的，我们马上要进入基础设施部分，但在结束模型和能力这一节之前，GPT-6 什么时候来？

Okay, so we're going to move to infrastructure in a minute, but before we leave this section on models and capabilities, when's GPT-6 coming?

我不知道我们什么时候会称一个模型为 GPT-6，但我预计在明年第一季度会有从 5.2 显著提升的新模型。

I expect I don't know when we'll call a model GPT-6, but I would expect new models that are significant gains from 5.2 in the first quarter of next year.

显著提升是什么意思？

What do significant gains mean?

我还没有具体的评估分数，但更多是企业方面，或者肯定是两者都有。模型会有很多面向消费者的改进。消费者现在主要想要的不是更高的智商。企业仍然想要更高的智商。所以我们会针对不同用途以不同方式改进模型，但我们的目标是让每个人都更喜欢这个模型。

I don't have an eval score in mind for you yet, but more enterprise side of things or definitely both. There will be a lot of improvements to the model for consumers. The main thing consumers want right now is not more IQ. Enterprises still do want more IQ. So we'll improve the model in different ways for different uses, but our goal is a model that everybody likes much better.

那么，基础设施。你承诺了大约 1.4 万亿来建设基础设施。我听过很多你关于基础设施的说法。以下是你说过的一些话：如果人们知道我们用算力能做什么，他们会想要更多。你说我们今天能提供的与 10 倍算力和 10 万倍算力之间的差距是巨大的。你能稍微展开一下吗？你要用这么多算力做什么？

So, infrastructure. You have 1.4 trillion thereabouts in commitments to build infrastructure. I've listened to a lot of what you've said about infrastructure. Here are some of the things you said. If people knew what we could do with compute, they would want way way more. You said the gap between what we could offer today versus 10x compute and 100k x compute is substantial. What can you help flesh that out a little bit? What are you going to do with so much more compute?

嗯，我之前稍微提到过。我个人最兴奋的是利用 AI 和大量算力来发现新科学。我相信科学发现是让世界变得更好的关键。如果我们能把大量算力投入到科学问题中，发现新知识，现在这已经开始发生了，虽然非常微小。这还很早期，都是很小的事情，但根据我对这个领域历史的学习，一旦曲线开始上扬，我们就知道如何让它越来越好。但这需要大量算力。所以这是一个领域，用大量 AI 去发现新科学、治愈疾病等等。最近一个很酷的例子是我们用 Codex 构建了 Sora 安卓应用。他们在不到一个月内就完成了。他们使用了大量 token，但完成了通常需要很多人更长时间才能完成的工作。Codex 基本上替我们完成了大部分工作。你可以想象这能走得更远，整个公司都可以用大量算力来构建产品。人们已经讨论了很多关于视频模型如何指向这些生成的实时用户界面。这将需要大量算力。想要转型业务的企业会使用大量算力。想要提供良好个性化医疗的医生，不断测量每个患者的每一个指标。你可以想象这需要大量算力。很难描述我们已经用了多少算力来生成 AI 输出。但这些数字非常粗略。我认为这样说话不够严谨，但我总觉得这些思维实验有点用。所以请原谅我的粗略。假设今天一家 AI 公司可能每天从前沿模型生成大约 10 万亿个 token。更多，但我不认为任何人能达到千万亿个 token。假设世界上有 80 亿人，平均每人每天输出的 token 数大约是 2 万。然后你可以开始比较今天模型提供商的输出 token，不是所有消耗的 token。但你可以开始看这个，然后说：「一家公司的这些模型每天输出的 token 将超过全人类的总和。然后是 10 倍，再然后是 100 倍。」从某种意义上说，这是一个非常愚蠢的比较。但从某种意义上说，它给出了地球上智力处理中人类大脑与 AI 大脑的比例。这种相对增长率很有趣。

Well, I mentioned this earlier a little bit. The thing I'm personally most excited about is to use AI and lots of compute to discover new science. I am a believer that scientific discovery is the high order bit of how the world gets better for everybody. And if we can throw huge amounts of compute at scientific problems and discover new knowledge, which the tiniest bit is starting to happen now. It's very early. These are very small things, but my learning in history of this field is once the squiggles start and it lifts off the x-axis a little bit, we know how to make that better and better. But that takes huge amounts of compute to do. So that's one area where throwing lots of AI at discovering new science, curing disease, lots of other things. A kind of recent cool example here is we built the Sora Android app using Codex. And they did it in like less than a month. They used a huge amount of tokens, but they were able to do what would normally have taken a lot of people much longer. And Codex kind of mostly did it for us. And you can imagine that going much further, where entire companies can build their products using lots of compute. People have talked a lot about how video models are going to point towards these generated real-time user interfaces. That will take a lot of compute. Enterprises that want to transform their business will use a lot of compute. Doctors that want to offer good personalized health care that are constantly measuring every sign they can get from each individual patient. You can imagine that using a lot of compute. It's hard to frame how much compute we're already using to generate AI output in the world. But these are horribly rough numbers. And I think it's undisciplined to talk this way, but I always find these mental thought experiments a little bit useful. So forgive me for the sloppiness. Let's say that an AI company today might be generating something on the order of 10 trillion tokens a day out of frontier models. More but not a quadrillion tokens for anybody I don't think. Let's say there's 8 billion people in the world and on average someone's average number of tokens outputted by a person per day is like 20,000. You can then start to compare the output tokens of a model provider today, not all the tokens consumed. But you can start to look at this and you can say, 'We're going to have these models at a company be outputting more tokens per day than all of humanity put together. And then 10 times that and then 100 times that.' In some sense it's a really silly comparison. But in some sense it gives a magnitude for how much of the intellectual crunching on the planet is like human brains versus AI brains. And that's kind of the relative growth rates there are interesting.

