英伟达CEO黄仁勋：OpenAI将成万亿美元级巨头，三大定律预示AI计算需求的爆发

主讲人: Bg2 Pod

来源: 查看

发布时间: 2025-10-06

摘要:

英伟达CEO黄仁勋与主持人探讨了AI计算的未来，预测OpenAI将成为下一个万亿美元级的超大规模公司。他详细阐述了预训练、后训练和推理这“三大扩展定律”对计算需求的巨大推动作用，并讨论了公司与OpenAI的深度合作、ASIC竞争的本质，以及美国在吸引全球AI人才方面的战略。

对OpenAI未来前景的预测

I think that OpenAI is likely going to be the next multi-trillion dollar hyperscale company.

我认为OpenAI很可能会成为下一个万亿美元级的超大规模公司。

重新聚首与AI时间感

Welcome to Invidia. Uh, oh, and nice glasses. Um, those actually look really good on you. The problem is now everybody's going to want you to wear them all the time. They're going to say, "Where are the red glasses?" I can vouch for that. So, it's been over a year since we did the last pod.

欢迎来到英伟达。哦，你这新眼镜真不错。它们戴在你脸上看起来真棒。问题是，现在所有人都希望你一直戴着它们，他们会说：“红色的眼镜去哪儿了？”我能证明这一点。所以，自我们上次播客以来已经过去了一年多。

推理计算需求的巨大飞跃

Yeah. Over 40% of your revenue today is inference, but inference is about ready because of chain of reasoning. Yeah. Right. It's about It's about to go up by a billion times, right? By by a million x by a billion. That's right. That's the part that most people have, you know, haven't completely internalized. This is that industry we were talking about. This is the industrial revolution.

是的。你今天超过40%的收入来自推理，但推理即将到来，因为有了推理链。是的。没错。它即将增长十亿倍，对吧？增长一百万倍，增长十亿倍。没错。这是大多数人还没有完全内化的部分。这就是我们一直在谈论的那个行业。这就是工业革命。

AI时间线的飞速发展

Honestly, it's it's felt like you and I have had a continuation of the pod every day since then. You know, in AI time, it's been about a hundred years. I was re-watching the pod recently and the many things that we talked about that stood out. The most the one that that was probably most profound for me was you pounding the table that you know remember at the time there was kind of a slump in terms of pre-training and people were like, "Oh my god, the end of pre-training, right? The end of pre-training. We're not going. We're overbuilding. This is about a year and a half ago. And you said inference isn't going to 100x, a thousandx, it's going to 1 billionx.

老实说，感觉我们俩从那时起每天都在延续这个播客。你知道，在人工智能的时间里，这感觉像是一百年。我最近重温了那期播客，我们谈到的很多事情都给我留下了深刻印象。最让我感到深刻的是你当时拍着桌子说，记得那时候预训练似乎出现了一段低迷期，人们还说：“我的天，预训练要结束了，对吧？预训练要结束了。我们不能再过度建设了。这是大约一年前的事情了。而你说推理不会是100倍、1000倍的增长，而是会增长10亿倍。”

三大扩展定律的整合

Mhm. Which brings us to where we are today. You know, you announced this huge deal. We ought to start there. I underestimated. Let me just go on record. I estimated we we now have three scaling laws, right? We have pre-training scaling law. We have post-training scaling law. Post-training is basically like uh AI practicing. Yes, practicing a skill until it gets it right. And so it tries a whole bunch of different ways and and uh in order to do that, yeah, you've got to do inference. So now training and inference are now integrated in reinforcement learning. Really complicated. And so that's called post training. And then the third is inference. The old way of doing inference was one shot, right? But the new way of doing inference, which we appreciate, is thinking. So think before you answer. Yeah. And so now you have three scaling laws. The the longer you think, the better the quality answer you get. While you're thinking, you do research, you go check on some ground truth. And you you learn some things, you think some more, you go learn some more, and then you generate an answer. Don't just generate right off the bat. And so thinking, post-training, pre-training, we now have three scaling laws, not one.

嗯。这就把我们带到了今天。你知道，你宣布了这笔巨大的交易。我们应该从那里开始。我低估了。让我公开声明一下。我估计我们现在有三个扩展定律，对吧？我们有预训练扩展定律。我们有后训练扩展定律。后训练基本上就像是人工智能在练习。是的，练习一项技能直到做好为止。所以它会尝试很多不同的方法，嗯，为了做到这一点，是的，你必须进行推理。所以现在训练和推理在强化学习中结合起来了。这非常复杂。所以这就叫做后训练。第三个是推理。过去推理的方式是一次性的，对吧？但我们所欣赏的新的推理方式是思考。所以，三思而后答。是的。所以现在你有三个扩展定律。你思考得越久，得到的答案质量就越高。在你思考的时候，你做研究，你去核实一些事实真相。你学到了一些东西，你再多想一想，再去学习更多，然后你生成一个答案。不要立即给出答案。所以，思考、后训练、预训练，我们现在有了三个扩展定律，而不是一个。

对推理增长的信心增强

You knew that last year, but is your level of confidence this year in the inferences going to 1 billionx and where that will take the levels of intelligence is it higher? Are you more confident this year than you were a year ago? I'm more confident this year and the reason for that is because look at the agent systems now and AI is no longer a language model and AI is a system of language models and they're all running concurrently maybe using tools. Some of us using tools, some of us doing research and yeah, there's a whole bunch of stuff and it's all multimodality and look at all the video that's being generated. I mean, it's just crazy stuff.

你去年就知道这一点了，但你今年对推理将增长10亿倍以及这将把智能水平带到何种程度的信心是否更高？你比一年前更有信心吗？我今年的信心更足了，原因在于看看现在的智能体系统，人工智能不再是语言模型，人工智能是一个语言模型的系统，它们都在并发运行，可能使用工具。有些人使用工具，有些人做研究，是的，有很多东西，而且都是多模态的，看看正在生成的所有视频。我的意思是，这太疯狂了。

OpenAI Stargate 合作的深入理解

It really brings us to, you know, kind of the seminal moment this week that everybody's talking about the massive deal. You announced a couple days ago with OpenAI Stargate where that you're going to be a preferred partner, invest hundred billion dollars in the company over a period of time. They're going to build 10 gigs and if they used Nvidia for those 10 gigs that could be upwards of 400 billion in revenue to Nvidia. So help us understand you just tell us a little bit about that partnership what it means to you right and why that investment makes so much sense for Nvidia.

这确实把我们带到了本周每个人都在谈论的那个开创性的时刻，就是你几天前宣布的与OpenAI Stargate的巨额交易，你将成为首选合作伙伴，在一段时间内向该公司投资1000亿美元。他们将建造10个千兆瓦的设施，如果他们为这10个千兆瓦的设施使用英伟达，那可能为英伟达带来高达4000亿美元的收入。所以请帮助我们理解，告诉我一些关于这次合作、它对你的意义，以及为什么这项投资对英伟达如此有意义。

投资OpenAI的原因：下一家万亿美元巨头

So first of all that I'll answer that last question first and then I'll come back and present my way through. I think that OpenAI is likely going to be the next multi-trillion dollar hyperscale company. Okay. I think you and I Why do you call it a hyperscale company? Hyperscale like uh like Meta is a hyperscale. Uh Google's a hyperscale. They're going to have consumer and enterprise services and and uh they are very likely going to be the world's next multi-t trillion dollar hyperscale company. Yes. And I think you would agree with that. I agree. If that's the case, the opportunity to invest before they get there, this is some of the smartest investments we can possibly imagine. And you got to invest in things, you know, right? And it turns out we happen to know this space. And so the opportunity to invest in that, the return on that money is going to be fantastic. So we love the opportunity to invest. We don't have to invest, right? And it's not required for us to invest, but they're giving us the opportunity to invest. Fantastic thing.

首先，我先回答最后一个问题，然后再介绍我的看法。我认为OpenAI很可能会成为下一个万亿美元级的超大规模公司。好的。你为什么称之为超大规模公司？超大规模就像Meta一样，谷歌也是超大规模。他们将拥有消费者和企业服务，而且他们极有可能成为世界上下一家万亿美元级的超大规模公司。是的。我想你也会同意。我同意。如果是这样的话，在他们达到那个水平之前进行投资，这是我们能想到的最明智的投资之一。而且你知道，你必须投资于你知道的领域，对吧？事实证明，我们恰好了解这个领域。所以投资于此，这笔钱的回报将会非常可观。所以我们非常喜欢这次投资的机会。我们不一定非要投资，对吧？投资不是我们的要求，但他们给了我们投资的机会。这太棒了。

英伟达与OpenAI合作的三个项目

Now let me start from the beginning. So we're partnering with with OpenAI in several projects. First the first project is the buildout of Microsoft Azure. We're going to continue to do that and that that partnership is going fantastically. We're going we have several years of buildout to do hundreds of billions of dollars of work just to do there. Right. The second is the OCI buildout and uh I think there's some five, six, seven gigawatts that are about to be built out. And so we're working with OCI and OpenAI and and SoftBank to build that out, right? Those projects are contracted. Um we're working on it. Lots of work to do. And then the third the third is Core Weave, right? And so uh all of Core Wee 4, I'm talking about OpenAI still. Yes. Okay. Everything in the context of OpenAI.

现在让我从头开始。我们正在与OpenAI在几个项目中合作。首先，第一个项目是微软Azure的建设。我们将继续这样做，而且这次合作进展得非常顺利。我们有几年的建设工作要做，仅此一项就涉及数千亿美元的工作。第二个是OCI的建设，嗯，我想有五六七吉瓦的设施即将建成。所以我们正在与OCI、OpenAI和软银合作来建设它，对吧？这些项目都有合同。我们正在努力。有很多工作要做。然后是第三个，第三个是Core Weave，对吧？我说的都是关于OpenAI的。是的。好的。所有内容都在OpenAI的背景下。

助力OpenAI自建AI基础设施

And so so the question is what is this new partnership? This this new partnership is about helping OpenAI working partnering with OpenAI to build their own selfbuild AI infrastructure for the first time, right? And so this is us working directly with OpenAI at the chip level, at the software level, at the systems level, at the AI factory level um to help them become a fully operated hyperscale company that I mean this is going to go on for some time. And it's going to supplement it's going to supplement the amount of you know they're going through two exponentials as you know right the first exponential is the number of customers is growing exponentially and the reason for that is the AI is getting better the use case is getting better just about every every application is connected to open now and so they're going through the the usage exponential the second exponential is the computational exponential of every use.

所以，那么问题来了，这个新的合作是什么？这个新的合作是帮助OpenAI与OpenAI合作，首次建立他们自己的自建AI基础设施，对吧？所以这是我们直接与OpenAI在芯片层面、软件层面、系统层面、AI工厂层面合作，帮助他们成为一个完全运营的超大规模公司，我的意思是这将持续一段时间。它将补充——你知道，他们正在经历两个指数级增长。第一个指数增长是客户数量呈指数级增长，原因是AI变得越来越好，用例越来越好，几乎每一个应用现在都与OpenAI相关联，所以他们正在经历使用率的指数增长。第二个指数增长是每次使用的计算指数。

思考能力对推理需求的指数级影响

Yes. Right. Yes. Instead of just a oneshot inference is now thinking before it answers. Yeah. And so these two exponential is compounding their compute requirements. And so we we got to build out these all these different projects. And so this last one is an additive on top of everything that they've already announced, all the things that we're already working on with them. It's additive on top of that and uh it's going to support the you know this incredible exponential growth.

是的。对。是的。推理不再是一次性的，而是现在在回答之前会思考。是的。所以这两个指数正在使他们的计算需求复合增长。所以我们必须建立所有这些不同的项目。所以最后一个项目是对他们已经宣布的所有事情、我们正在与他们合作的所有事情的补充。它是附加的，它将支持，你知道，这种惊人的指数增长。

从外包到自建基础设施的转变

One of the things you said there that's really interesting to me is kind of you know they're going to be high probability multi-t trillion dollar company in your mind. I think it's a great investment. At the same time, you know, they're selfbuilding. You're helping them self-build their data centers. So, here to four, they've been outsourcing to Microsoft to build the data center. Now they want to build full stack factories themselves. They want to do they want to they want to basically have a relationship with us the way that that Elon and X has relation. Correct. I mean, Elon and Exel built. Exactly.

你刚才说的一件事对我来说非常有趣，那就是在你看来，他们很可能成为一家高概率的万亿美元级公司。我认为这是一项伟大的投资。同时，你知道，他们正在自建。你在帮助他们自建数据中心。所以，到目前为止，他们一直将数据中心的建设外包给微软。现在他们想自己建造全栈工厂。他们想——他们基本上想和我们建立一种关系，就像埃隆和X（指X.AI）与我们建立关系一样。没错。我的意思是，埃隆和X.AI已经建立起来了。没错。

基础设施自建带来的超大规模优势

But I think that's this is a very big deal when you think when you think about the advantage that Colossus had. Yeah, they're building full stack. That is a hyperscaler because if they don't use the capacity, they could sell it to somebody else. In the same way Stargate, they're building monstrous capacity. They think they'll need to use most of it, but it puts them in a position to sell it to somebody else as well. It sounds very much like AWS or GCP or Azure. That's what you're saying. Yeah, I I think they'll likely use it themselves and um just think the case of X, they'll likely use it themselves. Um but they would like to have the the same direct relationship with us, direct working relationship and direct purchasing relationship. Um uh Meta just as with Zuck and Meta has with us uh it's exactly a direct um our relationship with uh between us and Sunund and Google direct our partnership with Satia and Azure direct.

但我认为，当你考虑到Colossus所拥有的优势时，这是一个非常重大的举动。是的，他们正在构建全栈，这就是一个超大规模运营商，因为如果他们不使用这些产能，他们可以卖给别人。同样，Stargate正在建设巨大的产能。他们认为他们需要使用大部分，但这使他们有能力将多余的产能出售给其他人。这听起来很像AWS、GCP或Azure。你说的就是这个意思。是的，我认为他们很可能会自己使用。嗯，想想X（指X.AI）的情况，他们很可能会自己使用。但他们希望与我们有同样的直接关系，直接的工作关系和直接的采购关系。嗯，Meta就像扎克伯格和Meta与我们的关系一样，这正是我们与Sundar和Google之间直接的关系，与Satia和Azure的合作也是直接的。

规模化带来的直接采购优势

Isn't that right? And so they've gotten to a large enough scale that they believe it's time for them to start building these direct relationships. So, I'm delighted to support that and and all of and and Satia knows it and Larry knows it and everybody everybody's aware of what's going on and everybody's very supportive of it.

不是吗？所以他们已经达到了一个足够大的规模，他们认为现在是时候建立这些直接关系了。所以我很高兴支持这一点，而且Satia知道，拉里也知道，每个人都知道正在发生什么，每个人都非常支持。

华尔街对英伟达增长预期的分歧

So, one of the things I find mysterious, right? You know, you just mentioned Oracle 300 billion Colossus what they're building. We know what the sovereigns are building. We know what the hyperscalers are building. You know, Sam's talking in terms of trillions. But of the 25 sellside analysts on Wall Street who cover your stock, if I look at the consensus estimate, it basically has your growth flatlining starting in 2027. 8% growth 2027 through 2030. Okay, that is the 25 people in their only job. They get paid to forecast the growth rate for Nvidia. So clearly we're comfortable with that by the way. Right. Look, we're comfortable with that. Okay, we have no trouble beating the numbers on a regular basis, right? No, I understand that. But, but there is this interesting disconnect.

所以，有一件事我觉得很神秘，对吧？你知道，你刚提到了甲骨文3000亿的Colossus他们正在建造的东西。我们知道主权国家在建造什么。我们知道超大规模运营商在建造什么。你知道，萨姆（指Sam Altman）谈论的是万亿级别。但在华尔街覆盖贵公司股票的25位卖方分析师中，如果我看共识估计，它基本上预测到你们的增长从2027年开始趋于平缓。2027年到2030年增长8%。好吧，这是25个只做一件事的人。他们靠预测英伟达的增长率拿薪水。所以很明显，我们对此很满意。对。看，我们对此很满意。好吧，我们定期超越这些数字没有任何问题，对吧？不，我理解。但这里存在这种有趣的脱节。

解释华尔街预期的不一致性

No, right. I hear it every day on CNBC and Bloomberg. And I think it goes to, you know, some of these questions around, you know, uh, shortages leading to a glut that they don't believe. They say, "Okay, we'll give you credit for '26, but '27, you know, maybe we'll have too much and you're not going to need that." But it is interesting to me and I think it's important to point out that the your consensus forecast is that this won't happen right and we also put together forecast uh you know for the company taking into account all of these numbers and what it shows me is still even though we're two and a half years into the age of AI a massive divergence of belief between what we hear Sam Alman saying you saying Sundar saying Satcha is saying and what Wall Street still believes and you know again you're comfortable with that. I also don't think it's inconsistent.

不，没错。我每天在CNBC和彭博社上都能听到。而且我认为这与一些关于短缺导致供过于求的问题有关，他们不相信。他们说：“好吧，我们承认26年的预测，但27年，你知道，也许我们会供过于求，你就不需要那些了。”但对我来说这很有趣，我认为重要的是要指出，你们的共识预测是这种情况不会发生，对吧？我们还根据所有这些数字为公司做出了预测，它向我表明，尽管我们已经进入AI时代两年半了，但在萨姆·阿尔特曼、你、桑达尔、萨蒂亚所说的内容和华尔街仍然相信的内容之间，存在着巨大的信念分歧，而且你再次对此感到满意。我也不认为这有什么不一致。

第一大定律：通用计算的终结与加速计算的未来

Okay. So explain that a little bit. So first of all uh for the builders we're supposed to be building for opportunity right? We're we're builders. Let me give you three points to think through and and and these three points uh it'll help you um hopefully uh be more comfortable with Nvidia in this future. So the first point and this is the laws of physics point. This is the most important point that general general purpose computing is over and the future is accelerated computing and AI computing. That's the first point. And so the way to think about that is there's how much how many trillions of dollars of computing infrastructures in the world that has to be refreshed. Right. Right. And when it gets refreshed it's going to be accelerated comput.