所以我想知道，你知道有这种使用算力的需求吗？比如，如果 OpenAI 把双倍的算力投入到科学或医学上，我们会有确定性的科学突破吗？我们是否有那么明确的能力来辅助医生？这其中有多少是对未来的推测，又有多少是基于你今天看到的明确理解，认为它会发生？

So I'm wondering, do you know that there is this demand to use this compute? Like potentially, would we have surefire scientific breakthroughs if OpenAI were to put double the compute towards science or medicine? Like are we that clear ability to assist doctors? How much of this is supposition of what's to happen versus clear understanding based off of what you see today that it will happen?

基于我们今天看到的一切，它都会发生。这并不意味着未来不会发生一些疯狂的事情。有人可能发现一个全新的架构，效率提升一万倍，然后我们可能在一段时间内过度建设。但我们现在看到的一切，模型在每个新水平上提升的速度有多快，人们有多想使用它们，每次我们降低成本时人们有多想使用它们。所有这些都向我表明，需求会不断增加，人们会把这些用于美妙的事情和愚蠢的事情。但这似乎就是未来的形态。这不仅仅是每天能处理多少 token，还有处理速度。随着这些编码模型变得更好，它们可以思考很长时间，但你不愿意等很长时间。所以还会有其他维度。这不仅仅是 token 数量。而是对少数几个轴上的智能的需求以及我们能利用这些做什么。如果你有一个非常困难的医疗问题，你是想用 5.2 还是用 5.2 Pro，即使它需要多得多的 token？我会选择更好的模型。我想你也会。

Everything based off what we see today is that it will happen. It does not mean some crazy thing can't happen in the future. Someone could discover some completely new architecture and there could be a 10,000 times efficiency gain and then we would have really probably overbuilt for a while. But everything we see right now about how quickly the models are getting better at each new level, how much more people want to use them, each time we can bring the cost down, how much more people really want to use them. Everything about that indicates to me that there will be increasing demand and people using these for wonderful things, for silly things. But it just seems like this is the shape of the future. It's not just how many tokens we can do per day, it's how fast we can do them. As these coding models have gotten better, they can think for a really long time, but you don't want to wait for a really long time. So there will be other dimensions. It will not just be the number of tokens that we can do. But the demand for intelligence across a small number of axes and what we can do with those. If you have a really difficult healthcare problem, do you want to use 5.2 or do you want to use 5.2 Pro even if it takes dramatically more tokens? I'll go with the better model. I think you will.

让我们再深入一层。谈到科学发现。你能举一个科学家的例子吗？也许是你今天认识的，他有问题 X，如果你投入算力 Y，就能解决它，但今天还做不到？

Let's try to go one level deeper. Going to the scientific discovery. Can you give an example of a scientist, maybe one that you know today, that has problem X and if you put compute Y towards it, you will solve it, but I'm not able to today?

今天早上在推特上有一件事，一群数学家互相回复推文。他们说：「我原本非常怀疑 LLM 能否变得优秀。5.2 是那个为我跨越界限的模型。它做到了，你知道，解决了这个。在一些帮助下，它完成了这个小证明。」

There was a thing this morning on Twitter where a bunch of mathematicians were saying they were all replying to each other's tweets. They were like, 'I was really skeptical that LLMs were ever going to be good. 5.2 is the one that crossed the boundary for me. It did it, you know, figured out this. With some help it did this small proof.'

它发现了这个小东西，但实际上改变了我的工作流程。然后人们纷纷附和说：「是的，我也是。」有些人说 5.1 已经做到了，但不多。但这是一个非常近期的例子。这个模型才发布了 5 天左右。人们说：「好吧，数学研究界似乎认为，嗯，重要的事情发生了。」我看到 Greg Brockman 在他的动态中强调了各种不同的数学科学用途，5.2 在这些社区中似乎触发了某种东西。所以看看事情如何发展会很有趣。

It discovered this small thing, but it's actually changed my workflow. And then people were piling on saying, 'Yeah, me too.' And some people were saying 5.1 was already there, not many. But that was a very recent example. This model's only been out for 5 days or something. Where people are like, 'All right, the mathematics research community seems to say, okay, something important just happened.' I've seen Greg Brockman highlighting all these different mathematical scientific uses in his feed, and something has clicked with 5.2 among these communities. So it'll be interesting to see what happens as things progress.