好的。那么请解释一下。首先，对于建设者来说，我们应该为机会而建设，对吧？我们是建设者。让我给你三个观点供你思考，这三个观点——希望——能让你对英伟达的未来感到更舒适。第一个观点，这是物理定律的观点。这是最重要的一点：通用计算已经结束了，未来是加速计算和AI计算。这是第一个观点。思考它的方式是：世界上有多少万亿美元的计算基础设施需要更新？对吧？对吧？当它更新时，它将是加速计算。

现有计算的加速迁移

That's right. And so the first thing you have to realize is that general purpose computing and nobody disputes that. Everybody goes, "Yeah, we completely agree with that. General purpose computing is over. Moore's law is dead." People say these things. And so what does that mean? So general purpose computing is going to go to accelerated computing. Our partnership with Intel is recognizing that general purpose computing needs to be fused with accelerated computing to create opportunities for them. Is that right? And so one, general purpose computing is shifting to accelerated computing and AI. Two, the first use case of AI is actually already everywhere, right? It's in search recommender engines, isn't that right? In shopping. The basic hyperscale computing infrastructure used to be CPUs doing recommenders, right? Is now going to GPUs doing AI, right? So you just take classical computing, it's going to accelerated computing AI. You take hyperscale computing is going from CPUs to accelerated computing and AI and then now that's the second point just feeding the metas the Googles the bite dances the Amazons and take their classical traditional way of doing hyperscaling and moving into AI that's hundreds of billions of dollars and and because that may be four billion people on the planet today if you take Tik Tok meta into account that's Google into account who are already demanding workloads that are driven by accelerated comput.

没错。所以你首先要意识到，通用计算，没有人对此有争议。每个人都说，“是的，我们完全同意。通用计算已经结束了。摩尔定律已经死了。”人们都这么说。那这意味着什么呢？通用计算将转向加速计算。我们与英特尔的合作认识到，通用计算需要与加速计算融合，为他们创造机会。对吧？所以，第一，通用计算正在转向加速计算和人工智能。第二，人工智能的第一个用例其实已经无处不在了，对吧？它在搜索推荐引擎中，不是吗？在购物中。过去，超大规模计算基础设施是CPU在做推荐，对吧？现在将转变为GPU在做AI，对吧？所以你把经典计算转为加速计算AI。你把超大规模计算从CPU转变为加速计算和AI，现在这是第二个要点，就是满足Meta、Google、字节跳动、亚马逊的需求，把他们传统地使用超大规模计算做推荐的方式转移到AI上，这涉及数千亿美元，而且因为今天地球上可能有40亿人（考虑到TikTok、Meta和Google），他们已经在要求由加速计算驱动的工作负载。

新兴应用的巨大机遇

That's exactly right. And so there a simp without even thinking about AI creating new opportunities. It's about AI shifting how you used to do something to the way new way of doing something. Okay. And then now let's talk about the future. I just so far I've only spoken kind of largely about just mundane stuff. Just mundane stuff. The old way is now wrong. You're going to go, you're no longer going to use uh uh fuel light lanterns. You're going to go to electricity. That's all. Right. Okay. And you no longer, you know, prop planes. You're going to go to jets. That's all. And so, you know, so far, you know, that's all I've talked about. And then now that the incredible thing is when you go to AI, when you go to accelerated computing, then what happens? What are the new applications that emerge as a result? And that's all the AI stuff that we're talking about. And that's the that opportunity.

完全正确。所以，在不考虑AI创造新机会的情况下，这是一种简单的方法。它是关于AI如何改变你过去做某事的方式，转变为新的做事方式。好的。然后现在我们来谈谈未来。到目前为止，我主要谈论的都是些琐碎的事情。琐碎的事情。旧的方式现在是错误的。你不会再使用燃油灯了。你会转向电力。就这样。对吧？好的。你不再使用螺旋桨飞机了。你会去乘坐喷气式飞机。就这样。所以，你知道，到目前为止，我谈论的都是这些。然后，令人难以置信的是，当涉及到AI，涉及到加速计算时，会发生什么？随之出现的新应用是什么？这就是我们谈论的所有AI内容。这就是那个机会。

AI超级计算机：增强人类智能

What is it? How do what does that look like? Well, the simple way of thinking about that is where motors replace labor and physical activity. We now have AI. These AI supercomputers, these AI factories that I talk about, they're going to generate tokens to augment human intelligence, right? And human intelligence represents what 55 65% of the world's GDP. Let's call it $50 trillion. And that $50 trillion is going to get augmented by something. And so let's you just let's come back to a single person. Suppose I were to hire a $100,000 employee. And I augmented that $100,000 employee with a $10,000 AI. Yes. And that $10,000 AI as a result made that $100,000 employee twice more productive, three times more productive. Would I do it? Heartbeat. I I'm doing it across every single person in our company right now. Right. Every single co-agents. That's right. Every That's right. Every single software engineer, every single chip designer in our company already has AIS working with them. 100% coverage. As a result, the number of chips we're building is better. The number is growing. The pace at which we're doing it is right. And so we're we're growing faster as a company. As a result, we're hiring more people. Our productivity is greater. Our top line's greater. Our profitability is greater. What's not to love about that?

它是什么？它看起来像什么？思考它的简单方法是：在马达取代劳动力和体力活动的地方，我们现在有了人工智能。我所说的这些AI超级计算机、这些AI工厂，它们将生成Token来增强人类智能，对吧？人类智能代表着全球GDP的55%到65%。我们称之为50万亿美元。这50万亿美元将得到某种程度的增强。所以我们回到一个人身上。假设我要雇佣一个年薪10万美元的员工。我用价值1万美元的AI来增强这个年薪10万美元的员工。是的。结果是，这个1万美元的AI使这个10万美元的员工的生产力提高了两倍、三倍。我会这样做吗？毫不犹豫。我现在正在我们公司的每一个人身上都这样做。对。每个联合智能体。没错。每个。没错。我们公司的每一位软件工程师、每一位芯片设计人员都已经有AI在与他们合作了。100%覆盖。因此，我们构建的芯片数量更多。数量在增长。我们完成的速度是对的。所以我们作为一家公司的增长速度更快。因此，我们雇佣了更多的人。我们的生产力更高。我们的收入更高。我们的盈利能力更强。这有什么不爱的呢？

亿美元级AI基础设施投资的逻辑

Now apply the Nvidia story to the world's GDP. Yeah. And so what's likely to happen is that that $50 trillion is augmented by let's pick a number 10 trillion that $10 trillion needs to run on a machine. M now the reason that AI is different than it in the past in a way software was written a priori and then it runs on a CPU and it doesn't it runs it a a person would operate it in the future of course AI is generating tokens but a machine has to generate the tokens and it's thinking so that software is running all the time whereas in the past the software was written once now the software is in fact writing all the time it's thinking in order for the AI to think it needs a factory. And so let's say that that 10 trillion of token generated 50% gross margins and 5 trillion of it needs a factory needs an AI infrastructure. So if you told me that on an annual basis the capex of the world was about $5 trillion I would say the math seems to make sense.

现在将英伟达的故事应用于全球GDP。是的。所以很可能会发生的是，那50万亿美元被我们选定的数字——10万亿美元——所增强，这10万亿美元需要在机器上运行。现在人工智能与过去不同之处在于，过去软件是先验编写的，然后在CPU上运行，它不会——一个人会在未来操作它，当然AI正在生成Token，但机器必须生成Token，而且它在思考，所以那个软件一直在运行，而在过去软件只写一次，现在软件实际上一直在编写，它在思考，为了让AI思考，它需要一个工厂。所以，假设那10万亿美元的Token生成带来了50%的毛利率，其中5万亿美元需要一个工厂，需要一个AI基础设施。所以如果你告诉我，全球每年的资本支出约为5万亿美元，我会说这个数学似乎说得通。

市场规模的扩大：从4000亿美元到5万亿美元的TAM

Yeah. And that's kind of the future, right? Yeah. the going from Excel general purpose computing to accelerated computing replacing all the hyperscales with AI and then now augmenting human intelligence for the world's GDP and today that market is about our estimate is about 400 billion annually. Yeah. So the the TAM you know is is a four to 5x increase over where it is today. Yeah. Eddie last night, Eddie Woo at Alibaba said between now and the end of the year and excuse me now and the end of the decade, they're going to increase their data center power by 10x. Right. Right. You just said how much? 4x. There you go. There you go. Yeah. They're going to increase power by 10x. And we we correlate to power. Nvidia's revenue is almost correlated to power. Isn't that right? Yeah, that's right. Yeah. Because one other thing, what he what else did he say? Yeah. He said token generation is doubling every few months. Yeah. What's that saying? The the perf per watt has to keep on going exponentially. That's why Nvidia's like cranking it out with perf per watt. And revenue per watt is, you know, watt is basically revenues in this future.

是的。这就是未来，对吧？从Excel通用计算到加速计算，用AI取代所有超大规模运营商，然后现在增强全球GDP的人类智能。今天，这个市场我们估计每年约4000亿美元。是的。所以总潜在市场（TAM）比现在增长了四到五倍。是的。昨天晚上，阿里巴巴的Eddie Woo说，从现在到今年年底——抱歉，从现在到十年底，他们的数据中心电力将增加10倍。对吧？对吧？你刚才说了多少？4倍。就是这样。就是这样。是的。他们将电力增加10倍。我们的收入几乎与电力相关。不是吗？是的，没错。是的。因为他说的另一件事是什么？他说Token生成每隔几个月就会翻一番。是的。那句名言是什么？每瓦性能（perf per watt）必须持续指数级增长。这就是为什么英伟达正在通过每瓦性能不断突破。而每瓦收入……你知道，在这个未来，瓦特基本上就代表收入了。

GDP增长的加速与分析师预期的偏差

Embedded in this assumption, I find it very fascinating historical context, right? For 2,000 years, basically GDP did not grow. Okay? And then we get the industrial revolution, GDP accelerates. We get the digital revolution, GDP accelerates. And it basically what you're saying, and Scott Besson has said it, he said, I think we're going to have 4% GDP growth next year. Basically, what you're saying is the world's GDP growth is going to accelerate because now we are giving the world billions of co-workers that will do work for us. And if GDP is an amount of output for a fixed amount of labor and capital, right, it has to accelerate. It has to, right? It has to look at what's going on with AI as a result of the technology of AI. And that technology of AI, let's just call it the large language models and all the AI agents. It's now creating a new industry of AI agents. There's no question about that. Okay. So, so that's OpenAI is the fastest growing revenue company in history, right? and they're growing exponentially, right? And so, so AI itself is a fast growing industry because of AI needs a factory behind it, right? An infrastructure behind it. There's this industry is growing. My industry is growing. And because my industry is growing, the industry underneath it is growing. Energy is growing. Power shell. This is the This is like renaissance for the energy industry, isn't that right? Nuclear energy, you know, gas turbines. I mean, look at all of those companies in the in the infrastructure ecosystem underneath us. They're doing incredibly well. Everybody's growing.

在这个假设中嵌入了一个我发现非常迷人的历史背景，对吧？两千年来，GDP基本上没有增长。好的？然后我们迎来了工业革命，GDP加速增长。我们迎来了数字革命，GDP加速增长。基本上你的意思是，斯科特·贝森也说过，他说，我认为我们明年GDP增长将达到4%。你的基本意思是，世界GDP增长将加速，因为我们现在正在给世界提供数十亿的同事来为我们工作。如果GDP是固定劳动力和资本的产出量，对吧，它就必须加速。它必须加速，对吧？看看AI技术带来的变化。这种AI技术，我们姑且称之为大型语言模型和所有AI智能体，它正在创造一个AI智能体的新行业。毫无疑问。好的。所以，OpenAI是历史上收入增长最快的公司，对吧？而且他们的增长是指数级的，对吧？所以AI本身就是一个快速增长的行业，因为AI需要背后的工厂，对吧？需要背后的基础设施。这个行业在增长。我的行业在增长。由于我的行业在增长，它下面的行业也在增长。能源在增长。电力。这对能源行业来说就像文艺复兴，不是吗？核能，你知道，燃气轮机。我的意思是，看看我们下面基础设施生态系统中的所有这些公司。它们都做得非常好。每个人都在增长。

关于“泡沫”与“循环收入”的担忧

These numbers have everybody talking about a glutter bubble, right? Zuckerberg said last week on a podcast, you know, he said, "Listen, I think it's quite possible at some point that we will have an air pocket and Meta may in fact overspend by $10 billion or whatever, but he said it doesn't matter. It's so existential to the future of his business that it's a risk that they have to take. But when you think about that, it sounds a little bit like prisoners dilemma. Right. And walk us again through these are very happy prisoners. Walk us again through Right. Today our estimate is that we're going to have a 100 billion of AI revenue in 2026 excluding meta and excluding you know the GPUs running recommender engines. Okay. So there's or search or correct. So there there's other stuff but let's call it 100 billion. What is that industry anyways? What is the industry already in hypers scale? What is the hypers scales you know between trillions?

这些数字让每个人都在谈论供应过剩或泡沫，对吧？扎克伯格上周在一次播客中说，你知道，他说：“听着，我认为很有可能在某个时候我们会遇到一个‘气泡’，Meta 实际上可能会超支100亿美元或多少，但他又说这没关系。这对他的业务未来太重要了，是他们必须承担的风险。”但当你想到这一点时，听起来有点像囚徒困境。对吧？然后我们再来回顾一下，这些都是非常快乐的囚徒。我们再回顾一下。对。今天我们的估计是，到2026年，我们的AI收入将达到1000亿美元，不包括Meta，也不包括你所说的运行推荐引擎的GPU。好的。所以还有搜索或者……对。所以还有其他的东西，但我们姑且称之为1000亿美元。这个行业到底是什么？超大规模行业本身已经有多少收入了？超大规模运营商的收入在万亿之间？

AI收入的内生性增长

Yeah exactly by the way that industry is going to AI before anybody starts at zero. You got to start there. But I think the skeptics would say we need to go from a 100 red billion of AI revenue in '26 to at least a trillion of AI revenue in 2030. Okay. You just were talking a minute ago about five trillion when you look at kind of global GDP. If you do did a bottoms up, can you see your way to a trillion dollars of AI driven revenues from a hundred billion over the course of the next 5 years? Are we growing that fast? Yes. And I would also say we're already there.

顺便说一句，这个行业在任何人从零开始之前就已经转向AI了。你必须从那里开始。但我认为怀疑论者会说，我们需要从26年的1000亿美元AI收入增长到2030年至少一万亿美元的AI收入。好的。你刚才还谈到了5万亿美元，当你从全球GDP的角度来看。如果你做自下而上的估算，你能在未来5年内看到从1000亿美元增长到一万亿美元的AI驱动收入吗？我们增长那么快吗？是的。而且我还会说我们已经到了那里。

迭代与增量增长

Okay, let's just talk about incremental from where we are. Now we can talk about incremental from where we are today, right? As you do your bottoms up or your tops down, I just heard your tops down about percentage of global GDP. Yeah. What is the percentage probability that you think we'll have a glut will run into a glut in the next three or four or five years? Right. It's a distribution of we don't know the future. It's a distribution of power until until we fully convert all general purpose computing to accelerated computing and AI. Until we do that, yes, I think the chances are extremely low. Okay. Okay. And that will take a few years. That'll take a few years. Yeah. Yeah. Let me ask one more and then until all recommender engines are AI based. until all content generation is AI based because content generation consumer oriented content generation is very largely recommender systems and so on so forth um and all of that's going to be AI generated until until all of the stuff what classically was hypers scale now transitions to AI you know everything from shopping to e-commerce to you know all that stuff until everything goes over because but all this new build right when we're talking about trillions. We're investing ahead of where we are.

好的，我们只谈论从我们目前的水平开始的增量。现在我们可以谈论从我们今天的水平开始的增量，对吧？当你进行自下而上或自上而下的分析时，我刚才听到了你关于占全球GDP百分比的自上而下的分析。是的。你认为在未来三、四年或五年内，我们遇到供应过剩的概率是多少？对。这是一个概率分布，我们不知道未来。这是一个权力分布，直到我们完全把所有的通用计算转换为加速计算和AI。在我们完成之前，是的，我认为机会非常低。好的。好的。这需要几年时间。这需要几年时间。是的。是的。再问一个，直到所有的推荐引擎都基于AI。直到所有的内容生成都是基于AI的，因为面向消费者的内容生成在很大程度上是推荐系统等等，所有这些都将是AI生成的，直到所有经典意义上的超大规模应用都过渡到AI，你知道，从购物到电子商务，再到所有这些东西，直到一切都过去……因为所有这些新建设，当我们谈论万亿的时候，我们是在我们现有水平之上进行投资。

投资的驱动力：需求而非财务工程

Um, you know, is that like will? Are you obliged to invest the money even if you see a slowdown or a kind of a glut coming or is this one of these things that you're just waving the flag to the ecosystem to say get out and build and at some point in time if we see some of this slow down, we can always pull back on the level of investment. Actually, it's the other way because we're at the end of the supply chain, right? And so, we respond to demand. Okay? And right now all the VCs will tell you and you guys know the demand the short there's a shortage of compute in the world not because there's a shortage of GPUs in the world. Okay, if they give me an order I'll build it. Mhm. Right. We've over the last couple years we've really plumbed the supply chain. So all of the supply chain behind me from wafer starts to co-ass HBM memories you know all of that technology. We've really geared up. Yeah. If we need to double we'll double. Yes. Okay. So the supply chain is ready. Now we're just waiting for demand signals and when when uh the the CSPs and the hyperscalers and our customers uh do their annual plan and they give us you know their forecast um we respond to that and we build to that.

嗯，你知道，这是“will”吗？即使你看到放缓或某种供应过剩，你也有义务投资这些钱吗？还是说这只是你向生态系统挥舞旗帜，告诉他们出去建设，总有一天如果我们看到放缓，我们可以随时减少投资水平？实际上是反过来的，因为我们在供应链的末端，对吧？所以，我们对需求做出反应。好的？现在所有的风险投资公司都会告诉你，而且你们知道需求——世界上计算资源短缺，不是因为世界上GPU短缺。好的，如果他们给我一个订单，我就会去建。嗯。对。在过去的几年里，我们真正深入研究了供应链。所以我后面所有的供应链，从晶圆启动到CoWoS HBM内存，你知道所有这些技术。我们真的做好了准备。是的。如果我们必须翻倍，我们就会翻倍。是的。好的。所以供应链已经准备好了。现在我们只是在等待需求信号，当……当CSP和超大规模运营商以及我们的客户做他们年度计划并给我们他们的预测时，我们做出回应并据此建设。

客户预测的持续低估

Now what's what's going on of course is that every one of their forecasts that they provide us turns out to have been wrong right because they under forecasted and so now we're always in a scramble mode. Mhm. And so we've been in the scramble mode now for you know a couple of years and it's whatever forecast we've been given has been always significant increase from last year but not not enough. Satcha last year seemed to be pulling back a little bit. You know seemed to be you know some people called him the adult in the room tamping down kind of some of these these expectations. A few weeks ago he said hey I've also built two gigs this year and we're going to accelerate in the future. Do you see some of the traditional hyperscalers that may have been moving a little slower than let's call it a Coreweave or or or Elon X or maybe a little slower than Stargate? Do you see them all? It it sounds like to me they're all leaning in more now and they're all also because of the second exponential.

当然，现在发生的情况是，他们向我们提供的每一个预测结果都被证明是错误的，对吧？因为他们低估了需求。所以我们现在总是在争分夺秒的状态下。嗯。所以我们已经处于这种争分夺秒的状态下好几年了，无论我们得到的预测是什么，都比去年有显著增加，但仍然不够。萨蒂亚去年似乎有点退缩了。你知道，似乎……有些人称他为“房间里的成年人”，在抑制这些预期。几周前他说，嘿，我今年也建了两个千兆瓦的设施，而且我们将来会加速。你是否看到一些传统超大规模运营商的动作可能比Coreweave、Elon X，或者Stargate稍微慢一些？你是否看到他们都在？听起来他们现在都在更加积极地投入，而且他们也都在，因为第二个指数增长。

第二个指数增长：推理取代一次性决策

Okay. We've already had one exponential we were experiencing which was the adoption rate of AI, the engagement of AI was growing exponentially. Yes. The second exponential that kicked in was reasoning. Yeah. That was the conversation we had one year ago. One year ago. Yeah. We said, "Hey, listen. The moment you take AI from one shot, memorizing an answer and Right. Memorizing and generalizing, that's basically pre-training." Yeah. So memorizing an answer, you know, what's 8* 8? Just memorize it. Okay. And so memorizing an answer and generalizing, that was one shot AI. Now a year ago, reasoning came about for sure research came about, tool use came about, and now you're a thinking AI 1 billion X. It's going to use a lot more compute.

好的。我们已经经历了一个指数增长，那就是AI的采用率和参与度呈指数级增长。是的。开始的第二个指数增长是推理。是的。这就是我们一年前的对话。一年前。是的。我们说：“嘿，听着。当你把AI从一次性、记住答案，以及记住和泛化（这基本上就是预训练）中解放出来时。”是的。所以记住一个答案，你知道，8乘以8是多少？记住就行了。好的。所以记住一个答案和泛化，这就是一次性AI。现在一年前，推理确实出现了，研究出现了，工具使用出现了，现在你是一个思考型AI，增长10亿倍。它将使用更多的计算资源。

传统超大规模运营商的普遍低估

Certain hyperscale customers to your point had internal workloads that they had to migrate anyways from um from general purpose computing to accelerated computing. So they built through the cycle. I think maybe some hyperscalers had different workloads so they weren't quite sure how quickly they could digest it but everyone has now concluded that they dramatically underbuilt.

正如你所指出的，一些超大规模客户本就必须将内部工作负载从通用计算迁移到加速计算。所以他们度过了这个周期。我想也许一些超大规模运营商有不同的工作负载，所以他们不确定他们能消化多快，但现在所有人都得出了结论：他们严重低估了需求。

数据处理市场的巨大机遇

One of the applications that my favorite is just good oldfashioned data processing structured data and unstructured data. Just good oldfashioned data processing. And very soon we're going to announce a very big initiative of accelerated data processing. M data processing represents the vast majority of the world's CPUs today. It still completely runs on CPUs. You know, if you go to data bricks, it's mostly CPUs. You go to snowflakes, mostly CPUs. Uh SQL processing at Oracle, mostly CPUs. Everybody's using CPUs to do SQL structured data. In the future, that's all going to move to AI data. That is one gigantic massive market that we're going to move to. But you need you need a you need everything that Nvidia does requires acceleration layers and requires you know domain specific right data processing lies recipes we got to go build that but that's coming.

我最喜欢的一个应用就是老式的CPU驱动的数据处理，包括结构化数据和非结构化数据。老式的数据处理。很快我们将宣布一项非常大的加速数据处理计划。数据处理代表了当今世界上绝大多数CPU的使用场景。它仍然完全在CPU上运行。你知道，如果你去Databricks，大部分是CPU。你去Snowflake，大部分是CPU。甲骨文的SQL处理，大部分是CPU。每个人都在使用CPU来处理结构化SQL数据。未来，所有这些都将转向AI数据处理。这是一个巨大的、庞大的市场，我们将向其迁移。但你需要，英伟达所做的一切都需要加速层，需要特定领域的，对吧，数据处理的“食谱”，我们必须去构建它，但它正在到来。

循环收入指控的回击：系统成本的考量

So one of the push backs I you know I turned on CNBC yesterday they were like oh glut bubble when I turned on Bloomberg it was about roundtipping and circular revenues okay and so for the benefit of people you know at home know these arrangements are when companies enter into a misleading transaction that artificially inflates revenue without any underlying economic substance. So in other words, growth propped up by financial engineering, not by customer demand. And the canonical case everybody's referencing of course is Cisco and Nortell from the last bubble 25 years ago. So when you guys or Microsoft or Amazon are investing in companies that are also your big customers, in this case you guys investing in Open AI while Open AI is buying tens of billions of chips, just remind us and remind everybody else like what is it what are the analysts on Bloomberg and otherwise getting wrong when they're hyperventilating about circular revenues or about roundtipping?

所以，我昨天打开CNBC时，他们谈论供应过剩的泡沫。当我打开彭博社时，他们谈论的是回购和循环收入，好吧？所以为了让家里的朋友们了解，这些安排是指公司进行误导性交易，在没有任何潜在经济实质的情况下人为地夸大收入。换句话说，增长是由财务工程支撑的，而不是由客户需求支撑的。每个人都在引用的典型案例当然是25年前的思科和北电网络。所以当你们、微软或亚马逊投资那些同时也是你们大客户的公司时，比如你们投资OpenAI，而OpenAI正在购买数百亿美元的芯片时，请提醒我们和提醒其他所有人，彭博社和其他分析师在对循环收入或回购过度反应时，究竟错在哪里？

投资与业务的解耦

10 gigawatts is like $400 billion, right? Something like that. and and that $400 billion dollars will have to be largely funded by their offtake, right? Their revenue which is growing exponentially. It has to be funded by their capital, the money they've raised through equity and whatever debt they can raise. Those are the three vehicles. And the equity that they could raise and the debt that they could raise has something to do with the confidence of the revenues that they could sustain for sure. And so smart investors and smart lenders will consider all of these factors. Fundamentally, that's what they're going to do. That's their company. It's not my business. And of course, we have to stay very close to them to make sure that we build in support of their continued growth. Okay? And so, um, there's the revenue side of it and has nothing to do with the investment side of it. The investment side of it is not tied to anything. It's an opportunity to invest in them. And as we were mentioning earlier, this is likely going to be the next multi-trillion dollar hyperscale company. And who doesn't want to be an investor in that? You know, my only regret is that that they invited us to invest early on. I remember those conversations and we were so poor, you know, that we were so poor, we didn't invest enough, you know, and I should have given them all my money.

10吉瓦大约是4000亿美元，对吧？大概是这个数。而这4000亿美元将主要由他们的承购量资助，对吧？他们呈指数级增长的收入。它必须由他们的资本资助，他们通过股权和任何能筹集的债务筹集的资金。这三种方式。他们可以筹集的股权和债务，与他们能够维持的收入信心有关。所以精明的投资者和精明的贷款人会考虑所有这些因素。从根本上说，他们会这么做。那是他们的公司。不是我的事。当然，我们必须与他们保持非常密切的联系，以确保我们支持他们持续增长。好的？所以，嗯，这是收入方面，与投资方面无关。投资方面与任何事情都没有关系。这是一个投资他们的机会。正如我们之前提到的，这很可能成为下一家万亿美元级的超大规模公司。谁不想成为它的投资者呢？你知道，我唯一的遗憾是他们邀请我们早期投资。我记得那些谈话，我们当时太穷了，你知道，我们太穷了，投资不够，你知道，我真应该把所有的钱都给他们。

对英伟达芯片的依赖与投资的务实性

And the reality is if you guys don't do your jobs and keep up with, you know, if if Ver Rubin doesn't turn into a good chip, they can go get other chips and put them in these data centers, right? There's no obligation that they have to use your chips and and like you said, you're looking at this as an opportunistic equity investment. The other thing I would say and we've made some great investments. I got to put it out there, you know, we invested in XAI, we invested in Corewave. Incredible. Yeah. Yeah. How smart was that? Yeah. As I go back to this, the other fundamental thing it seems to me is, you know, you're putting it out there. You're saying this is what we're doing. And the underlying economic substance here, right? It's not that you're just some somehow sending revenues back and forth between the two companies. We got people sending money every month for Chat GPT, a billion and a half monthly users using the product. You just said every enterprise in the world is either going to do this or they will die. Every sovereign views this as existential to their national security and economic security as as nuclear power. What person, company or nation says intelligence is bas basically optional for yeah for us. I mean it's fundamental to them.

现实是，如果你们没有做好工作跟上进度，你知道，如果Ver Rubin没有变成一个好芯片，他们可以去拿别的芯片放进这些数据中心，对吧？他们没有义务必须使用你们的芯片，而且就像你说的，你把这看作是一项机会主义的股权投资。我想说的另一件事是，我们做了一些很棒的投资。我得说出来，你知道，我们投资了XAI，我们投资了Corewave。太棒了。是的。那多明智啊？是的。当我回到这里时，另一个基本点似乎是，你知道，你把它说出来。你说这就是我们正在做的事情。这里的潜在经济实质是什么，对吧？不是说你们只是在两家公司之间来回传送收入。我们有用户每月为ChatGPT支付费用，每月有十亿半的用户在使用该产品。你刚才说世界上每一个企业都要么这样做，要么就会消亡。每个主权国家都认为这对他们的国家安全和经济安全至关重要，就像核能一样重要。什么人、公司或国家会说，智能对我们来说基本上是可有可无的？我的意思是，这对他们来说是根本性的。

固态硬盘（SSD）与AI：每年发布周期的驱动力

Well the automation of intelligence I beat the demand question to death. So let's jump in a little bit to system design and I'm I'm I'm going to turn to Clark here in a sec second on that. But in 2024, you switched to your annual release cycle, right with Hopper. You then had a massive upgrade which required, you know, significant data center overhaul with Grace Blackwell in 2025, and in the back half of '26, we're going to get Vera Rubin. '27 we'll get Ultra and '28 Fineman. How is the annual release cycle going? Okay. What were the main goals of going to an annual release cycle? And did AI inside Nvidia allow you to execute the annual release cycle?

人工智能的自动化……我对需求问题已经讨论得够多了。所以我们来谈谈系统设计，我马上要请Clark来谈谈。但在2024年，你转向了年度发布周期，对吧，随着Hopper的推出。然后你有了巨大的升级，需要进行重大的数据中心改造，Grace Blackwell在2025年，26年下半年我们将迎来Vera Rubin，27年是Ultra，28年是Fineman。年度发布周期进展如何？好的。转向年度发布周期的主要目标是什么？英伟达内部的AI是否让你能够执行年度发布周期？

AI对加速和规模的决定性作用

Yeah, the answer is yes. On the back on the last question, without it, um, Nvidia's velocity, our pace, our scale would be limited. And so without without AI these days, um, it's just simply not possible to build what we built. Now, um, why do we do it? There's something that remember uh uh Eddie said it uh at his earnings call um or his conference. Uh Satia has said it, Sam has said it. The token generation rate is going up exponentially. Yeah. And the customer use is going up exponentially. Um I think they're at 800 million weekly active users or something like that. Yes. I mean that's less than two years from chat GPT, right? And each of those users is generating massively more tokens because they're using inference time reasoning. That's right. Exactly. And so so the first thing is because the token generation rate is going up so incredibly two exponentials. Yeah. On top of each other, we have to un unless we increase the performance at incredible rates, the cost of token generation will keep growing because Mo's law is dead, right? Because transistors basically cost the same every single year now. And power is largely the same. And between those two fundamental laws, unless we come up with new technologies to drive the cost down, even if there's a slight difference in growth, you give somebody a discount of a few percent. How's that going to make up for two exponentials?

是的，答案是肯定的。关于最后一个问题，如果没有AI，英伟达的速度、我们的步伐、我们的规模都会受到限制。所以如果没有AI，这些天，要构建我们所构建的东西，简直是不可能的。那么，我们为什么要这样做呢？还记得Eddie在财报电话会议上说的……萨蒂亚说过，萨姆也说过。Token生成速度呈指数级增长。是的。客户使用量也在呈指数级增长。我认为他们的周活跃用户数达到了8亿左右。是的。我的意思是，这距离ChatGPT还不到两年，对吧？而且这些用户中的每一个都在生成更多的Token，因为他们在使用推理时间的推理能力。没错。完全正确。所以，第一点是，因为Token生成速度如此惊人地增长，两个指数。是的。相互叠加，除非我们以惊人的速度提高性能，否则Token生成的成本将持续增长，因为摩尔定律已经死了，对吧？因为晶体管在如今每年成本基本相同。电力也基本相同。在这两条基本定律之间，除非我们拿出新技术来降低成本，即使增长率有微小差异，你给别人打几折，这怎么能弥补两个指数的差距呢？

极致的系统级代码设计

And so we have to increase our per performance annually at a pace that keeps up with that exponential. Yeah. So in the case of in the case of um uh going from uh uh Kepler to yeah all the way to Kepler all the way to um uh Hopper was probably 100,000x that was the beginning of the AI journey for Nvidia. 100,000x in 10 years. Okay. Between Hopper and Blackwell, we increased because of MVLink 72, right? 30x in one year and then we'll get another X factor again with Reuben and then we'll get another X factor with Fineman. And the way we we do that is because the transistors aren't really helping us very much, right? Moors law is largely the density is growing up but going up but the performance is not. And so if that's the case, one of the challenges that we have to do is we have to break the entire problem down at the system level and change every chip at the same time and all the software stack and all the systems all at the same time. The ultimate extreme code design. Nobody's ever codees at this level before, right? We change the CPU, revolutionize the CPU, a GPU, the networking chip, the MVLink scale up, the Spectrum X scale out. Somebody said I heard somebody said, "Oh yeah, it's just Ethernet." Yeah. Right. Okay. So, Spectrum X Ethernet is not just Ethernet. And people are starting to discover, oh my god, the X factors is pretty incredible, right? You know, Nvidia's Ethernet business, the just Ethernet business is the fastest growing Ethernet business in the world. Yeah. And so, so scale out and of course now we have to uh build even larger systems. So, we scale across um multiple AI factories connected together. And then we do this at an annual pace. And so we now have an exponential of exponentials going ourselves from technology. And that allows our customers to drive the cost of tokens down. Keep making those tokens smarter and smarter with pre-training and post-training and thinking. And as a result, when the AI gets smarter, they get more used. When they get more used, they're going to grow exponentially.

所以我们必须每年以跟上这种指数级增长的速度来提高我们的性能。是的。所以从Kepler到Hopper，我们增加了大约10万倍，这是英伟达AI之旅的开始。十年10万倍。好的。在Hopper和Blackwell之间，由于NVLink 72，我们在一年内实现了30倍的提升，然后随着Rubin我们会得到另一个X因子，然后随着Fineman我们会得到另一个X因子。我们之所以能做到这一点，是因为晶体管并没有真正帮我们太多，对吧？摩尔定律在很大程度上是密度在增加，但性能没有。所以，如果是这样的话，我们必须做的一个挑战是，我们必须在系统层面分解整个问题，同时改变每一个芯片、所有软件栈和所有系统。极致的协同设计（code design）。以前没有人在这个层面上进行协同设计，对吧？我们改变了CPU，革命性地改变了CPU、GPU、网络芯片、NVLink的扩展（scale up），Spectrum X的扩展（scale out）。有人说，我听到有人说，“哦，这只是以太网。”是的。对吧？好的。所以，Spectrum X以太网不仅仅是以太网。人们开始发现，我的天哪，这个X因子非常惊人，对吧？你知道，英伟达的以太网业务，纯粹的以太网业务是世界上增长最快的以太网业务。是的。所以，扩展出去，当然现在我们必须构建更大的系统。所以我们在连接起来的多个AI工厂之间进行扩展。然后我们每年都这样做。所以我们自己从技术上获得了指数的指数增长。这使得我们的客户能够降低Token的成本，通过预训练、后训练和思考，不断让Token变得更智能。结果是，当AI变得更聪明时，它们的使用率就会增加。当它们使用率增加时，它们就会呈指数级增长。

什么是极端协同设计（Extreme Code Design）？

For people who may not be as familiar. Yeah. What is extreme code design? Extreme code design means that you have to optimize the model, algorithm, system, and chip at the same time. You have to innovate outside the box, right? Because Moore's law said you just have to keep making the CPU faster and faster. Everything got faster. You were innovating within a box. Just make that chip faster. Yeah. Well, if that chip doesn't go any faster, then what are you going to do? Innovate outside the box. And so Nvidia really changed things because we did two things. We invented CUDA, invented GPUs, and we invented the idea of code design at a very large scale. That's why there's all these industries we're in. We're creating all these libraries and code design. Number one, full stack extreme is even beyond software and GPUs. It's now at the data center level switches and networking and you know all of that all of that software in the switches and the networking and the nicks the scale up the scale out optimizing across all of that as a result of that blackwell to hopper is 30x no moors law could possibly achieve that right and so that's extreme and that comes from the extreme code design that's because Nvidia has that's why we got into networking and switching and scale up and scale out and scale across and building CPUs and building GPUs and building nicks. You know that that's the reason why Nvidia is so rich in software and people we we check in more open-source software in the world than just about anybody except one other company. I think it's AI2 or something like that. And so so we have such enormous richness of software and that's just in AI. Don't forget computer graphics and digital biology and autonomous vehicles and you know the amount of software we produce as a company is incredible that allows us to do deep and extreme code designs.