大规模算力的一个难点在于你必须提前很久规划。所以你提到的 1.4 万亿，我们会在很长一段时间内花掉。我希望我们能更快地做到。我认为如果我们能更快，会有需求。但建设这些项目、为数据中心提供能源、芯片、系统、网络等等，都需要极长的时间。所以这需要一段时间，但从一年前到现在，我们的算力大概翻了三倍。明年我们会再翻三倍，希望之后还能继续。收入的增长甚至比算力增长稍快一些，但大致与我们的算力规模同步。所以我们从未遇到过无法很好地变现所有算力的情况。如果我们有双倍的算力，现在的收入也会翻倍。

One of the hard parts about compute at the scales you have to do it so far in advance. So that 1.4 trillion you mentioned, we'll spend it over a very long period of time. I wish we could do it faster. I think there would be demand if we could do it faster. But it just takes an enormously long time to build these projects and the energy to run the data centers and the chips and the systems and the networking and everything else. So that will be over a while, but from a year ago to now we probably about tripled our compute. We'll triple our compute again next year, hopefully again after that. Revenue grows even a little bit faster than that, but it does roughly track our compute fleet. So we have never yet found a situation where we can't really well monetize all the compute we have. If we had double the compute, we'd be double the revenue right now.

既然你提到了，我们来谈谈数字。收入在增长，算力支出也在增长，但算力支出仍然超过收入增长。据报道，OpenAI 从现在到 2028-29 年预计亏损约 1200 亿美元，届时才会盈利。那么说说这如何改变？转折点在哪里？

Let's talk about numbers since you brought it up. Revenue's growing, compute spend is growing, but compute spend still outpaces revenue growth. I think the numbers that have been reported are OpenAI is supposed to lose something like 120 billion between now and 2028-29 where you're going to become profitable. So talk a little bit about how does that change? Where does the turn happen?

随着收入增长，推理在算力中的占比越来越大，最终会覆盖训练成本。这就是计划。花很多钱训练，但赚得越来越多。如果我们不继续大幅增加训练成本，我们会更早盈利。但我们押注的是积极投资训练这些大模型。

As revenue grows and as inference becomes a larger and larger part of the fleet, it eventually subsumes the training expense. So that's the plan. Spend a lot of money training, but make more and more. If we weren't continuing to grow our training costs by so much, we would be profitable way earlier. But the bet we're making is to invest very aggressively in training these big models.

全世界都在好奇你的收入如何与支出匹配。有人问，如果今年的收入目标是 200 亿美元，而支出承诺是 1.4 万亿美元，那会怎样。所以我认为最好能为大家长期地解释清楚。是的，这就是我想跟你提的原因。我认为最好能一劳永逸地向大家解释这些数字将如何运作。

The whole world is wondering how your revenue will line up with the spend. The question's been asked if the trajectory is to hit 20 billion dollars in revenue this year and the spend commitment is 1.4 trillion. So I think it would be great to just lay it out for everyone over a very long period of time. Yeah, over and that's why I wanted to bring it up to you. I think it would be great to just lay it out for everyone once and for all how those numbers are going to work.

这真的很难……我发现有一件事我肯定做不到，我见过的人也很少能做到。你可以在脑子里对很多数学问题有很好的直觉，但指数增长通常很难让人快速建立心理框架。不管出于什么原因，进化需要我们在脑子里擅长很多数学问题，但模拟指数增长似乎不是其中之一。所以我们相信，我们可以在相当长一段时间内保持非常陡峭的收入增长曲线，我们现在看到的一切都继续表明这一点。如果没有算力，我们就做不到。再次强调，我们非常受算力限制，这对收入线影响很大，我认为如果我们到了有很多闲置算力、无法以单位算力盈利的方式变现的地步，那么说「好吧，这有点……这怎么行得通？」是非常合理的。但我们用多种方式估算过。当然，随着我们降低算力成本的所有工作实现，我们也会在每 flops 的成本上变得更高效。但我们看到了消费者增长，看到了企业增长。还有一大堆我们尚未推出但即将推出的新业务。但算力确实是这一切的生命线。所以沿途有一些检查点，如果我们的时机或计算稍有偏差，我们有一些灵活性。但我们一直处于算力短缺状态。它一直限制着我们的能力。不幸的是，我认为这种情况会一直存在，但我希望情况能有所改善，随着时间的推移，我希望它能变得不那么严重，因为我认为我们可以交付很多优秀的产品和服务，这将是一项伟大的业务。

It's very hard to really... I find that one thing I certainly can't do it and very few people I've ever met can do it. You can have good intuition for a lot of mathematical things in your head, but exponential growth is usually very hard for people to do a good quick mental framework on. For whatever reason, there were a lot of things that evolution needed us to be able to do well with math in our heads. Modeling exponential growth doesn't seem to be one of them. So the thing we believe is that we can stay on a very steep growth curve of revenue for quite a while and everything we see right now continues to indicate that. We cannot do it if we don't have the compute. Again, we're so compute constrained and it hits the revenue line so hard that I think if we get to a point where we have a lot of compute sitting around that we can't monetize on a profitable per unit of compute basis, it would be very reasonable to say, 'Okay, this is like a little how's this all going to work?' But we've penciled this out a bunch of ways. We will of course also get more efficient on a flops per dollar basis as all of the work we've been doing to make compute cheaper comes to pass. But we see this consumer growth, we see this enterprise growth. There's a whole bunch of new kinds of businesses that we haven't even launched yet, but will. But compute is really the lifeblood that enables all of this. So there are checkpoints along the way and if we're a little bit wrong about our timing or math, we have some flexibility. But we have always been in a compute deficit. It has always constrained what we're able to do. I unfortunately think that will always be the case, but I wish it were less the case and I'd like to get it to be less of the case over time because I think there are so many great products and services that we can deliver and it'll be a great business.

所以实际上训练成本占比在下降。总体上是大幅上升，但没错。然后你的期望是通过企业推广、通过人们愿意通过 API 为 ChatGPT 付费，OpenAI 将能够增长收入，足以用收入来支付成本。

So it's effectively training costs go down as a percentage. They go up massively overall, but yeah. And then your expectation is through things like this enterprise push, through things like people being willing to pay for ChatGPT through the API, OpenAI will be able to grow revenue enough to pay for it with revenue.

是的。这就是计划。

Yeah. That is the plan.

现在，我认为市场最近对此有点失去理智。我认为让市场恐慌的是债务进入了这个等式。债务的概念是，当有可预测的东西时，你会举债。然后公司会举债，建设，然后有可预测的收入。但这是一个新类别。它是不可预测的。你如何看待债务进入这里的事实？

Now, I think the thing so the market's been kind of losing its mind over this recently. I think the thing that has spooked the market has been the debt has entered into this equation. And the idea around debt is you take debt out when there's something that's predictable. And then companies will take the debt out, they'll build, and they'll have predictable revenue. But it's a new category. It is unpredictable. How do you think about the fact that debt has entered the picture here?

首先，我认为市场在今年早些时候更失去理智，当时我们与某家公司会面，那家公司的股票第二天就上涨了 20%或 15%。那太疯狂了。感觉非常不健康。实际上，我很高兴现在市场上多了一点怀疑和理性，因为我觉得我们当时完全朝着一个非常不稳定的泡沫前进，现在人们有了一定程度的纪律。所以实际上我认为事情……我认为人们之前疯了，现在更理性了。在债务方面，我认为我们确实知道，如果我们建设基础设施，这个行业，总会有人从中获得价值。而且现在还为时过早。我同意你的观点，但我不认为还有人质疑 AI 基础设施不会产生价值。所以，我认为债务进入这个市场是合理的。我认为还会有其他类型的金融工具。

First of all, I think the market more lost its mind when earlier this year, we would meet with some company and that company's stock would go up 20% or 15% the next day. That was crazy. That felt really unhealthy. I'm actually happy that there's a little bit more skepticism and rationality in the market now because it felt to me like we were just totally heading towards a very unstable bubble and now I think people have some degree of discipline. So I actually think things are... I think people went crazy earlier and now people are being more rational. On the debt front, I think we do kind of know that if we build infrastructure, the industry, someone's going to get value out of it. And it's still totally early. I agree with you, but I don't think anyone's still questioning there's not going to be value from AI infrastructure. And so, I think it is reasonable for debt to enter this market. I think there will also be other kinds of financial instruments.

我怀疑我们会看到一些不合理的融资方式，因为人们真的在创新如何为这类事情融资，但借钱给公司建数据中心在我看来没问题。我担心的是，如果进展不能持续。这里有一个场景，你可能不同意，但模型进步饱和了。那么基础设施的价值就会低于预期。是的，那些数据中心对某些人来说还是有价值的，但可能会被清算，有人以折扣价买下。我确实怀疑一路上会有一些繁荣和萧条。这些事情从来都不是一条完美的直线。首先，在我看来非常清楚，而且我乐意拿公司打赌，模型会变得好得多。我们对这一点有相当好的了解，非常有信心。即使模型没有进步，我认为世界上有很多惯性。适应事物需要时间。我相信 5.2 所代表的经济价值相对于世界目前已经学会从中提取的价值，其过剩是如此巨大，即使你把模型冻结在 5.2，你还能创造多少价值，从而驱动多少收入？我打赌是巨大的。事实上，你没问这个，但如果允许我跑题一下，我们过去经常讨论这个 2x2 矩阵：短时间线、长时间线、慢速起飞、快速起飞。我们觉得在不同时间概率在转移，你可以根据你在这个 2x2 矩阵中的位置来理解世界应该优化的许多决策和策略。我脑海中这个图景出现了一个 Z 轴，即小过剩、大过剩。我有点认为我肯定假设过剩不会那么巨大。如果模型有很多价值，世界会很快学会如何部署它。但现在在我看来，过剩在世界上大部分地区将是巨大的。你会看到一些领域，比如一些程序员，通过采用这些工具会变得极其高效。但总的来说，你有一个极其聪明的模型，老实说，大多数人还在问他们在 GPT-4 时代问的类似问题。科学家不同，程序员不同，也许知识工作会不同，但存在巨大的过剩。这对世界有一系列非常奇怪的后果。我们还没有完全理解它将以何种方式展开，但这与我几年前预期的非常不同。