对于那些可能不太熟悉的人来说。是的。什么是极端协同设计？极端协同设计意味着你必须同时优化模型、算法、系统和芯片。你必须进行箱外创新，对吧？因为摩尔定律说你只需要不断地让CPU变得更快。一切都变快了。你是在一个盒子里进行创新。只是让那个芯片更快。是的。好吧，如果那个芯片不再变快，那你会怎么做？在盒子外面进行创新。所以英伟达确实改变了局面，因为我们做了两件事：我们发明了CUDA，发明了GPU，我们发明了大规模协同设计的理念。这就是为什么我们涉足了所有这些行业。我们正在创建所有这些库和协同设计。第一，全栈的极端性甚至超越了软件和GPU。它现在已经到了数据中心级别的交换机和网络，你知道所有这些……交换机、网络和网卡中的所有软件，以及扩展（scale up）、扩展（scale out），对所有这些进行优化。因此，Blackwell到Hopper的提升是30倍，没有摩尔定律可以做到这一点，对吧？所以这就是极端，它来自于极端的协同设计。这是因为英伟达拥有……这就是为什么我们涉足网络和交换，以及扩展和跨域扩展，以及构建CPU和构建GPU以及构建网卡。你知道，这就是为什么英伟达在软件和人才方面如此丰富。我们向开源软件的贡献量仅次于几乎所有其他公司，除了另一家公司。我认为是AI2或类似的公司。所以我们在软件方面拥有如此巨大的丰富性，而且这仅限于AI。别忘了计算机图形学、数字生物学和自动驾驶汽车，你知道我们公司产生的软件量是惊人的，这使我们能够进行深入和极端的协同设计。

年度发布周期对竞争格局的影响

I heard from one of your competitors you know yes he's doing this because it helps drive down the cost of token generation but at the same time your annual release cycle makes it almost impossible for your competitors to keep up. um the supply chain gets locked up more because you're giving three-year visibility to your supply chain. So now the supply chain has confidence as to what they can build to. So do you think about this? Wait wait wait before you before you ask the question. Think about this. In order for us to do several hundred billion dollars a year of AI infrastructure buildout. Yes. Think about how much capacity we had to go start a year ago. Yes. We're talking about building hundreds of billions of dollars of wafer starts and DRAM buys and are you you guys talking Yeah. This is now at a scale that hardly any company can keep up with.

我从你的一位竞争对手那里听说，你知道，是的，他这样做是因为这有助于降低Token生成的成本，但同时你的年度发布周期使得你的竞争对手几乎不可能跟上。嗯，供应链被锁定了更多，因为你给你的供应链提供了三年的可见性。所以现在供应链对他们可以构建什么有了信心。你对此怎么看？等等，等等，在你问问题之前。想想这个。为了让我们每年完成数千亿美元的AI基础设施建设。是的。想想我们一年前不得不启动了多少产能。是的。我们谈论的是建造价值数千亿美元的晶圆启动和DRAM采购，你们在谈论吗？是的。这现在的规模几乎没有哪家公司能跟上。

英伟达竞争优势的强化

So would you say your competitive moat is greater today than it was three years ago? Yeah. You know, first of all, there's just more competition than ever before, but it's harder than ever before. Mhm. And the reason why I say that is because um wafer cost is getting wafer costs are getting higher which means that unless you do code design at an extreme scale you're just not going to be able to deliver the X factor growth number one number and so you you know unless you unless you're working on six seven eight chips a year right that's amazing thing it's not about building an ASIC it's about building an AI factory system and this system that has a lot of chips in and they're all co-designed and together they deliver that you know that 10x factor that we get uh almost regularly. Okay. So number one the code design is extreme. The second thing is that the scale is extreme. When your customers deploy a gigawatt that's a you know 400,000 500,000 GPUs right getting 500,000 GPUs to work together is a miracle. I mean it's just a miracle. And so they your customers are taking enormous risk on you to go buy all of this. You got to ask yourself what customer would place a $50 billion PO on an architecture, right? On an unproven architecture, a new one, right? A new chip. You just take out a whole new chip. You're as excited as you are about it, you know, and everybody's excited for you and and you just show the first silicon, right? Who's going to give you $50 billion PO, right? And why would you start $50 billion worth of wafers for a chip that just taped out? But for Nvidia, we could do that because our architecture is so proven. So number the the scale of our customer is so incredible. Now the scale of our supply chain is incredible, right? Who's going to start all of that stuff, pre-build all of that stuff for a company unless they know that Nvidia can deliver through? Isn't that right? and they believe that we can we can deliver through to all the customers around the world. They're willing to start several hundred billion dollars at a time. This is just the scale is incredible.

所以你会说，你今天的竞争护城河比三年前更深了吗？是的。你知道，首先，竞争比以往任何时候都多，但难度也比以往任何时候都大。嗯。我说这句话的原因是，晶圆成本越来越高，这意味着除非你进行极端的协同设计，否则你根本无法实现X倍的增长，这是第一点。所以，你知道，除非你每年都在研究六七八个芯片，对吧，这是一件了不起的事情，这不是关于构建一个ASIC，而是关于构建一个AI工厂系统，这个系统包含很多芯片，它们都是协同设计的，共同实现了我们几乎定期获得的10倍提升。好的。所以第一点，协同设计是极端的。第二点是规模是极端的。当你的客户部署一个吉瓦时，那就是40万到50万个GPU，对吧？让50万个GPU协同工作是一个奇迹。我的意思是，这简直是个奇迹。所以他们——你的客户为了购买所有这些东西，对你承担了巨大的风险。你得问问自己，谁会对一个架构下500亿美元的采购订单，对吧？对一个未经证明的架构，一个新的架构，对吧？一个新的芯片。你刚推出一个全新的芯片，你对它感到非常兴奋，你知道，每个人都为你感到兴奋，然后你只展示了第一个硅片，对吧？谁会给你500亿美元的采购订单，对吧？你为什么要为一个刚刚定型的芯片启动价值500亿美元的晶圆？但是对于英伟达来说，我们可以做到，因为我们的架构经过了充分的证明。所以第一点，我们客户的规模是如此不可思议。现在我们供应链的规模也是不可思议的，对吧？谁会为一个公司启动、预先构建所有这些东西，除非他们知道英伟达可以交付？不是吗？他们相信我们可以向全球所有客户交付。他们愿意一次性启动数千亿美元的订单。规模是惊人的。

GPU与ASIC之争的本质：系统与可编程性

To that point, you know, one of the biggest key debates and controversies in the world is this question of GPUs versus A6s, Google's TPUs, Amazon's Tranium, and it seems like everyone from ARM to OpenAI to Enthropic are, you know, rumored to be building one. Mhm. Last year you said, you know, we're building systems, not chips, and you're driving performance through every single part of that stack. You also said that many of these projects may never get to production scale. But given like the most of them, most of them, given the seeming success of Google's TPUs, yeah, you know, how are you thinking about this evolving landscape today?

说到这一点，你知道，世界上最主要的争论和争议之一就是GPU与ASIC的较量，谷歌的TPU，亚马逊的Tranium，似乎从ARM到OpenAI再到Anthropic都在传闻要构建自己的芯片。嗯。去年你说，你知道，我们构建的是系统，而不是芯片，你通过这个堆栈的每一个部分来推动性能。你也说过，这些项目中的许多可能永远无法达到生产规模。但是考虑到它们中的大多数，考虑到谷歌TPU的表面成功，是的，你知道，你今天如何看待这种不断发展的格局？

ASIC的局限性与英伟达的平台优势

Yeah. And first of all, Mhm. the the advantage that that Google had is foresight. Remember, they started TPU1 before everything started. You know, this is no different than a startup. You're supposed to build a startup. You're supposed to create a startup before the market grows. Mhm. You're not supposed to come up as a startup when the market's a trillion dollars large. Mhm. You know the this fallacy and and all VCs know this this fallacy that a large market if you could just take a few percent market share you could be a giant company. That's actually fundamentally wrong. Yeah. You're supposed to take 100% of a tiny company, a tiny industry, which is what Nvidia did did, right? Which is what TPUs did. There were only the two of us. But you better hope that that industry gets really big. You're creating an industry.

是的。首先，嗯。谷歌拥有的优势是远见。还记得吗，他们在一切开始之前就开始研发TPU1了。你知道，这与初创公司没什么不同。你应该在市场增长之前就建立一家初创公司，你应该在市场增长之前就创造一家初创公司。你不应该在市场已经达到万亿美元规模时才作为初创公司出现。你知道这个谬论，而且所有风险投资人都知道这个谬论，即在一个大市场中，如果你能获得几分的市场份额，你就可以成为一家大公司。这实际上从根本上就是错误的。是的。你应该占据一个微小公司、一个微小行业的100%份额，这就是英伟达所做的，对吧？这就是TPU所做的。我们俩是仅有的两家。但你最好希望那个行业变得非常庞大。你在创造一个行业。

ASIC的演变与英伟达的整体视图

That's right. Right. and and I mean the Nvidia story you know which and so that's the challenge for the for people who are building A6s now it looks like a juicy market but remember this juicy market has evolved from a chip called a GPU to I just described an AI factory and you guys just saw I just announced a chip called CPX for context processing and diffusion video generation a very specialized workload but an important workload inside a data center I just prelude it to maybe AI data processing processors because guess what? You need long-term memory. You need short-term memory. The KV cache processing is really intense. AI memory is a big deal. You know, you kind of like your AI to have good memory and just dealing with all the KV caching around the system. Really complicated stuff. Maybe it wants to have a specialized processor. Maybe there's other things, right? So you you see that Nvidia's our our our viewpoint is now not GPU. Our viewpoint is looking at the entire AI infrastructure and what does it take for these incredible companies to get all of their workload through it which is diverse and changing. Look at the transformer. The transformer architecture is changing incredibly. If not for the fact that CUDA is easy to operate on and iterate on, how do they try all of their vast number of experiments to decide which one of the transformer versions, what kind of attention algorithm to use? How do you disagregate? CUDA helps you do all that because it's so programmable. And so the way to think about our our our business now is you look at when when all of these ASIC companies or ASIC projects start 3, four, five years ago, I got to tell you that industry was super adorable and simple. There was a GPU involved, right? But now it's giant and complex and in another two years it's going to be completely massive. The scale is going to be so large. And so I think that the battle the the battle of getting into a very large market as a nent player is just hard you know as you guys know even for the customers who perhaps are successful with AS6.

没错。没错。英伟达的故事，你知道，所以对于现在正在构建ASIC的人来说，这是一个挑战，看起来市场很诱人，但请记住，这个诱人的市场已经从一个叫做GPU的芯片发展到了我刚才描述的AI工厂，你们刚刚看到我宣布了一款名为CPX的芯片，用于上下文处理和扩散视频生成，这是一个非常专业化的工作负载，但却是数据中心内重要的工作负载。我只是把它预告为可能是AI数据处理处理器，因为你猜怎么着？你需要长期记忆。你需要短期记忆。KV缓存处理非常紧张。AI内存是一件大事。你知道，你希望你的AI有良好的记忆力，并处理系统周围所有的KV缓存。非常复杂的东西。也许它需要一个专门的处理器。也许还有其他事情，对吧？所以你看到了，英伟达的——我们的观点不再是GPU。我们的观点是着眼于整个AI基础设施，以及如何让这些了不起的公司处理所有多样化且不断变化的工作负载。看看Transformer。Transformer架构正在发生巨大的变化。如果不是因为CUDA易于操作和迭代，他们怎么能尝试他们大量的实验来决定使用哪种Transformer版本、哪种注意力算法呢？你如何分解？CUDA可以帮助你做所有这些，因为它具有高度的可编程性。所以我们现在看待我们业务的方式是，你看看当所有这些ASIC公司或ASIC项目在三、四、五年前开始时，我得告诉你，那个行业超级可爱又简单。其中涉及一个GPU，对吧？但现在它庞大而复杂，再过两年它将变得非常庞大。规模将是如此之大。所以我认为，作为一个新进入者进入一个非常大的市场的战斗，就像你们所知道的，即使对于那些可能在ASIC方面取得成功的客户来说，也很困难。

算力集群中的硬件平衡点

Yeah. Isn't there an optimal balance in their compute fleet like it's you know I think investors are very much binary creatures. They just want a yes or no black and white answer. But even even if you get the ASIC to work, isn't there an optimal balance because you think I'm buying the Nvidia platform, CPX is going to come out for prefill for, you know, for video generation, maybe a decode, you know, you know, a platform video. Exactly. Yeah. So there will be like many different um chips or parts to add to the Nvidia ecosystem, accelerated compute fleet, right? as new workloads are, you know, are are born. That's right. And, you know, people trying to tape out new chips today are not really anticipating what's happening a year from now. They're just trying to get a chip to work. That's right. Set another way, Google's a big GPU customer. Google's a big GPU customer. Um, if you look at and Google is a very special case. I mean, we just have to, you know, show respect where respect is really deserved. I mean, uh, TPU is on TPU7. Yes. Right. And so um and and uh and it's a challenge for them as well, right? And so so the work that they do is incredibly hard. So so I think the first thing to to let let me do a um you know remember there there are three categories of chips.

是的。他们的算力集群中难道没有一个最佳平衡点吗？我想投资者是非常二元对立的生物。他们只想得到一个是或否、非黑即白的答案。但即使你让ASIC工作了，难道没有一个最佳平衡点吗？因为你认为我要购买英伟达平台，CPX将用于预填充，用于视频生成，也许用于解码，你知道，一个平台视频。没错。是的。所以英伟达的加速计算集群生态系统将会有许多不同的芯片或组件加入，对吧？随着新工作负载的诞生。没错。而且，你知道，今天试图定型新芯片的人并没有真正预见到一年后会发生什么。他们只是在努力让芯片工作起来。没错。换句话说，谷歌是主要的GPU客户。谷歌是主要的GPU客户。嗯，如果你看——谷歌是一个非常特殊的案例。我的意思是，我们只是必须——你知道——在真正应得的地方表示尊重。我的意思是，TPU已经发展到TPU7了。是的。对吧？所以，嗯，而且……而且这对他们来说也是一个挑战，对吧？所以他们所做的工作非常艰巨。所以，我想我先来做个……你知道，还记得有三类芯片吗？

芯片的三大类别：架构芯片、ASIC和COOT

There's the category chips that are architectural. X86 CPUs, ARM CPUs, Nvidia GPUs. Architectural And it has an ecosystem above and and um uh the architecture uh allows uh has rich IP and rich ecosystem very complicated technology. It's built by the owners like us. Okay. There's AS6. I worked for the original company LSI Logic who invented the idea of A6s. As you know LSI Logic is not here anymore, right? And the reason for that is because AS6 is really fantastic. When the the market size is not very large, um it's easy to have somebody uh be a contractor to help you put the packaging of all that stuff together and do the manufacturing on your behalf and they charge you 50 60 points of margin. But when the market gets large for an ASIC, there's a new way of doing things called coot, customerowned tooling. And who would who would do something like that? Um Apple's as Apple Apple's smartphone chip the volume is so large they would never go pay somebody else 50 60% gross margin to be an ASIC. They do customer own tooling. And so um where will TPUs go when they when it becomes a large business? Customer own tooling. There's no question about it. And so a but there's a place for A6s. Uh video transcoders will never be too large. Um, smart nicks will never be too large. And so when when there's 10, 12, 15 as projects going on at an ASIC company, I'm not surprised by that, you know, because there there probably five smart nicks and four transcoders and, you know, are they all AI chips? Of course not. You know, and and if somebody were to build an embedded embedding processor for a specific recommender system and that was an ASIC, of course you could do that. But would you do that as the fundamental compute engine for AI that's changing all the time? You've got low latency workload. You got high throughput lo workload. You have token generation for chat. You have thinking workload. You have AI video generation workload. Is there a you know now you're talking about the workhorse backbone of your accelerated?

有架构芯片。X86 CPU、ARM CPU、英伟达GPU。架构性的，它有一个生态系统在其上，嗯……架构有丰富的IP和丰富的生态系统，是非常复杂的技术。由我们这样的所有者构建。好的。还有ASIC。我曾在发明ASIC概念的LSI Logic公司工作过。你知道，LSI Logic已经不复存在了，对吧？原因是ASIC确实很出色。当市场规模不是很大时，嗯，很容易有人作为承包商来帮助你把所有东西的封装组合起来，并代表你进行制造，他们向你收取50%到60%的毛利率。但是当一个ASIC的市场变得很大时，有一种新的做事方式叫做COOT，即客户自有工具。谁会做这样的事情呢？苹果的智能手机芯片，销量如此之大，他们绝不会付给别人50%到60%的毛利率来做ASIC。他们做客户自有工具。那么，当TPU发展成为一项大业务时，它们会走向何方？客户自有工具。毫无疑问。但是ASIC也有其用武之地。视频转码器永远不会太大。嗯，智能网卡永远不会太大。所以，当一家ASIC公司有10个、12个、15个项目在进行时，我并不感到惊讶，你知道，因为可能那里有五个智能网卡和四个转码器，它们都是AI芯片吗？当然不是。你知道，如果有人为特定的推荐系统构建一个嵌入式处理芯片，并且那是一个ASIC，你当然可以这样做。但你会把它作为AI的基础计算引擎吗？AI在不断变化。你有低延迟工作负载。你有高吞吐量工作负载。你有聊天的Token生成。你有思考工作负载。你有AI视频生成工作负载。你现在谈论的是你的加速器的中流砥柱吗？

英伟达的定位：AI基础设施的构建者

That's what Nvidia is all about. Again, dumb this down. It's like playing chess and checkers, right? The fact of the matter is the folks who are starting AS6 today whether it's Tranium or whether it's some of these other accelerators etc. they're building a chip that's a component of a much larger machine. you've built a very sophisticated system, platform, factory, whatever you want to call it, and now you're opening up a little bit, right? So, you mentioned CPXGPU, right? That is, it seems to me that it in some ways you're disaggregating the workloads to the best slice of the hardware for that particular domain.

这就是英伟达的全部意义所在。再说一遍，把它简单化。这就像下棋和跳棋，对吧？事实是，今天开始研究ASIC的人，无论是Tranium还是其他一些加速器等等，他们构建的芯片是更大机器的一个组成部分。你已经构建了一个非常复杂的系统、平台、工厂，无论你想叫它什么，现在你稍微开放一点了，对吧？所以，你提到了CPXGPU，对吧？在我看来，这在某些方面是将工作负载分解到最适合该特定领域的硬件切片上。

Dynamo与MV Fusion的开放性

Well, we did we announced this thing called Dynamo, right? disagregated orch AI workload orchestration and we open sourced it because the future AI factory is disagregated right and you launched MV Fusion that even said to your competitors including Intel which you just invested in that's right you know the way in which you participate in this factory that we're building because nobody else is crazy enough to try to build the entire factory but you can plug into that if you have a product that's good enough, compelling enough that the end user says, "Hey, we want to use this instead of an ARM GPU or we want to use this instead of your inference accelerator, etc. Is that correct?" We're delighted delighted to connect you in. Yeah. Tell us a little bit fusion such a great idea and we're so happy to partner with Intel on that. It takes takes the Intel ecosystem, you know, there most of the world's enterprise still runs on Intel. It takes the Intel ecosystem, takes the Nvidia AI ecosystem, accelerated computing, and we fused it together, right? And we did that with ARM, right? And there are several others we're going to be doing it with. And and uh that that opens up opportunities for both of us. It's a win for both of us. Great great win. I'll be a large customer of theirs and uh they're going to expose us to a a much much larger market opportunity.