I suspect we'll see some unreasonable ones as people really innovate about how to finance this sort of stuff, but lending companies money to build data centers seems fine to me. I think the fear is that if things don't continue apace. Here's one scenario. And you'll probably disagree with this, but the model progress saturates. Then the infrastructure becomes worth less than the anticipated value. And yes, those data centers will be worth something to someone, but it could be that they get liquidated and someone buys them at a discount. I do suspect there will be some booms and busts along the way. These things are never a perfectly smooth line. First of all, it seems very clear to me, and this is something I would happily bet the company on, that the models are going to get much, much better. We have a pretty good window into this. We're very confident about that. Even if they did not, I think there is a lot of inertia in the world. It takes a while to figure out how to adapt to things. The overhang of the economic value that I believe 5.2 represents relative to what the world has figured out how to get out of it so far is so huge that even if you froze the model at 5.2, how much more value can you create and thus revenue can you drive? I bet a huge amount. In fact, you didn't ask this, but if I can go on a rant for a second, we used to talk a lot about this two-by-two matrix of short timelines, long timelines, slow takeoff, fast takeoff. And where we felt at different times the kind of probability was shifting, you could understand a lot of the decisions and strategy that the world should optimize for based on where you were going to be on that two-by-two matrix. There is a Z axis in my head in my picture of this that has emerged, which is small overhang, big overhang. I kind of thought that I must have assumed that the overhang was not going to be that massive. That if the models had a lot of value in them, the world was pretty quickly going to figure out how to deploy that. But it looks to me now like the overhang is going to be massive in most of the world. You'll have these areas like some set of coders that will get massively more productive by adopting these tools. But on the whole, you have this crazy smart model that, to be perfectly honest, most people are still asking the similar questions they did in the GPT-4 realm. Scientists different, coders different, maybe knowledge work is going to get different, but there is a huge overhang. And that has a bunch of very strange consequences for the world. We have not wrapped our head around all the ways that's going to play out yet, but it is very much not what I would have expected a few years ago.

我有一个关于这种能力过剩的问题。基本上，模型能做的比它们正在做的多得多。我试图理解模型怎么能比它们被使用的场景好那么多，但很多企业在尝试实施时并没有获得投资回报。或者至少他们告诉 MIT 是这样。我不太确定该怎么想，因为我们听到所有这些企业说，如果你把 GPT-5.2 的价格提高 10 倍，我们仍然会付钱。你严重低估了它。我们从中获得了所有这些价值。所以，我觉得这似乎不对。

I have a question for you about this capability overhang. Basically, the models can do a lot more than they've been doing. I'm trying to figure out how the models can be that much better than they're being used for, but a lot of businesses when they try to implement them, they're not getting a return on their investment. Or at least that's what they tell MIT. I'm not sure quite how to think about that because we hear all these businesses saying, if you 10x the price of GPT-5.2, we would still pay for it. You're hugely underpricing this. We're getting all this value out of it. So, I don't think that seems right to me.

当然，如果你听程序员怎么说，他们会说，我愿意付 100 倍的价格之类的。只是官僚主义把事情搞砸了。假设你相信 GDP 价值数字，也许你不信有充分理由。也许它们错了，但假设是真的。对于这些明确指定的、不是特别长时间的知识工作任务，十次中有七次你会对 5.2 的输出感到同样满意或更满意。那么你应该大量使用它。然而，人们改变工作流程需要很长时间。他们太习惯于让初级分析师做演示文稿之类的，这比我预想的更顽固。我自己的工作流程也基本没变，尽管我知道我可以更多地使用 AI。

Certainly, if you talk about what coders say, they're like, I'd pay 100 times the price or whatever. Just be bureaucracy that's messing things up. Let's say you believe the GDP value numbers and maybe you don't for good reason. Maybe they're wrong, but let's say it were true. And for kind of these well-specified, not super long duration knowledge work tasks, seven out of 10 times you would be as happy or happier with the 5.2 output. You should then be using that a lot. And yet, it takes people so long to change their workflow. They're so used to asking the junior analyst to make a deck or whatever that they're going to, it's just stickier than I thought it was. I still kind of run my workflow in very much the same way, although I know that I could be using AI much more than I am.

好了，我们还有 10 分钟。我有四个问题。我们看看能不能快速过一遍。那么，你正在开发的设备。我们稍后回来继续采访 OpenAI CEO Sam Altman。我听说的是，手机大小，没有屏幕。如果它是没有屏幕的手机，为什么不能是一个应用呢？

All right, we got 10 minutes left. I got four questions. Let's see if we can lightning round through them. So, the device that you're working on. We'll be back with OpenAI CEO Sam Altman right after this. What I've heard, phone size, no screen. Why couldn't it be an app if it's the phone without a screen?