我们确实发布了一个名为Dynamo的东西，对吧？分离式AI工作负载编排（disaggregated orch AI workload orchestration），我们将其开源了，因为未来的AI工厂是分离的，对吧？你推出了MV Fusion，它甚至对你的竞争对手（包括你刚刚投资的英特尔）说，没错，你知道，你可以通过你所拥有的足够好的、足够有说服力的产品插入我们正在构建的这个工厂，以至于最终用户会说：“嘿，我们想用这个而不是ARM GPU，或者我们想用这个而不是你的推理加速器，等等。是这样吗？”我们非常高兴能让你参与进来。是的。Fusion是个很棒的主意，我们很高兴能与英特尔合作。它接纳了英特尔的生态系统，你知道，世界上大部分企业仍然运行在英特尔之上。它接纳了英特尔的生态系统，接纳了英伟达的AI生态系统、加速计算，并将它们融合在一起，对吧？我们也与ARM做到了这一点，对吧？我们还将与其他几家公司这样做。嗯，这对我们双方都开辟了机会。这对双方都是双赢。是一个巨大的胜利。我将成为他们的大客户，他们将让我们接触到更大得多的市场机会。

成本为王：英伟达系统的价值远超免费ASIC

Yeah. that's deeply related to this idea is the argument you've made that kind of um c shock some people where you say our competitors building A6s they could literally all their chips are cheaper already today but they could literally price them at zero our objective is they could price them at zero and you would still buy an Nvidia system because the total cost of operating that system power data center land etc the intelligence out is still a better bet than buying a chip even if it's given to you for free because the land power and shell is already $15 billion, right? Yeah. So, we've taken a crack at kind of the math on that. But walk us through your math because I think for people who don't spend as much time here that it just doesn't compute. How could it possibly be that you were pricing your competitor's chips at zero given the expense of your chips and it still is a better bet? There's two ways to think about it. Um, one way is um, uh, let's just think about it from a perspective of revenues. Yes. Okay. So everybody's power limited and let's say uh you were able to secure two more gigawatts of power. Well, that two gawatts of power you would like to have translate to revenues. Yes. So your performance or tokens per watt was twice as high as somebody else's token per watt because you did I did deep and extreme code design, right? and my performance was much higher per unit energy, then my customer can produce twice as much revenues from their data center. And who doesn't want twice as much revenues? and and and if somebody gave them a 15% discount, you know, the difference between our gross margins, which is called the 75 points, and somebody else's gross margins, call it the 50 to 65 points, is not so much as to make up for the 30 times difference between Black Wall and Hopper. Let's pretend Hopper Hopper is an amazing chip, an amazing system. Let's pretend somebody else's ASIC is Hopper. Yeah. Black Wall's 30 times. So you've got to give up 30x revenues in that one gigawatt. Mhm. It's too much to give up. So even if they gave it to you for free, you you you only have 2 gigawatts to work with. Your opportunity cost is so insanely high. You would always choose the best perf per watt.

与此相关的是你提出的一个论点，这个论点可能会让一些人感到震惊：你说的竞争对手制造的ASIC，他们的芯片今天可能已经更便宜了，但他们可以把价格定为零，我们的目标是，他们可以把价格定为零，而你仍然会购买英伟达的系统，因为运营该系统的总成本——电力、数据中心、土地等等——产生的智能，仍然比购买一个免费的芯片要划算，因为土地和电力成本本身就高达150亿美元，对吧？是的。所以我们试着计算了一下，但请向我们解释一下你的计算过程，因为对于那些不常接触这方面的人来说，这根本想不通。考虑到你们芯片的昂贵，你怎么可能给竞争对手的芯片定价为零，而这仍然是一个更好的选择呢？有两种思考方式。嗯，一种方式是……让我们从收入的角度来思考。是的。好的。所以每个人的电力都是有限的，假设你能再获得两吉瓦的电力。你希望获得的这两吉瓦电力将转化为收入。是的。所以你的每瓦性能或每瓦Token数是你竞争对手的两倍，因为我进行了深入和极端的协同设计，对吧？我的每单位能量的性能要高得多，那么我的客户就可以从他们的数据中心产生两倍多的收入。谁不想要两倍多的收入呢？而且，如果有人给他们15%的折扣，你知道，我们的毛利率（我们称之为75个点）与别人（我们称之为50到65个点）的毛利率差异，不足以弥补Blackwell和Hopper之间30倍的差距。让我们假设Hopper是一个了不起的芯片、一个了不起的系统。假设别人另一个ASIC是Hopper。是的。Blackwell是30倍。所以在这一个吉瓦的电力中，你必须放弃30倍的收入。嗯。放弃太多了。所以即使他们免费给你，你只有2吉瓦的电力可供使用。你的机会成本高得离谱。你会永远选择最好的每瓦性能。

持续的技术迭代与竞争优势

So I heard this from one of the CFOs at one of the hyperscalers that given the performance improvement right that's coming out of your chips again precisely to that point tokens per per gig um and power being the limiting factor right that they had to upgrade uh to the new cycle so when you look ahead at Ruben at Ruben Ultra at Fineman does that trajectory continue we're building what six seven chips a year now yeah and and each one that's part of that system. That's right. And those that system software is everywhere and it takes it takes the integration and the optimization across all of those six seven chips to deliver on the 30x blackwell. Now imagine I'm doing this every single year. Bam bam bam bam bam bam. And so if you build one ASIC in that soup of AS6 in that soup of chips and we're optimizing across that, you know, it's a hard problem to solve.

我从一位超大规模运营商的首席财务官那里听说了这一点，鉴于你们芯片带来的性能提升，正是基于每吉瓦Token数以及电力成为限制因素，他们不得不升级到新的周期。所以当你展望Rubin、Rubin Ultra、Fineman时，这个轨迹会持续吗？我们现在每年要制造六七个芯片，是的，而且每个芯片都是那个系统的一部分。没错。而且那个系统的软件无处不在，需要对所有这六七个芯片进行集成和优化，才能实现Blackwell的30倍提升。现在想象一下我每年都在做这个。砰砰砰砰砰砰。所以如果你在这个由ASIC组成的汤、一堆芯片中构建了一个ASIC，而我们正在跨越这些进行优化，你知道，这是一个很难解决的问题。

竞争壁垒的加深：规模与协同设计

This does bring me back to where we started about the competitive moat. We've been covering this and investors for a while. We're investors throughout the ecosystem and in competitors of yours, you know, from Google to to Broadcom. But when I really just first principles around this and say are you increasing or decreasing your competitive mode, you move to an annual cadence. You're co-developing with a with a supply chain. The scale is massively bigger than anybody anticipated which requires scale both of balance sheet and of development. Right? the moves you made both through acquisition and organically with things like Envy Fusion, CPX, which we just talked about. All of those things together cause me to believe that your competitive mode is increasing visav at least in so far as building out the factory or the system. It's at least surprising. But but I think it's interesting that your multiple is much lower than most of those other people. And I think part of that has to do with this law of large numbers. A $4.5 trillion company couldn't possibly get any bigger. But I asked you this a year and a half ago as you sit here today. If the market's going to AI workloads are going to 10x or 5x, you know, we know what capex is doing, etc. Is there any conceivable world in your mind where your top line in 5 years isn't two or 3x bigger than it is in 2025? Like what's the probability that it's actually not not much higher than it is today given those advantages?

这确实让我回到了我们开始谈论竞争护城河的地方。我们和投资者已经讨论了很长时间。我们是整个生态系统以及您的竞争对手（从谷歌到博通）的投资者。但当我真正从第一性原理来思考这个问题，并问你是否在增加或减少你的竞争护城河时，你转向了年度节奏。你与供应链共同开发。规模比任何人预期的都要大得多，这需要资产负债表和开发的规模。对吧？你所做的举动，无论是通过收购还是有机增长，比如我们刚才谈到的Envy Fusion、CPX。所有这些因素加在一起，使我相信你的竞争护城河正在增加，至少在构建工厂或系统方面是这样。这至少是令人惊讶的。但我认为有趣的是，你们的市盈率远低于其他大多数公司。我认为一部分原因在于大数法则。一家4.5万亿美元的公司不可能再变大了。但我一年前半就问过你这个问题，以你今天坐在这里的情况来看。如果市场转向AI工作负载将增长10倍或5倍，你知道资本支出正在做什么等等，在你看来，有没有任何可能的世界，五年后你的收入不会比2025年高出两三倍？考虑到这些优势，它比今天高出不多的可能性有多大？

英伟达：首家万亿级公司与AI基础设施提供商

I'll answer it this way. Our opportunity as I described it is much larger than the consensus. I'll say it here. I think Nvidia will likely be the first 10 trillion dollar company. And I would I've been here long enough, it wasn't that long ago, just a decade ago, as you well remember, that people said there could never be a trillion dollar company. Now we have 10, right? And today the world's bigger, right? And today, this is this is the back to the exponentials around GDP and the growth. The world is bigger and and and people misunderstand what we do. They they uh they remember we're a chip company, right? And we are we build chips. Boy, do we build chips and build the most amazing chips in the world. But NVIDIA is really an AI infrastructure company. You know, we are your AI infrastructure partner and our partnership with OpenAI is a perfect demonstration of that. Yeah. That we are their AI infrastructure partner and we work with people in a lot of different ways. You know, you you you would uh we we don't require anybody to buy everything from us. Um we don't we don't require that they buy uh the full rack. They could buy a chip. They could buy a component. They could buy our networking. They could buy our We have customers buying only our CPU. You know, just buy our GPUs and buy somebody else's CPUs and somebody else's networking. You know, we're kind of okay selling any way you like to buy. You know, my only request is just buy a little something from us.

我将这样回答。正如我所描述的，我们的机会比共识要大得多。我在这里说。我认为英伟达很可能成为首家10万亿美元的公司。而且我在这里待了足够长的时间，十年前，你还记得，人们说永远不会有万亿美元的公司。现在我们有10家了，对吧？而今天世界更大了，对吧？而今天，这是回到GDP和增长的指数级增长。世界更大了，人们误解了我们所做的事情。他们……他们……他们记得我们是一家芯片公司，对吧？我们是，我们制造芯片。我们确实制造了世界上最令人惊叹的芯片。但英伟达真的是一家AI基础设施公司。你知道，我们是您的AI基础设施合作伙伴，我们与OpenAI的合作正是完美的证明。是的。我们是他们的AI基础设施合作伙伴，我们以许多不同的方式与人合作。你知道，你……我们不要求任何人必须从我们这里购买所有东西。我们不要求他们购买整套机架。他们可以买一片芯片。他们可以买一个组件。他们可以买我们的网络设备。他们可以买我们的……我们有客户只买我们的CPU。你知道，只买我们的GPU，然后买别人的CPU和别人的网络设备。你知道，我们很乐意以任何你喜欢的方式销售。你知道，我唯一的要求是你从我们这里买点东西就行了。

埃隆·马斯克和AI超级计算机的构建能力

You know, you said, you know, this isn't just about better models. We also have to build. We have to we have to have world class builders. And you said, you know, the most world-class builder maybe that we have in the country is Elon Musk. And we talked about Colossus one and what he, you know, what he was doing there, standing up a couple hundred thousand, you know, at the time H100s, H200s in a coherent cluster. Now he's working on Colossus 2, you know, which may be 500,000 GBs, um, millions of H100 equivalents in a coherent cluster. I would not be surprised if he gets to a gigawatt before anybody else does in one. Yeah. So say a little bit about that. The advantage of being, you know, the builder who, you know, isn't just building the software and the models, but understands what it takes to to build those clusters.

你知道，你说，这不只是关于更好的模型。我们还必须建设。我们必须拥有世界级的建设者。你说，也许我们国家最世界级的建设者是埃隆·马斯克。我们谈到了Colossus 1以及他正在做的事情，他正在建立一个由数十万个H100、H200组成的连贯集群。现在他正在研究Colossus 2，它可能是50万个GB，数百万个H100等效物在一个连贯的集群中。我不会惊讶如果他能比任何人都先达到一个吉瓦的规模。是的。所以请多谈谈这一点。作为建设者，你的优势在于，你不仅仅是构建软件和模型，而且你理解构建这些集群需要什么。

构建AI超级计算机的复杂性与融资挑战

Well, you know, these AI supercomputers are complicated things. The technology is complicated. Procuring it is complicated because of financing issues. Securing the land power and shell, powering it is complicated. Building it all, bringing it all up. I mean, these are this is unfortunately the most complex systems problem humanity has ever endeavored. And and so Elon has has a great advantage that in his head um all of these systems are interoperating and um and the and the interdependencies um are are um you know resides in one head including the financing. Yes. And so he's a big GPT. He's a big supercomputer himself. He's the Yeah. the ultimate GPU. Yeah. Yeah. And so so he has a great advantage there. Yeah. And and he has a great sense of urgency. He Yeah. He has a has a real desire to to build it and and so when when will comes together with with skill. Yeah. You know, unbelievable things can happen. Yes. Yeah. Quite unique.

你知道，这些AI超级计算机是复杂的东西。技术很复杂。由于融资问题，采购它很复杂。确保土地、电力和外壳，为其供电很复杂。建造这一切，启动这一切。我的意思是，不幸的是，这是人类迄今为止尝试过的最复杂的系统问题。所以埃隆有一个巨大的优势，在他的脑海中，所有这些系统都在互操作，而且……而且相互依赖性都存在于一个人的脑海中，包括融资。是的。所以他是一个大GPT。他本身就是一个大超级计算机。他是……是的。终极GPU。是的。是的。所以他在那里有一个巨大的优势。是的。而且他有一种强烈的紧迫感。他……是的。他有一种真正的愿望去建造它，所以当意志与技能结合在一起时，你知道，不可思议的事情就会发生。是的。是的。非常独特。

主权AI：国家安全的关键

Something you've been so involved in is I want to talk about sovereign AI. I want to talk about China and the global AI race that's going on. You know, when I look back at you 30 years ago, you couldn't have imagined you were going to be hanging out in palaces with airs and the king this week and you're at the White House all the time. The president has said that you and Nvidia are critical to uh US, you know, national security. So when when you look at that first just contextualize for me like it's hard to believe that you would be in those places if sovereigns didn't view this at least as existential as important as maybe we did nuclear in the 1940s right we don't have a Manhattan project today at least funded by the government but it's funded by Nvidia it's funded by open AI it's funded by Meta it's funded by Google we have companies today the size of nation states and thank god God for America, right? Who are funding something that it appears to me presidents and kings think are think is existential to their future economic and national security. Would you agree with that?

你深度参与的事情是，我想谈谈主权AI。我想谈谈中国和正在进行中的全球AI竞赛。你知道，当我回顾30年前的你时，你无法想象这周你会和国王们在宫殿里闲逛，而且你总是在白宫。总统说你和英伟达对美国的国家安全至关重要。所以当你看到这一切时，首先请为我梳理一下，这很难让人相信，如果主权国家不认为这至少像我们认为1940年代的核能一样重要，你就能出现在那些地方，对吧？我们今天没有政府资助的曼哈顿计划，但它由英伟达、OpenAI、Meta、谷歌资助。我们今天有规模堪比国家级的公司，感谢上帝，感谢美国，对吧？他们正在资助一些事情，在我看来，总统和国王们认为这对他们未来的经济和国家安全至关重要。你同意吗？

AI的重要性超越原子弹

Nobody needs atomic bombs. Everybody needs AI. Well said. Okay. Here. Here. Yeah. Here. And and so that's a very very large difference. Um AI AI as you know is modern software. I just that's where I started from general purpose computing to accelerated computing from human written code line at a time to AI written code that foundation can't be forgotten we've reinvented computing there's not a new species on earth we just reinvented computing and everybody needs computing it needs to be democratized which is the reason why everybody all of these all of the countries realize they have to get into the AI world because everybody needs to stay in computing. There's nobody in the world that says, "Guess what? You know, I used to use computers yesterday. I'm pretty good with, you know, clubs and fire tomorrow, you know, and so everybody needs to move into computing. It's just it's just being modernized. That's all."

没有人需要原子弹。每个人都需要AI。说得好。好的。这里。这里。是的。这里。所以这是一个非常非常大的区别。嗯，AI，如你所知，是现代软件。我就是从通用计算到加速计算开始的，从手工编写代码到AI编写代码，这个基础不能被遗忘。我们重新发明了计算，地球上没有新的物种，我们刚刚重新发明了计算，而每个人都需要计算，它需要民主化，这就是为什么所有国家都意识到他们必须进入AI世界，因为每个人都需要留在计算领域。世界上没有人会说：“你猜怎么着？我昨天还在用电脑。我对用棍棒和火把过日子还不错，你知道，所以每个人都需要进入计算领域。它只是正在现代化。就这样。”

主权AI能力与文化价值观的编码

Okay. Number one, um it it is the case that that in order to participate in AI, you have to encode within AI, your your history, your culture, your values. And and of course AI is getting smarter and smarter so that even the core AI is able to learn these things fairly quickly. You don't have to start from the ground, you know, from ground zero. And so I I think that that every country um needs to have some sovereign capability. I recommend that they all use OpenAI, they all use Gemini, they all use, you know, these open models use Grok and I think they I recommend they all use anthropic. Um but they should also dedicate resources to learn how to build AI. And the reason for that is because they need to learn how to build it not just for language models, but they need to build it for industrial models, manufacturing models, national security, national security models. There's a whole bunch of intelligence they had to go cultivate themselves. So they they ought to have sovereign capability. Every country should develop it.