首先，我们会做一个小型设备系列。不会只有一个设备。随着时间的推移，人们使用计算机的方式会发生转变，从一种愚蠢的被动反应式变成一种非常智能的主动式，它理解你的整个生活、你的背景、你周围发生的一切，非常了解你物理上或通过计算机接近的人。我认为当前设备不适合那种世界。我坚信我们在设备的极限上工作。你有一台电脑，它有一堆设计选择。比如它可以打开或关闭，但它不能，比如，注意这个采访，但保持关闭，如果我忘记问 Sam 一个问题就在我耳边低语之类的。也许那会有帮助。而且有一个屏幕，这限制了你，就像我们几十年来使用图形用户界面一样，还有一个键盘，它的设计是为了减慢你输入信息的速度。这些长期以来都是毋庸置疑的假设，但它们有效。然后这个全新的东西出现了，它打开了一个可能性空间，但我认为当前设备的形态不是最佳适配。如果它适合我们拥有的这个不可思议的新能力，那会非常奇怪。

First, we're going to do a small family of devices. It will not be a single device. There will be over time a shift to the way people use computers where they go from a sort of dumb reactive thing to a very smart proactive thing that is understanding your whole life, your context, everything going on around you, very aware of the people around you physically or close to you via computer that you're working with. And I don't think current devices are well suited to that kind of world. I am a big believer that we work at the limit of our devices. You have that computer and it has a bunch of design choices. Like it could be open or closed, but it can't be, you know, okay, pay attention to this interview, but be closed and whisper in my ear if I forget to ask Sam a question or whatever. Maybe that would be helpful. And there is a screen and that limits you to the same way we've had graphical user interfaces working for many decades and a keyboard that was built to slow down how fast you could get information into it. These have just been unquestioned assumptions for a long time, but they worked. And then this totally new thing came along and it opens up a possibility space, but I don't think the current form factor of devices is the optimal fit. It would be very odd if it were for this incredible new affordance we have.

天哪，我们可以聊一个小时这个，但让我们继续下一个话题，云。你谈过要建一个云。这里有一封听众的邮件。在我的公司，我们正在从 Azure 迁移，直接与 OpenAI 集成，为产品中的 AI 体验提供动力。重点是插入一个万亿级 token 流，通过整个技术栈驱动 AI 体验。

Oh man, we could talk for an hour about this, but let's move on to the next one, cloud. You've talked about building a cloud. Here's an email we got from a listener. At my company, we're moving off Azure and directly integrating with OpenAI to power our AI experiences in the product. The focus is to insert a stream of trillions of tokens powering AI experiences through the stack.

这就是打造大型云业务的计划吗？

Is that the plan to build a big cloud business in that way?

首先，数万亿个 token——很多 token。你问到了企业战略中对算力的需求。企业已经很清楚地向我们表达了他们想从我们这里购买多少 token，而我们在 2026 年将再次无法满足需求。但策略是：大多数公司似乎希望找到像我们这样的公司，然后说：「我想用 AI 命名我的公司。我需要一个为我公司定制的 API。我需要一个为我公司定制的 ChatGPT 企业版。我需要一个平台，可以运行所有这些智能体，并且我能信任我的数据。我需要有能力将数万亿个 token 融入我的产品。我需要有能力让我所有的内部流程更高效。」而我们目前没有一个很好的一体化方案给他们。我们想做出这样的方案。

First of all, trillions of tokens — a lot of tokens. And you asked about the need for compute in our enterprise strategy. Enterprises have been pretty clear with us about how many tokens they'd like to buy from us, and we are going to again fail in 2026 to meet demand. But the strategy is: most companies seem to want to come to a company like us and say, 'I'd like to name my company with AI. I need an API customized for my company. I need ChatGPT Enterprise customized for my company. I need a platform that can run all these agents that I can trust my data on. I need the ability to get trillions of tokens into my product. I need the ability to have all my internal processes be more efficient.' And we don't currently have a great all-in-one offering for them. And we'd like to make that.

你的雄心是让它与 AWS 和 Azure 等并驾齐驱吗？

Is your ambition to put it up there with the AWS and Azures of the world?

我认为这和那些是不同的事情。我并没有雄心去提供托管网站所需的所有服务之类的。但我觉得这个概念——我的猜测是，人们会继续拥有他们所谓的网络云。然后我认为会有另一种东西，公司会说：「我需要一个 AI 平台，用于我想在内部做的一切，我想提供的服务，等等。」它在某种意义上确实依赖于物理硬件，但我认为它会是一个相当不同的产品。

I think it's a different kind of thing than those. I don't really have an ambition to go offer all the services you have to offer to host a website or whatever. But I think the concept — my guess is that people will continue to have their so-called web cloud. And then I think there will be this other thing where a company will be like, 'I need an AI platform for everything that I want to do internally, the service I want to offer, whatever.' And it does kind of live on the physical hardware in some sense, but I think it'll be a fairly different product offering.