好的。第一点，嗯，事实是，为了参与AI，你必须在AI中编码你的历史、你的文化、你的价值观。当然，AI变得越来越聪明，以至于核心AI也能相当快地学习这些东西。你不必从头开始，你知道，从零开始。所以我认为每个国家都应该拥有一定的主权能力。我建议他们都使用OpenAI，都使用Gemini，都使用，你知道，这些开放模型，使用Grok，我认为他们……我建议他们都使用Anthropic。但他们也应该投入资源来学习如何构建AI。原因是他们需要学习如何构建它，不仅仅是语言模型，他们还需要为工业模型、制造模型、国家安全、国家安全模型来构建它。他们必须自己培养一整套的智能能力。所以他们应该拥有主权能力。每个国家都应该发展它。

AI基础设施建设对各国的重要性

And is that what you see? Is that what you're hearing around the world? They all realize it. They all realize they all realize it. And they they all are going to be customers of OpenAI and Throbic and Grock and Gemini, but they all really need to also build their own infrastructure. And this is this is the big idea that that what Nvidia does is we're building infrastructure. Just as every country needs energy infrastructure, the communications and internet infrastructure, now every single country needs AI infrastructure.

你看到的是这样吗？这是你环游世界听到的吗？他们都意识到了。他们都意识到……他们都意识到了。他们都将是OpenAI、Anthropic、Grok和Gemini的客户，但他们都真的需要建立自己的基础设施。这是个大想法：英伟达所做的是构建基础设施。正如每个国家都需要能源基础设施、通信和互联网基础设施一样，现在每个国家都需要AI基础设施。

美国在AI领域的人才与政策

So you let's start with the rest of the world. You know, our our good friend David Saxs. Um the AIS are doing a heck of a job. We are in so lucky. Yeah. To have David and Shriram in Washington DC. Um doing you know and David doing Yeah. doing AI in the AISR. Uh this at what a what a smart move by President Trump to put them in the White House. Um because during this pivotal time Yes the technology is complicated. Shriram is the only person in Washington DC that I think knows CUDA. Yeah. Um and and which is strange anyways. But but I I just love the fact that during this pivotal time when technology is complicated, policy is complicated, the impact to the future of our nation is so great that we have somebody who is clear-minded, dedicating the time to understand the technology and thoughtful to uh help us through that. And it would seem to me, yeah, I'm going back to the Manhattan Project analogy. Yeah. Right. that you have a president who understands how existential this is. You have governors like Greg Abbott in Texas who want to remove regulations to accelerate because they understand how important it is. You have secretaries right at energy and Doug Bergram at interior and Lutnik at commerce who also understand how important this is, how pro- energy they are. Could you imagine? Could you imagine the alternative if we had an administration right now who is not pro energy and want energy to grow in our nation so that we could have AI? I find it I just can't even think about it.

所以，让我们从世界其他地区开始。你知道，我们的好朋友戴维·萨克斯（David Sacks）。嗯，AI顾问们做得非常出色。我们在华盛顿特区有戴维和什里拉姆（Shriram），戴维在做AI顾问。特朗普总统让他们进入白宫真是个明智的举动。因为在这个关键时刻，是的，技术很复杂。什里拉姆是我认为在华盛顿特区唯一懂CUDA的人。是的。嗯，这很奇怪。但我只是喜欢这样一个事实：在这个技术复杂、政策复杂、对我们国家未来影响巨大的关键时刻，我们有头脑清晰的人，投入时间去理解技术，并深思熟虑地帮助我们度过难关。在我看来，是的，我要回到曼哈顿计划的比喻。对吧？总统明白这有多么重要。德克萨斯州的州长格雷格·阿博特（Greg Abbott）希望取消管制以加速发展，因为他们理解这有多重要。能源部长、内政部长道格·伯格拉姆（Doug Berrgram）和商务部长卢特尼克（Lutnik）也明白这有多重要，他们对能源有多么支持。你能想象吗？你能想象一下，如果我们现在的政府不是亲能源的，不希望我们国家的能源增长以便我们拥有AI，那会是怎样一番景象？我简直无法想象。

从核能到AI：产业与政府的合作

I find it ironic that that that just a couple years ago we were saying China's building a 100 nuclear reactors. They're so far ahead of us. Like that's the the the primitive to AI. But now you have people when we go to build it, everybody says, "Oh, it's a glut, right?" Like it seems to me that this is something that the government, it is in their interest. And we have industry and government working together in a way that I haven't seen in a long time. You've been around a long time. You you're very close with President Trump at this stage. Help us understand like what is the nature of industry government relationships? We saw that dinner last week with all the CEOs. You know, you spent a lot of time. Is it unique? Have you seen anything like this in your career over the last 30 years?

我发现很有讽刺意味的是，就在几年前，我们还在说中国正在建造100座核反应堆。他们在我们前面遥遥领先。这就像是AI的原始阶段。但现在，当我们去建设时，每个人都说，“哦，这是供应过剩，对吧？”在我看来，政府对此是感兴趣的。产业和政府之间的合作方式，在我很长一段时间里都没见过。你在这个领域很久了。你在这个阶段与特朗普总统关系密切。帮我们理解一下产业和政府关系是什么性质的？我们上周看到了所有CEO参加的晚宴。你知道，你花了很多时间。这独特吗？在你过去30年的职业生涯中，你见过这样的事情吗？

特朗普政府对增长和科技领先的重视

It was it was hard to go to DC in the past as you know. Uh getting an appointment is almost impossible, right? Uh President Trump has a open door to leaders who wants to come in and uh help them understand the future. Um this is an administration that believes in growth. Fundamentally, President Trump wants America to grow. Yeah. If we can grow economically, we will be strong militarily. If we could be if we could grow economically, we will be secure. I've never met somebody who is secure who's poor. Being being rich as a nation is an essential part of national security. And he knows that. He also wants America to win the AI the the AI race. This is going to be a very long-term race. And um and he understands that this is a pivotal time. He wants the technology industry to run. He wants everybody in the world to be built on American technology. These are sensible, logical things. You know, the opposite is strange to me. If I take everything and I just reversed it, we want our country not to grow and because we don't want our country to grow, we don't need any energy because we know we need energy to grow and so let's not have any energy and in fact we don't want our technology industry to lead. Uh he understands that our technology industry is our national treasure. Correct. And that technology like corn and steel and things in the past are now such fundamental trade opportunities. It's an essential part of trade. And why would you not want American technology to be coveted by everyone so that it could be used for trade?

如你所知，过去去华盛顿特区是很困难的。嗯，约到一个会面几乎是不可能的，对吧？特朗普总统对希望进来了解未来的领导人持开放态度。这是一个相信增长的政府。从根本上说，特朗普总统希望美国发展。是的。如果我们能在经济上发展，我们就会在军事上强大。如果我们能在经济上发展，我们就会安全。我从未见过一个安全但贫穷的人。一个国家富有是国家安全的一个基本组成部分。他知道这一点。他还希望美国赢得AI竞赛。这将是一场非常长期的竞赛。而且……而且他明白这是一个关键时刻。他希望科技行业蓬勃发展。他希望世界上所有的一切都建立在美国技术之上。这些都是合理、合乎逻辑的事情。你知道，相反的情况对我来说很奇怪。如果我把所有的一切都反过来，我们不希望我们的国家发展，因为我们不希望国家发展，所以我们不需要任何能源，因为我们知道我们需要能源才能发展，所以我们就不搞能源了，事实上我们不希望我们的科技行业领先。他明白我们的科技行业是我们的国家财富。没错。这项技术就像过去的玉米和钢铁一样，是如此基本的贸易机会。它是贸易的基本组成部分。你为什么不希望美国技术受到所有人的追捧，以便它可以用于贸易呢？

推广美国技术与加速出口许可

Right. So let's talk about you know the internet. Yeah. Google spread around the world. Yeah, we had democratic values spread around the world by way of search and Google didn't have to go to Washington to get permission to do it. It just happened. We diffused our technology around the world. David Sachs has been crystal clear of the need to accelerate export licenses so that the American AI stack wins around the world. Right? We're talking chips, we're talking models, we're talking data centers, etc. We know a year and a half ago that wasn't happening. was a concept that was called small yard tall fence or something like that. A small yard tall fence and and the irony of it was it was described in such a way and it was recommended in policy in such a way it was a small yard tall fence around America. That was the strange part. I think President Trump's got it right that we want to maximize exports. We want to maximize American influence around the world. We're supposed to maximize those things. And do you see those licenses coming? Are you seeing the acceleration in Washington? I know it's being said at the top, but are you seeing it flow down through government that's accelerating us around the world? Secretary Lutnick was all over it. Great. Yeah.

对。所以我们来谈谈互联网。是的。谷歌遍布全球。是的，我们的民主价值观通过搜索传播到世界各地，谷歌不必去华盛顿申请许可。它就是发生了。我们把我们的技术传播到了全世界。戴维·萨克斯已经非常明确地指出了加速出口许可的必要性，这样美国AI技术栈就能在全世界获胜。对吧？我们谈论芯片、模型、数据中心等等。我们知道一年半前，这种情况还没有发生。有一个概念叫做“小院高墙”之类的。一个小院子，高高的围墙，而它的讽刺之处在于，它被描述和推荐给政策的方式，是围绕着美国的一道小院高墙。那是奇怪的部分。我认为特朗普总统的看法是正确的，我们希望最大化出口。我们希望最大化美国在世界上的影响力。我们应该最大化这些事情。你看到这些许可要来了吗？你在华盛顿看到加速了吗？我知道高层在谈论，但你是否看到它向下渗透到政府层面，从而加速我们在世界范围内的发展？卢特尼克部长非常关注此事。很好。是的。

中国市场的竞争与不应采取的“单边解除武装”

So, now let's talk about China. You know what most people may not realize is I think you understand China as well as any leader in the United States. We've been there for 30 years. Been there for 30 years. What most people don't realize is is up until a couple years ago, you had dominant market share within China in terms of 95% market share. 95% market share in the most important thing arguably. And you have said that our biggest goal that we as a country could have under the guise of somehow trying to slow them down is we've unilaterally disarmed. We forced Nvidia out of China, which has allowed Huawei to accelerate on the back of monopoly profits within China. And I just saw this morning, you're seeing announcements out of Huawei and Baba and others that they're going to build data centers around the world. Now Huawei has a three-year plan to pass Nvidia funded by the monopoly profits in the biggest AI market in the world. So it's looking like your admonition that this is a huge mistake to hand China monopoly markets is coming true. The president said, you know, a a after kind of the ban on H20s, now we have a situation where you can sell, you know, chips to China, but there's a 15% export tax. But now it appears that the Chinese perhaps offended by statements out of the United States are saying no, Nvidia is not allowed to sell here. Now where do we stand today between Nvidia and China? And can you reiterate kind of what you think we as a country should be doing to put ourselves in a best position to win the AI race around the world?

现在我们来谈谈中国。大多数人可能没有意识到的是，我认为你对中国的了解不亚于美国任何一位领导人。我们已经在那里待了30年。已经待了30年了。大多数人没有意识到的是，直到几年前，你在中国拥有主导的市场份额，高达95%。在可以说最重要的领域拥有95%的市场份额。你说过，我们国家在试图减缓他们的同时，可以做的最大的事情就是我们单方面解除了武装。我们将英伟达赶出了中国，这使得华为得以在中国垄断利润的基础上加速发展。我今天早上刚看到，华为、阿里巴巴等公司宣布他们将在世界各地建立数据中心。现在华为有一个三年计划，要在全球最大的AI市场中，依靠垄断利润超越英伟达。所以，你曾告诫我们不要将垄断市场拱手让给中国是一个巨大的错误，这个警告似乎正在成真。总统说，在禁止H20之后，现在的情况是你可以向中国出售芯片，但有15%的出口税。但现在看来，中国人可能因为美国的一些言论而感到被冒犯，他们说不，不允许英伟达在这里销售。那么，我们今天在英伟达和中国之间处于什么位置？你能重申一下你认为我们国家应该做些什么才能使我们处于赢得全球AI竞赛的最佳位置吗？

竞争与共存：中国应是开放市场

We have a competitive relationship with China. We should acknowledge that China rightfully should want their companies to do well. We I don't I don't for a second begrudge them for them. They should do well. They should they should give them as much support as they like. It's all their prerogative. And don't forget that China has some of the best entrepreneurs in the world because they came from some of the best STEM schools in the world. They're they're the most hungry in the world. Yes. 996 as you know. This is a very producing the most AI engineers in the world. 96. So the audience knows 9 in the morning to 9 at night 6 days a week. That is their culture. Yeah. Okay. We're up against a formidable, innovative, hungry, fast moving, underregulated. Yeah. Okay. People don't realize this. They are very lightly regulated, right? Less regulated, ironically, than we are in a capitalist system. That's right. People think that they're centrally governed. But remember, the genius of China was distributed economic systems. Yeah. And so all of these 33 provinces and all the mayor economy has driven enormous amount of internal competition, internal economic vibrancy, which of course has some of its side effects. But this is a vibrant, entrepreneurial, high techch, modern industry. And two, one, uh, some of the things I I heard, uh, they could never build AI chips. That just sounded insane. Two, uh, that China can't manufacture. China can't manufacture. If there's one thing they could do is manufacture. And three, they're years behind us. Is it two years, three years? Come on. They're nanconds behind us. Nanoc. Yeah, they're nanconds behind us. And so we've got to go compete.

我们与中国存在竞争关系。我们应该承认，中国理应希望他们的公司做得好。我一秒钟都不会因此而责怪他们。他们应该做得好。他们应该给予他们喜欢的任何支持。这是他们的特权。别忘了，中国拥有世界上最优秀的创业者之一，因为他们来自世界上最好的STEM学校。他们是世界上最渴望成功的。是的。你知道的996。他们正在培养世界上最多的AI工程师。96。所以听众们知道，早上9点到晚上9点，一周工作6天。这就是他们的文化。是的。好的。我们面对的是一个强大、创新、饥渴、快速发展、监管不力的……是的。好的。人们没有意识到这一点。他们的监管非常宽松，对吧？具有讽刺意味的是，他们的监管比我们在资本主义体系中还要宽松。没错。人们认为他们是中央治理的。但请记住，中国的天才在于其分散的经济体系。是的。所以所有这33个省份和所有的地方经济都推动了巨大的内部竞争、内部经济活力，当然这带来了一些副作用。但这是一个充满活力、创业精神、高科技、现代化的产业。第二，我听到的一些说法，他们说他们永远造不出AI芯片。这听起来太疯狂了。第二，他们说中国不会制造。中国不会制造。如果说有一件事他们能做到，那就是制造。第三，他们落后我们好几年了。是两年，是三年吗？算了吧。他们只落后我们纳秒级。纳秒级。是的，他们只落后我们纳秒级。所以我们必须去竞争。

合作共赢的理性期待

Yeah, we've got to go compete. And so so the question then becomes um what's in the best interest uh what's in the best interest of China of course um is that they have a vibrant industry. Uh they also publicly say and rightfully I believe they believe this is that they want China to be an open market. They want to attract uh foreign investment. They want companies to come to China and compete in the marketplace, right? And I believe that they I hope I believe and I hope that would return to that in our context. Answering your question, what do I see in the future? I do hope because they they say it um their leaders say it and I take it at face value and I believe it because I think it makes sense for China that what's in the best interest of China is for foreign companies to invest in China, compete in China and for them to also have vibrant competition themselves and they would also like to come out of China and participate around the world. That is I think is a fairly sensible outcome and we what we need to do as a country is to enable our technology industry which today is the I'm privileged to be working in an industry that is our national treasure. We have to acknowledge it is our national treasure. It is our best industry. It is our single best industry.

是的，我们必须去竞争。所以问题就变成了，什么最符合中国的利益？当然，最符合中国利益的是他们拥有一个充满活力的产业。他们也公开表示——而且我相信他们是正当地相信这一点——他们希望中国成为一个开放市场。他们希望吸引外国投资。他们希望公司来中国竞争市场，对吧？我相信他们……我希望，我认为他们……我希望我们能在我们的背景下回到这一点。回答你的问题，我对未来有何看法？我确实希望，因为他们说了，他们的领导人说了，我完全相信，因为我认为这对中国最有利的是外国公司投资中国、在中国竞争，同时他们自己也有充满活力的竞争，他们也希望走出去，参与到世界各地。我认为这是一个相当明智的结果，而我们作为一个国家需要做的是赋能我们的技术产业，今天我非常荣幸能身处这个我们称之为国家财富的行业。我们必须承认它是我们的国家财富。它就是我们最好的产业。它就是我们唯一的最好的产业。

竞争的本质：我们相信美国科技的胜利

Yeah. Why would we not allow this industry to go compete for its survival for for this industry to go and proliferate the technology around the world so that we could have the world be built on top of American technology so that we can maximize our economic success magn uh maximize our geopolitical influence u maximize this this technology industry during such a vibrant time such a pivotal time to allow it to thrive. The skeptic says Jensen just wants to sell more chips and if he can sell them to China, great. He'll sell them to China. He doesn't care about, you know, what that means for America. That's a skeptic. Now, can I just can I just address the skeptics? Just because I want America in ecosystem and economy to grow doesn't make me wrong. Right. Right. Okay. So, first of all, everything that has been said so far that's been made up so far about U.S. China has proven to be wrong. The facts are just wrong. The ground truth is wrong. And so just because we want America to win, just because we want this industry to grow, doesn't make me wrong. Correct. And I think anybody who knows you and now the president, certainly myself, you deeply care about the country. You deeply want the United States of America to win the global AI race. You just happen to believe and I think you have as much experience or more experience than anyone that it enures to our advantage the probability of us winning the global AI race actually goes up if you are competing in China because it allows us to tap into half of the world's AI engineers keeping them you know in this ecos let's be clear with the companies we're talking about here bite dance Alibaba etc these are companies that are largely owned by American investors Yeah. Right. Right. Like these are global companies that are building recommender engines that by the way extraordinary technologies, incredible companies. And so I think and I'm hopeful that the argument that you're making visav China, which is a harder argument than diffusion to the rest of the world. I understand that. And that's why I thought when the president said, you know, I don't know, it's a flip of a coin. Maybe Jensen's right. Maybe the other guys are right. If Jensen's willing to put a little bit of 15% into the US Treasury as a hedge on that, then I'll go for it. Um, but I was really disappointed on the heels of that.