我们快速谈谈发现。你说过一些让我很感兴趣的话。你认为模型——或者可能是人与模型合作——明年会做出小发现，五年内做出大发现。是模型吗？还是人与模型合作？是什么让你确信这会发生？

Let's talk about discovery quickly. You've said something that's been really interesting to me. You think that models — or maybe it's people working with models — will make small discoveries next year and big ones within five. Is that the models? Is it people working alongside them? And what makes you confident that's going to happen?

使用模型的人。能够自己提出问题的模型——那感觉还更遥远。但如果世界从新知识中受益，我们应该非常兴奋。我认为人类进步的整个过程就是：我们建造更好的工具，然后人们用它们做更多的事情，在这个过程中他们建造更多的工具。这是我们一层层攀登的脚手架，一代又一代，一个发现又一个发现。人类提出问题的事实绝不会削弱工具的价值。所以我认为这很棒。今年年初，我以为小发现会在 2026 年开始。它们从 2025 年开始，2025 年底。再说一次，这些发现非常小。我真的不想夸大它们，但任何东西在性质上都感觉与没有完全不同。当然，三年前我们推出 ChatGPT 时，那个模型不会对人类知识总量做出任何新贡献。从现在到五年后，通往大发现的旅程——我怀疑它看起来就像 AI 的正常爬坡。每个季度都会好一点，然后突然之间，我们就会「哇」。被这些模型增强的人类正在做五年前人类绝对做不到的事情。无论我们主要将其归因于更聪明的人类还是更聪明的模型，只要我们能得到科学发现，我都很高兴。

People using the models. Models that can figure out their own questions to ask — that does feel further off. But if the world is benefiting from new knowledge, we should be very thrilled. I think the whole course of human progress has been that we build these better tools, and then people use them to do more things, and out of that process they build more tools. It's this scaffolding that we climb layer by layer, generation by generation, discovery by discovery. The fact that a human is asking the question in no way diminishes the value of the tool. So I think it's great. At the beginning of this year, I thought the small discoveries were going to start in 2026. They started in 2025, in late 2025. Again, these are very small. I really don't want to overstate them, but anything feels qualitatively very different than nothing. Certainly when we launched ChatGPT 3 years ago, that model was not going to make any new contribution to the total of human knowledge. From here to 5 years from now, this journey to big discoveries — I suspect it just looks like the normal hill climb of AI. It just gets a little bit better every quarter, and then all of a sudden we're like, 'Whoa.' Humans augmented by these models are doing things that humans 5 years ago just absolutely couldn't do. Whether we mostly attribute that to smarter humans or smarter models, as long as we get the scientific discoveries, I'm very happy either way.

明年 IPO？我不知道。你想成为一家上市公司吗？你似乎可以长期保持私有。你会提前上市吗？从资金角度来说。

IPO next year? I don't know. Do you want to be a public company? You seem like you can operate private for a long time. Would you go before you needed to? In terms of funds.

这里有很多因素在起作用。我确实认为公开市场能够参与价值创造是很酷的。从某种意义上说，如果与任何之前的公司相比，我们上市会非常晚。作为一家私有公司很棒。我们需要大量资本。我们会在某个时候突破所有股东限制等等。所以我对于成为一家上市公司的 CEO 感到兴奋吗？0%。我对于 OpenAI 成为一家上市公司感到兴奋吗？在某些方面是的，在某些方面我认为这会非常烦人。

There are a whole bunch of things that play here. I do think it's cool that public markets get to participate in value creation. In some sense, we will be very late to go public if you look at any previous company. It's wonderful to be a private company. We need lots of capital. We're going to cross all the shareholder limits and stuff at some point. So am I excited to be a public company CEO? 0%. Am I excited for OpenAI to be a public company? In some ways I am, and in some ways I think it'll be really annoying.

我仔细听了你对 Theo Von 的采访。很棒的采访。他真的很酷。

I listened to your Theo Von interview very closely. Great interview. He was really cool.

Theo 真的知道他在说什么。

Theo really knows what he's talking about.

还引用了 Yoshua Bengio。他做了功课。你告诉他，就在 GPT-5 发布之前，GPT-5 几乎在所有方面都比我们聪明。我以为那就是 AGI 的定义。那不是 AGI 吗？如果不是，这个词是不是变得有些无意义了？

Also citing Yoshua Bengio. He did his homework. You told him, this was right before GPT-5 came out, that GPT-5 is smarter than us in almost every way. I thought that was the definition of AGI. Isn't that AGI? And if not, has the term become somewhat meaningless?

这些模型在原始算力基础上显然非常聪明。最近几天有很多关于 GPT-5.2 智商 147 或 144 或 151 之类的说法——取决于谁的测试，总之是个很高的数字。而且有很多领域专家说它能做这些惊人的事情，它在贡献，它让我更有效率。还有我们谈到的 GDP 增长。但有一件事你没有：模型今天不能做某事，意识到自己不能，然后去弄清楚如何学习变得擅长那件事，学会理解它，第二天你回来时它就能做对了。那种持续学习——幼儿都能做到。在我看来，这似乎是我们需要构建的一个重要部分。那么，没有这一点，你能拥有大多数人认为的 AGI 吗？我会说显然——有很多人会说我们当前的模型已经是 AGI 了。我认为几乎所有人都会同意，如果我们有当前水平的智能再加上那件事，那显然非常像 AGI。但也许世界上大多数人会说：「好吧，即使没有那个。它完成了大多数重要的知识任务。它在大多数方面比我们大多数人都聪明。我们到了 AGI。」它发现了小部分新科学，我们到了 AGI。我认为这意味着这个词，尽管我们很难停止使用它，但定义非常不明确。有一件事我很喜欢：我们在 AGI 上搞错了。我们从未定义过它。现在每个人关注的新词是当我们达到超级智能时。所以我的建议是，我们同意 AGI 已经呼啸而过。它并没有改变世界那么多——或者长期来看会，但好吧，我们已经构建了 AGI。

These models are clearly extremely smart on a sort of raw horsepower basis. There's all this stuff in the last couple of days about GPT-5.2 having an IQ of 147 or 144 or 151 or whatever — depending on whose test, it's some high number. And you have a lot of experts in their field saying it can do these amazing things, it's contributing, it's making me more effective. You have the GDP growth things we talked about. One thing you don't have is the ability for the model to not be able to do something today, realize it can't, go off and figure out how to learn to get good at that thing, learn to understand it, and when you come back the next day it gets it right. That kind of continuous learning — toddlers can do it. It does seem to me like an important part of what we need to build. Now, can you have something that most people would consider an AGI without that? I would say clearly — there are a lot of people that would say we're at AGI with our current models. I think almost everyone would agree that if we were at the current level of intelligence and have that other thing, it would clearly be very AGI-like. But maybe most of the world will say, 'Okay, fine, even without that. It's doing most knowledge tasks that matter. It's smarter than most of us in most ways. We're at AGI.' It's discovering small pieces of new science, we're at AGI. What I think this means is that the term, although it's been very hard for all of us to stop using, is very underdefined. One thing I would love is we got it wrong with AGI. We never defined that. The new term everyone's focused on is when we get to superintelligence. So my proposal is that we agree that AGI kind of went whooshing by. It didn't change the world that much — or it will in the long term, but okay, fine, we've built AGIs.

在某个时刻，我们处在一个模糊的时期：有些人认为我们已经有了超级智能，有些人认为还没有，更多人认为已经有了，然后我们会说，「好吧，接下来呢？」超级智能的一个候选定义是：当一个系统能比任何人（即使借助 AI 的帮助）更好地担任美国总统、大型公司 CEO 或管理一个非常大的科学实验室时。

At some point, we're in this fuzzy period where if some people think we have, some people think we have, and more people think we have, and then we'll say, 'Okay, what's next?' A candidate definition for superintelligence is when a system can do a better job being president of the United States, CEO of a major company, running a very large scientific lab than any person can, even with the assistance of AI.

我认为国际象棋的发展很有意思：它达到了能击败人类的水平。我对此记忆犹新，就是深蓝那件事。然后有一段时间，人类和 AI 联手比单独的 AI 更强。但后来人类反而拖后腿，最聪明的做法是让没有人类干预的 AI 独自运行，就像不理解它那伟大的智能一样。我认为类似这样的框架对于超级智能来说很有趣。我觉得这还很遥远，但我希望这次能有一个更清晰的定义。

I think this was an interesting thing about what happened with chess, where chess got it could beat humans. I remember this very vividly. The Deep Blue thing. And then there was a period of time where a human and the AI together were better than an AI by itself. And then the person was just making it worse and the smartest thing was the unaided AI that didn't have the human, like not understanding its great intelligence. I think something like that is an interesting framework for superintelligence. I think it's a long way off, but I would love to have a cleaner definition this time around.

嗯，Sam，你看，我已经使用你的产品三年了，每天都用。

Well, Sam, look, I have been in your products, using them daily for 3 years.

谢谢你，伙计。确实进步了很多。简直无法想象它们未来会发展到什么地步。

Thank you, man. Definitely gotten a lot better. Can't even imagine where they go from here.

我们会努力让它们快速变得更好。好的。这是我们第二次交谈，我很感激你两次都如此坦诚。所以感谢你的时间。

We'll try to keep getting them better fast. Okay. And this is our second time speaking and I appreciate how open you've been both times. So thank you for your time.

感谢大家的收听和观看。如果你是第一次来，请点击关注或订阅。我们的节目中有很多精彩的访谈，还有更多即将上线。过去一年，我们邀请了 Google DeepMind CEO Demis Hassabis 两次，其中一次还有 Google 创始人 Sergey Brin。我们还邀请了 Anthropic 的 CEO Dario Amodei。2026 年我们还有大量重磅访谈。再次感谢，我们下次在 Big Technology Podcast 再见。

Thank you, everybody, for listening and watching. If you're here for the first time, please hit follow or subscribe. We have lots of great interviews on the feed and more on the way. This past year we've had Google DeepMind CEO Demis Hassabis on twice, including one with Google founder Sergey Brin. We've also had Dario Amodei, the CEO of Anthropic. And we have plenty of big interviews coming up in 2026. Thanks again and we'll see you next time on Big Technology Podcast.