是的。我们为什么不让这个产业去竞争求生存，让这个产业去向全世界推广技术，这样我们就能让世界建立在美国技术之上，从而最大化我们的经济成功，最大化我们的地缘政治影响力，在这个充满活力的关键时刻，让这个技术产业蓬勃发展。怀疑论者说，黄仁勋只是想卖更多的芯片，如果他能卖给中国，那就太好了。他会卖给中国。他不在乎这对美国意味着什么。那是怀疑论者。现在，我能为怀疑论者辩护吗？仅仅因为我想让美国的生态系统和经济增长，并不意味着我是错的。对吧？对吧？好的。首先，到目前为止关于中美所说的一切，都被证明是错误的。事实就是错的。基本真相是错的。所以仅仅因为我们希望美国获胜，仅仅因为我们希望这个行业增长，并不意味着我是错的。没错。我想任何了解你和现在总统的人，当然包括我自己，都深切关心这个国家。你非常希望美利坚合众国赢得全球AI竞赛。你只是碰巧相信——而且我认为你的经验和任何人都一样多，甚至更多——如果你在中国竞争，我们赢得全球AI竞赛的概率实际上会增加，因为这使我们能够接触到世界上一半的AI工程师，让他们留在……我们说清楚，我们谈论的公司，如字节跳动、阿里巴巴等，这些公司很大程度上由美国投资者持有。是的。对吧？对吧？这些是构建推荐引擎的全球性公司，顺便说一句，它们是了不起的技术，了不起的公司。所以我想，而且我希望你针对中国提出的论点，这比向世界其他地区进行“扩散”的论点更难。我理解这一点。这就是为什么我想，当总统说，“我不知道，这就像抛硬币一样。也许黄仁勋是对的。也许其他人是对的。如果黄仁勋愿意拿出一点15%存入美国国债作为对冲，那我就支持。”嗯，但在这之后我真的很失望。

对中美关系的期望与贸易的必要性

Um, I think if the Chinese feel like they're being taken advantage of that we're going to send them chips that are, you know, 10 years old or something, then I then I get why why they had that response. H20 is really quite spectacular still. And I of course it it's it's not as good as Blackwell and and I get that. Yeah. Um I look, you know, we're I'm patient and and I believe that they're they're wise. They're thinking through their situation. Um they they have they have larger agendas to to deal with um rel you know visibly uh the United States. There are a lot of discussions going on. But I'll come back to the ground truth fundamental truth I believe that is in the best interest of China that Nvidia is able to serve that market and compete in that market. I fundamentally believe is in the best interest of China. It is of course um fantastic in the fantastic interest of the United States. Yeah, it is fun. But those two truths can coexist. It is possible for both to be true and I believe it is both true. And so I I um uh I I'm rather, you know, even though I tell all of our investors Yeah. that our guidance includes no China. Yeah. And I appreciate all of our investors to include no China in any of our guidance. We've got plenty of growth opportunities outside and we you know we've got all of that is true. It doesn't make China not important to us. It's very important to us. Anybody who thinks that the Chinese market is not important is has their head deep in the sand. Yeah. And so this is s one of the most important markets in the world. Smart markets as you know, smart people doing smart things and we want to be there. Yeah. And I think it's in the best interest of both countries that we are there. And so I think when I take a step back, I am confident that ultimately the wisdom will prevail. Yes. Yes. I've I've always been confident that wisdom prevails. I've always been confident that that truth prevails and uh it's taken me this far and uh I believe I believe that to be fundamentally true now. And so these things will get sorted out and we will have the opportunity to go compete in that China market.

嗯，我认为如果中国人觉得他们被占了便宜，我们会给他们发送10年前的芯片或者类似的东西，那我理解他们为什么会有那样的反应。H20仍然非常出色。当然，它不如Blackwell好，我明白这一点。是的。嗯，看，你知道，我很耐心，我相信他们很明智。他们正在思考他们的情况。他们有更大的议程要处理，你知道，很明显，美国，有很多讨论正在进行。但我会回到基本真相，我认为对中国最有利的是英伟达能够服务那个市场并在那个市场竞争。我从根本上相信，这对中国最有利。当然，这对美国也极其有利。是的，这很有趣。但这两个真相可以共存。两者都可能是真的，我相信两者都是真的。所以我……嗯……我更倾向于，即使我告诉我们所有的投资者，是的，我们的指引中不包括中国。是的。我感谢我们所有的投资者，将不包括中国纳入我们的任何指引中。我们在外部有充足的增长机会，你知道，这一切都是事实。但这并不意味着中国对我们不重要。它对我们非常重要。任何认为中国市场不重要的人，都是把头深深埋在沙子里了。是的。所以这是世界上最重要的市场之一。如你所知，是精明的市场，是聪明人在做聪明事，我们想在那里。是的。我认为我们都在那里对两国都是最有利的。所以我想，当我退后一步时，我有信心最终智慧会占上风。是的。是的。我一直相信智慧会占上风。我一直相信真理会占上风，嗯，它把我带到了今天，我相信这从根本上是正确的。所以这些事情都会得到解决，我们将有机会在中国市场竞争。

H-1B签证费用：对美国品牌和人才吸引力的影响

I'm not very political, but very topical is the administration's decision to charge 100,000 per H1B visa. Mhm. You've spent a lot of time with the president. Um you've called him our secret weapon in AI. I also know you want to recruit the best and brightest to our country. Yeah. So, how do you think about the decision to charge 100,000 per H1B visa? Does this make it easier or harder to recruit talent? And, you know, does perhaps it's a little different for large companies or small companies? Like, how do you think about it?

我不太关心政治，但当前的热门话题是政府决定对每份H-1B签证收取10万美元的费用。嗯。你和总统共度了很多时间。你称他为我们在AI领域的秘密武器。我也知道你想招募我们国家最优秀、最聪明的人才。是的。那么，你如何看待对每份H-1B签证收取10万美元的决定？这会使招募人才变得更容易还是更难？你知道，这对于大公司和小公司来说可能略有不同？你是怎么看的？

对H-1B费用的初步评价

I'm going to start with it's a great start. Hold on. You said it's a great start. It's a great start. I'm just going to start there. And the reason for that is this. That implies I don't I hope it's not the end, but I think it's a great start. I just hope it's not the end. Here's what I fundamentally believe. America has one a singular brand reputation that no country in the world has. And no country in the world is in a position or in the horizon to be able to say come to America and realize the American dream. M what country has the word dream behind it? Yes, it's part of its brand. We are utterly singular and you're talking to somebody who represents the American dream. My parents didn't have any money. Sent us over here. We started from nothing. You guys know I, you know, bust tables, wash dishes, clean toilets, and here I am. Yeah. This is the American dream. President Trump knows that we want legal immigrants. Yeah. there's a difference between legal immigrants and illegal immigrants. But the idea that it's a country that's free for all doesn't make sense. And so now the question is how do we go from the the idea that we want to protect fundamentally the American dream to dealing with illegal immigrants at such a large scale? Um how do we find a logical pragmatic solution? Right? So, the idea that he that that that we put a $100,000 price tag on H-1B um probably sets the bar a little too high, but as a first bar, it at least eliminates um illegal immigration, and that's a good start.

我将从“这是一个好的开始”说起。等等。你说这是个好的开始。这是一个好的开始。我将从那里开始。原因如下。这暗示着……我希望这不是结束，但我认为这是一个好的开始。我只希望它不是结束。我的基本信念是：美国拥有一个独特的品牌声誉，世界上任何国家都没有。世界上没有任何国家处于或有能力说“来美国实现美国梦”。哪个国家的名字后面带着“梦”这个词？是的，这是它品牌的一部分。我们是独一无二的，你正在和一位代表美国梦的人交谈。我的父母身无分文，把我们送到了这里。我们白手起家。你们知道我……你知道，我曾端过盘子、洗过碗、打扫过厕所，而现在我站在这里。是的。这就是美国梦。特朗普总统知道我们想要合法移民。是的。合法移民和非法移民是有区别的。但一个对所有人开放的国家这个想法是不合理的。所以现在的问题是，我们如何从保护美国梦的基本理念，转变为处理如此大规模的非法移民问题？嗯，我们如何找到一个合乎逻辑、务实的解决方案？对吧？所以，他——他给H-1B贴上10万美元的价格标签的想法，可能把门槛设得太高了，但作为一个初步的门槛，它至少可以消除对H-1B的滥用，这是一个好的开始。

聆听者与务实解决之道

How does it how does eliminate illegal immigration? Well, it at least it eliminates abuse of abuse of H-1B. Yeah. Yeah. At least. And and that's a good start. and at least we can have a conversation. So, one of the things that we know about President Trump, he's he he's a good listener. He actually listens. I mean, he listens to you, he listens to me, and he doesn't have to. And he listens to a lot of people, and he's integrating a lot of information, and this is obviously a very complicated issue. And so, so I think that that this is a fine start. It's a fine start. But I I I'm not confused that that anyone in the administration, anyone in the White House is confused that legal immigration, immigration is the foundation of the American dream and is the ultimate brand that we want to protect and that's the future we want to protect.

它如何消除非法移民？好吧，至少它消除了对H-1B的滥用。是的。是的。至少如此。这是一个好的开始，至少我们可以进行对话。我们知道关于特朗普总统的一件事，他是一个好的倾听者。他真的会听。我的意思是，他听你的，他听我的，而且他不必听。他听取很多人的意见，他正在整合大量信息，这显然是一个非常复杂的问题。所以，我认为这是一个不错的开始。这是一个不错的开始。但我并不感到困惑，行政部门的任何人、白宫的任何人都会感到困惑，即合法移民是美国梦的基础，也是我们想要保护的最终品牌，这也是我们想要保护的未来。

保持品牌吸引力的重要性

And I would also say it seems to me that certainly Saxs and other people in the administration know that we have to recruit the world's best and brightest. we should not sacrifice the greatness of the brand. Um, charging $100,000 or let's say, you know, it got lowered to 50 or whatever the case is, it does seem like it it it it tilts the playing field in favor of big companies who can effectively sponsor all these people, right? And it's more challenging for the startup ecosystem where people are already super expensive and now I got to pay this fee on top of it. It also has an un an unintended consequence. It might um accelerate investment outside United States, right? And so there there are unintended consequences, but like I said, start somewhere, move towards the right answer, right? You know, often times people want to go directly from a wrong answer, wrong condition. We don't want this condition where we're at, right? And directly jump to the perfect answer is hard to find, right? Just start somewhere. It's the entrepreneurial way.

我还会说，在我看来，当然萨克斯（Sachs）和其他政府人员都知道我们必须招募世界上最优秀、最聪明的人才。我们不应该牺牲品牌的伟大。嗯，收取10万美元，或者假设降低到5万美元，无论哪种情况，这似乎确实有利于那些能够有效地担保所有这些人[人才]的大公司，对吧？这对初创企业生态系统来说更具挑战性，因为那里的员工已经非常昂贵了，现在我还要支付这笔费用。它还会产生一个意想不到的后果。它可能会加速美国的海外投资，对吧？所以这里存在意想不到的后果，但就像我说的，先从某个地方开始，朝着正确的答案前进，对吧？你知道，很多时候人们想直接从错误的答案、错误的情况转向完美的答案，这很难找到，对吧？只是从某个地方开始。这是创业者的方式。

留住STEM人才的战略规划

It's important to me, you know, the president talked about before when he was running for office, he wanted to staple a green card to the, you know, to the diplomas of these STEM students. so smart people coming to the United States from from China AI researchers studying at Stanford like we want to keep them here we want to get you know and by the way if their families can't get here they're going to leave after a few years so you might even want to make it easier for their families to come here and others are you confident that we have a strategic plan in this administration you know this is a start but your convers conversations they give you confidence that we have a broader strategic plan to make sure we're recruiting the best and the brightest. I don't know that I have an answer for that. Okay. Um but I understand that where we're at is not where we want to be. Yeah. And I don't think anybody's lost lost their focus on, you know, the American dream, the importance of immigration, the importance of attracting all of the world's best talent to United States, create the conditions for them to stay here. I there are things that are done um from time to time that works against what I just described, right? Um making making foreign students uncomfortable being here in the brand threatens the brand. Um uh let's not let's not forget that that it's okay to be competitive with China, but be careful not to be tough on Chinese. And so we need to make sure that that slippery slope isn't crossed. Yeah. Um, you know, and so there there all of these things that goes along with with finesse and nuance. But the fact of the matter is we know where we want to be. We know we're in a difficult situation. We don't want to be here and President Trump doesn't have much time to move us in that direction. Right. And so to the extent that we move in that direction, I believe it's a good start. Agreed.

你知道，这对我来说很重要，总统在竞选时提到，他想把绿卡钉在这些STEM学生的文凭上。所以来自中国的聪明人，在斯坦福学习的AI研究人员，我们想让他们留在这里，你知道，顺便说一句，如果他们的家人来不了，他们几年后就会离开，所以你甚至可能想让他们的家人更容易来这里。其他人呢？你对这个政府是否有一个战略计划有信心吗？你知道，这是一个开始，但你的谈话让你有信心认为我们有一个更广泛的战略计划，以确保我们正在招募最优秀、最聪明的人才。我不知道我对这个问题是否有答案。好的。嗯，但我明白我们现在的处境不是我们想去的地方。是的。而且我不认为有人会失去对美国梦的关注，对移民重要性的关注，对吸引全世界最优秀人才来到美国、为他们创造留下来的条件的关注。我……有时会做一些事情，违背了我刚才描述的[目标]，对吧？嗯，让外国学生在这里感到不舒服，这会损害我们的品牌。嗯……让我们不要忘记，与中国竞争是可以的，但要小心不要对中国人太苛刻。所以我们需要确保不跨越那条滑坡。是的。嗯，你知道，所有这些事情都需要技巧和细微差别。但事实是，我们知道我们想去哪里。我们知道我们处于一个困难的境地。我们不想呆在这里，而特朗普总统没有太多时间将我们引向那个方向。对吧？所以，在某种程度上，我们朝着那个方向前进，我相信这是一个好的开始。同意。是的。

中国顶尖AI人才流失的严重性

I heard from a Chinese researcher leading one of our leading labs in the US that three years ago 90% of the top AI researchers graduating from universities in China wanted to come to the United States and did come to the United States to work in our leading labs and he guessed that today that's closer to 10 or 15%. Right? So seen a precipitous drop. That's precisely a concern that we have, right? So have you seen this? Have you you you know you're you're paying attention to both markets. Do you see this? And what are the things we need to do in order to reverse that?

我从一位领导美国顶尖实验室的中国研究人员那里听说，三年前，90%从中国大学毕业的顶尖AI研究人员想来美国，并确实来到了美国在我们的顶尖实验室工作。他估计，今天这个比例更接近10%或15%。对吧？所以看到了急剧下降。这正是我们所担心的，对吧？你看到这个了吗？你……你知道，你同时关注这两个市场。你看到了吗？为了扭转这种局面，我们需要做些什么？

吸引人才的关键指标（KPIs）

Definitely see a greater concern of of Chinese students um who who come here and um uh remain here. Yeah. and or many of them who come here for for school and are thinking about going elsewhere, right? Many of them thinking about Europe, right? And so so I think I think we need to be super super concerned about this. This is this is this is a source of existential crisis. This is definitely the early indicators of a future problem. Right. Right. you know, smart people's desire to come to to America and smart people smart students desire to stay, those are what I would call KPIs. Yes. Early indicators of future success. Yes.

我绝对看到了中国学生来到这里并留在这里的担忧越来越大。或者他们中的许多人来这里上学，然后考虑去别处，对吧？很多人在考虑欧洲，对吧？所以我想，我们必须对此非常非常担心。这是一个生存危机的来源。这绝对是未来问题的早期指标。对吧？对吧？你知道，聪明人想来美国，聪明学生想留下，这些都是我所说的关键绩效指标（KPI）。是的。未来成功的早期指标。是的。

品牌价值与人才吸引力的类比

I think of it a bit like the Warriors. You know, if they have an advantage of recruiting all the best players in the NBA, right, then they can continue to win championships. Y, but the second that recruiting pipeline, right, because of the brand of the Warriors gets diminished or something else happens, um, then they're not going to be able to recruit the best future players and you're not going to win championships. And when I you talk about the American dream so eloquently, that being brand USA, right? The right to come here and to to do what you've done. And you know, so I hope that the feedback to this administration, it's not just the administration, it's also just how we as a country talk about immigration. That's right. Right. This needs to be the place that welcomes the best and the brightest, that attracts, has a strategic plan for recruiting the best and the brightest and making sure that this is the place that they want to work.

我把它想象成勇士队。你知道，如果他们有招募NBA所有最佳球员的优势，对吧，那么他们就可以继续赢得总冠军。但是，如果由于勇士队的品牌，招募渠道受到影响，或者发生了其他事情，嗯，那么他们将无法招募到未来最优秀的球员，你也无法赢得总冠军。而你如此雄辩地谈论美国梦，那就是品牌美国，对吧？来到这里并成就一番事业的权利。你知道，所以我希望对本届政府的反馈，不只是针对政府，也关乎我们作为一个国家如何谈论移民。没错。对吧。这里需要成为一个欢迎最优秀、最聪明人才的地方，要具有吸引力，要有一个招募最优秀、最聪明人才的战略计划，并确保这里是他们想工作的地方。

“对华鹰派”的标签与爱国主义的悖论

As you know, there's in in there's a there's a phrase, and I didn't hear about this phrase until just a few years ago, China hawks. Yes. And and apparently that if you're a China hawk, you get to wear that label with pride. It's almost like a badge of honor, right? It's a badge of shame. There's no question it's a badge of shame. There's no question that although they want what's in the best interest of our country, and we all want what's in the best interest of our country, um destroying that pipeline, right, of the American dream, Yeah. is not patriotic, right? They they think they think they're doing the right thing for our country, but it's not patriotic. Not Not even a little bit. And so we we need to um continue to be uh the great country we are, to have the confidence, right, of a great country. Yes. Well said. And to have the confidence of a great country and and have somebody who wants to compete with us and to have the attitude, bring it on. Right. Right. Bring it on. Right. because I believe in I believe in our people. I believe in our people. I believe in the people that are here. I believe in our culture. I believe in our country. I believe in our system. Bring it on. And is it your take that that's where the president is? Like he's a he's a pragmatist. He's a he's a believer in the growth and the ability of the United States to compete. Um it seems to me that's where he is. There's no question President Trump is the bring it on president. Right. Right. And he doesn't seem to me like the reason I'm confident and I've said on this pod that I think he'll get a big deal done with China. I I really really do hope so. Yeah. And and I I think he he speaks he he speaks positively um uh uh with great respect and um great eloquence about about his relationship and the importance of China. Uh, not one time have I ever heard him say the word decouple, which we heard a lot in the last administration, right? Um, you can't decouple against um the single most the two most important relationships for the next century. That doesn't make any sense at all. Decoupling is exactly the wrong concept.

你知道，有这样一个词，我几年前才听说过，“对华鹰派”。是的。而且显然，如果你是一个对华鹰派，你可以自豪地戴上这个标签。这几乎是一种荣誉勋章，对吧？这是耻辱的徽章。毫无疑问，这是一种耻辱的徽章。毫无疑问，尽管他们想要符合我们国家的最佳利益，而我们所有人都想要符合我们国家的最佳利益，但摧毁美国梦的[人才]管道，是的，那是不爱国的，对吧？他们认为他们为我们的国家做正确的事情，但那不爱国。一点也不爱国。所以我们需要继续成为我们伟大的国家，拥有一个伟大国家的信心，对吧？是的。说得好。拥有一个伟大国家的信心，然后有一个想与我们竞争的人，并抱有“放马过来”的态度。对吧？对吧？放马过来。对吧？因为我相信……我相信我们的人民。我相信这里的人民。我相信我们的文化。我相信我们的国家。我相信我们的制度。放马过来。你的看法是总统也是这样吗？他是一个实用主义者。他相信美国的增长和竞争力。在我看来，他就是这样。毫无疑问，特朗普总统是“放马过来”的总统。对吧？对吧？在我看来，他似乎不是……我之所以有信心，并在本次播客中说过，我认为他会与中国达成一项大交易，我真的非常希望如此。是的。而且我认为他以积极的、带着极大的尊重和极大的口才谈论他与中国关系的重要性。我一次都没有听他说过“脱钩”这个词，这是我们在上届政府中经常听到的，对吧？你不能在未来一个世纪最重要（两个最重要的）关系上进行脱钩。这完全没有意义。脱钩恰恰是错误的理念。

合作与竞争的平衡：交易的艺术

Right? I mean, it seems to me he and Scott Besson are saying, "Listen, we need to make America great. We need to re-industrialize America. We need to balance and make sure that we have fair trade. Um that we protect industries that we need to help build that China helps us do that recognizing that we have helped them do it over the course of the last 25 years. But that ultimately he said the best way to understand me is I'm a great dealmaker. I make deals right whereas I think in other camps there are people who are iconoclastic or dogmatic. Uh you know it's the mere shimer view of China that there's a great power struggle. one was must win and one must lose versus this idea the idea that every country has to look exactly like ours, right? You know, and we we want diversity. You want America to win, but that doesn't have to come at the expense of poking an eye and telling somebody else they have to lose because we're that confident. Yeah, we're that confident. Because we're that mighty. Because we're that incredible. I've got no trouble, as you know, I've got no trouble working with all my colleagues in the ecosystem, right? And notice we just did the ultimate deal, right? Partnering with Intel, a company that spent most of its life trying to put us out of business, right? And I had no trouble partnering with them, right? You know, and so, and the reason for that is because number one, bring it on. Yes. And number two, the future is so much greater. It doesn't have to be all us or them. It could be us and them. Yeah. Yeah. But nonetheless, bring it on. Yeah. Agreed.

我的意思是，在我看来，他和斯科特·贝森（Scott Besson）都在说：“听着，我们需要让美国再次伟大。我们需要让美国重新工业化。我们需要平衡并确保我们有公平的贸易。嗯，我们需要保护我们需要帮助建立的产业，中国帮助我们做到这一点，同时认识到我们在过去25年中也帮助了他们。但最终，他说，理解我的最好方式是，我是一个伟大的交易制定者。我做交易，对吧？而在其他阵营中，有些人是反传统或教条的。你知道，他们对中国的看法是存在一场大国斗争，一方必须赢，一方必须输，而不是这个想法——这个想法是每个国家都必须看起来和我们一模一样，对吧？你知道，我们想要多样性。你希望美国获胜，但这不应该以戳人眼睛、告诉别人他们必须输为代价，因为我们足够自信。是的，我们足够自信。因为我们如此强大。因为我们如此不可思议。你知道，我与生态系统中的所有同事合作都没有问题，对吧？请注意，我们刚刚完成了终极交易，对吧？与英特尔合作，这家公司在其生命的大部分时间里都在试图让我们破产，对吧？我与他们合作没有问题，对吧？你知道，原因是：第一，放马过来。是的。第二，未来更广阔。不一定是全有或全无。可以是“我们和他们”。是的。是的。但无论如何，放马过来。是的。同意。

拥抱合作与美国梦的延续

You know, you you you mentioned something that's profoundly important to both of us. You and I have talked a lot about this, the American dream, you know, and it was, I think, Abraham Lincoln who said, "Fundamental to the American dream is the right to rise." Yeah. That's right. The belief that your kids can do better than you did. That's right. Right. And you you've experienced the right to rise. We've all experienced the right to rise in America. So, yeah, you go to Wikipedia, you look up American dreams, my picture, right? Yeah. And the ultimate American dream. And yet we live at this time where because of the nature of these technological systems, we have companies that are going to be worth 10 trillion. We'll probably have individuals that are worth a trillion. Those are the incentives that give people the the encouragement to rise. But at the same time, when we head into this age of abundance, something that I was deeply worried about was that too many people get left behind. Yeah. Right. And they feel left out and left behind. So it makes sense for them to attack this system of capitalism. Something that you and I worked on together and I'm deeply grateful for was the idea of invest America that we have to start every kid at birth on the capitalist right to rise journey give them a thousand bucks in great companies like Nvidia social security and and and open a high etc. Um and they benefit right as the as the comp country wins they win and they own it individually they can see it on their every kid is a shareholder in the future of America of America.

你知道，你……你提到了对我们俩都至关重要的事情。你我谈论了很多关于美国梦的事情，我想是亚伯拉罕·林肯说的：“上升的权利是美国梦的基础。”是的。没错。相信你的孩子可以比你做得更好。没错。对吧。你体验过上升的权利。我们都在美国体验过上升的权利。所以，是的，你去维基百科上查美国梦，我的照片就在那里，对吧？是的。终极的美国梦。然而，我们生活在这样一个时代，由于这些技术系统的性质，我们将出现价值10万亿美元的公司。我们可能还会出现价值一万亿美元的个人。这些是激励人们上升的动力。但同时，当我们进入这个富足的时代时，我非常担心的一件事是，太多人被落下了。是的。对吧？他们感到被排斥和抛弃。所以他们攻击这个资本主义体系是合理的。你我一起努力过的一件事，我深表感激的，就是“投资美国”的想法，即我们必须在每个孩子出生时就让他们踏上资本主义“上升的权利”之旅，给他们1000美元投资于英伟达这样伟大的公司、社会保障，以及……等等。他们受益，对吧？当国家获胜时，他们也获胜，他们个人拥有它，他们可以在他们的……每个孩子都是美国未来的一份子。

Invest America法案的通过

So on the 200 because of your support and I wanted to take the chance on this podcast and the support of Well, I want to thank you for starting it, for driving it. Yeah. Yeah. What a great idea. And you know, so this you're a genius. The Please this passed in the big beautiful bill. Most people don't even realize that yet. Starting in 2026, every child born forever more in the history of this country will start off with an investment account at birth. Yeah. Seen a thousand bucks in the best American companies and your company has agreed to add to the accounts of not only the kids who work for your employees but maybe even kids of others. I'm going to adopt schools, you know, and lots of philanthropists and companies. We think every company across America wonderful way for companies to give back, right? Yeah. as part of the 401k.

所以，因为你的支持，在200[年]……我想借这次播客的机会，感谢你发起了它，推动了它。是的。是的。多么棒的主意。你知道，你是个天才。请注意，这已写入那项宏伟的法案中。大多数人还不知道。从2026年开始，这个国家历史上每个出生的孩子，都将在出生时拥有一个投资账户。是的。将最好的美国公司中的1000美元……你的公司同意向不仅为你员工的孩子，甚至可能是其他人的孩子增加投资。我要收养学校，你知道，还有很多慈善家和公司。我们认为美国每家公司都应该……这是一个很棒的回馈方式，对吧？是的。作为401k的一部分。

应对指数级变化的社会契约演变

This seems to me to be part of the change in the social contract that needs to occur because if if we're seeing this exponential progress, we know that the the evolution of government in the social contract needs to keep up with it. Um obviously President Trump and and bipartisan group in the House and Senate passed this into law. So maybe just talk to us a little bit when you think about the the pace and magnitude of changes that are coming, right? Um I know you believe it will be a net good, but there also going to be a bunch of people displaced along the way. We probably need things like this and other things, right? In order to, you know, bring everybody along for the journey. There's several things that that President Trump has done and let me just start start there has done that is incredibly good for bringing everybody along. The first thing is reindustrializing America. Yeah. President Trump, Secretary Lutnik, the you know they're all in behind that all the work that they're doing encouraging companies to come build here in the United States, investing in factories and uh reskilling and upskilling that skilled labor workforce, right? Incredibly valuable to our country. um the idea that that we no longer uh uh uh make it only that you get a PhD or you go to you know one of the great schools and only in that way can you build a great life right and and deserve to have a great living uh we've got to change all that doesn't make any sense we love craft right I love people who make things with their hands and and we're now we're now going to go back and build things build magnificent incredible things I love that.

这似乎是我认为需要发生的社会契约变化的一部分，因为如果我们看到这种指数级的进步，我们就知道政府和社会契约的演变需要跟上这种变化。显然，特朗普总统和众参两院的两党都通过了这项法律。所以也许当你想到即将到来的变化的速度和幅度时，请和我们谈谈，好吗？我知道你相信这将是净积极的，但在这个过程中也会有一批人失业。我们可能需要这样的事情和其他事情，对吧？为了让每个人都跟上这次旅程。特朗普总统做了一些非常好的事情，让我从那里开始，这些事情对让每个人都跟上进程非常有益。第一件事是美国再工业化。是的。特朗普总统、卢特尼克部长，你知道，他们都支持这一点，他们正在做的一切工作都是鼓励公司来美国建厂，投资工厂，以及对技术工人进行再培训和技能提升，对吧？这对我们的国家非常有价值。我们不再只认为你必须获得博士学位或上最好的学校，只有这样你才能过上美好的生活，对吧？并享有丰厚的收入。我们必须改变这一切，这没有意义。我们热爱工艺，对吧？我喜欢那些用手制作东西的人，我们现在要回去建造宏伟、不可思议的东西。我喜欢这一点。

AI作为最大的“均衡器”

Yes that's going to transform form America. There's no question about that. There's a whole there's a whole a whole band of an economy, a whole band of society that that um uh has been largely left behind because we outsourced everything, right? Now, I'm not suggesting we insource everything, right? You know, all the people arguing about, you know, manufacturing tennis shoes and toothpicks, I mean, you know, that's denigrating a perfectly good discussion into some insane level. You know, we we've got to um recognize that re-industrializ reindustrializing America is is just fundamentally going to be transform transformative number one. Number two, and aspirational. Oh, it's fantastic. Elon taking us to Mars, watching spaceships caught with, you know, uh chopsticks out of the sky. This is not only great for the industrializing base of America, it's aspirational for Fantastic. That's right. And then and then, uh, of course AI. Yeah, it is the greatest equalizer. Just think everybody can have an AI now. The the ultimate equalizer. We've closed the technology divide. Remember the last time that somebody had to learn has wants to use a computer uh for their economic um or career benefit. They have to learn C++ or C or at least Python. Now they just have to learn human, you know. And so, and if you don't know how to program an AI, you tell the AI, "Hi, I don't know how to program an AI. How do I program an AI?" And the AI explains it to you or does it for you. It does it for you. And so, it's incredible, isn't that right? It's And we've now closed the technology divide with technology.

这将改变美国的面貌。毫无疑问。有一个完整的经济、一个完整的社会阶层，因为我们把所有东西都外包了，对吧？现在，我并不是说我们要把所有东西都内包回来，对吧？你知道，那些争论制造网球鞋和牙签的人，我的意思是，你知道，这是把一个完全合理的讨论贬低到某种疯狂的水平。你知道，我们必须认识到，美国再工业化从根本上将是变革性的，这是第一点。第二点，是鼓舞人心的。哦，太棒了。埃隆带我们去火星，看着宇宙飞船像用筷子一样从天空中接住东西。这不仅对美国的工业基础有利，也对……太棒了。没错。然后，当然还有AI。是的，它是最伟大的均衡器。想想看，每个人现在都可以拥有一个AI。终极的均衡器。我们已经消除了技术鸿沟。还记得上一次有人想为了他们的经济或职业利益而学习使用计算机时，他们必须学习C++或C，或者至少是Python。现在他们只需要学习如何用人类的语言沟通，你知道。所以，如果你不知道如何编程AI，你就告诉AI：“嗨，我不知道如何编程AI。我该如何编程AI？”然后AI向你解释，或者替你做。它替你做了。所以，这太不可思议了，不是吗？我们现在用技术消除了技术鸿沟。

AI用户与就业的未来

Yeah. This is something that every everybody's got to engage. You know, OpenAI has 800 million active users. Um gosh, it really really needs to be 6 billion. Yeah. Right. It really needs to be 8 billion soon. And so so I think that that that's number one. Then number two, uh and then number three, you know, I think the the um AI will change tasks. Yeah. The thing that people confuse is there are many tasks that will be eliminated. There are many tasks that will actually be created. But it is very likely that for many people their jobs are gainfully protected. Right. And so, for example, I'm using AI all the time. You're using AI all the time. My analysts are using AI all the time. My engineers, every one of them use AI continuously. And we're hiring more engineers. We're hiring more people. We're hiring across the board. The reason for that is because we have more ideas. Yes, we can now go pursue more ideas. The reason for that is because our company became more productive. And because we became more productive, we became more rich. We became more rich, we can hire more people to go after those ideas, right? The I the concept that that AI comes along and therefore there's going to be a mass destruction of jobs starts with the I starts with the premise that we have no more ideas. Right. It starts with the premise we have nothing left to do. Everything we're doing in our lives today. Yeah. This is the end. Yeah. And if somebody else were to do that one task for me, I have one task left. Now I have to sit there and wait for something. Yes. You know, wait for retirement, sit on my rocking chair. That idea doesn't make sense to me.

是的。这是每个人都必须参与的事情。你知道，OpenAI有8亿活跃用户。嗯，天哪，它真的需要达到60亿。是的。对吧？它很快就需要达到80亿。所以，我想这是第一点。然后是第二点，然后是第三点，我想AI会改变任务。人们混淆的一点是，有很多任务将被消除，也有很多任务实际上将被创造出来。但对于许多人来说，他们的工作很可能得到了有益的保护。对吧？例如，我一直在使用AI。你一直在使用AI。我的分析师一直在使用AI。我的工程师，每一个都在持续使用AI。我们正在招聘更多的工程师。我们正在招聘更多的人。我们正在全面招聘。原因是：我们有了更多的想法。是的，我们现在可以去追求更多的想法了。原因是我们的公司变得更有效率了。因为我们变得更有效率了，我们变得更富有。我们变得更富有，我们可以雇佣更多的人去追求那些想法，对吧？AI的出现，因此将导致大规模失业的观念，其前提是我们再也没有更多的想法了。对吧？其前提是我们没有什么可做的了。我们今天生活中所做的一切。是的。这就是终点。是的。如果别人帮我完成了那一个任务，我只剩下一个任务了。我现在必须坐着等待一些事情。是的。你知道，等待退休，坐在我的摇椅上。这个想法对我来说没有意义。

智能非零和博弈：创造更多工作机会

And so I think I think that intelligence is not a zero sum game. The more intelligent people I'm surrounded by, the more geniuses I'm surrounded by, surprisingly, the more ideas I have, the more problems I imagine that we can go solve, the more work we create, the more jobs we create. And so I think for for um I don't know what the world looks like in a million years that's going to be left for my my children. Um but for the next several decades, my sense is that economy is going to grow. Lots of new jobs are going to be created. Every job will be changed. Some job some jobs will be lost. And um uh we're not going to be, you know, riding horses on streets and those things it'll be fine. You know, humans are are famously skeptical and terrible at understanding compounding systems and they're even worse at understanding exponential systems that accelerate with size.

所以我想，我认为智能不是零和博弈。我身边越是聪明的人，我身边越是天才，出乎意料的是，我拥有的想法就越多，我想我们能解决的问题就越多，我们创造的工作就越多，我们创造的就业机会就越多。所以我想，对于……我不知道一百万年后留给我子孙后代的世界会是什么样子。嗯，但在接下来的几十年里，我的感觉是经济将会增长。会创造很多新的工作。每一份工作都会改变。一些工作会失去。嗯，我们不会，你知道，在街上骑马，那些事情会没事的。你知道，人类以怀疑著称，而且非常不擅长理解复利系统，他们更不擅长理解随规模而加速的指数系统。

应对指数级进展的唯一方法

We've talked about exponentials a lot today. You know, the great futurist Ray Kerszswhile said in the 21st century, we're not going to have a hundred years of progress. We're likely to have 20,000 years of progress. Right. Right. You said earlier, we're so fortunate to be living at this moment and contributing to this moment. I'm not going to ask you to look out 10 or 20 or 30 years because I think it's so challenging. But when we think about 2030 things like robots, 30 years is easier than 2030. Oh, really? Yeah. Yeah. Okay. So, I'll get I I'll grant you license to go out 30. as you think out over the course of I like these shorter time frames because they have to marry bits and atoms bits and Adams the hard part of building this stuff right because everybody's saying it's going to happen is interesting but not helpful exactly but if we have 20,000 years of progress reflect on that statement by Ray reflect on exponentials and how all of our listeners whether you work in government whether you're in a startup whether you're you know running a big company need to be thinking about the accelerating rate of change, the accelerating rate of growth and how you will, you know, be co-intelligent in this new world.

我们今天谈了很多关于指数级增长的话题。你知道，伟大的未来学家雷·库兹韦尔（Ray Kurzweil）说，在21世纪，我们不会有100年的进步。我们很可能会有20000年的进步。对吧？对吧？你刚才说，我们非常幸运能生活在这一刻并为此做出贡献。我不会要求你展望10年、20年或30年，因为我认为这太有挑战性了。但当我们想到2030年的机器人等事物时，30年比2030年更容易。哦，真的吗？是的。是的。好的。所以，我会给你许可，让你展望30年。你思考的时间跨度……我喜欢这些较短的时间框架，因为它们必须将比特和原子结合起来。构建这些东西的困难部分，对吧？因为每个人都在说它会发生，这很有趣，但没有帮助。确切地说，但如果我们有20000年的进步，请思考雷的这句话，思考指数级增长，以及我们的所有听众——无论你在政府工作，无论你在初创公司，无论你……你知道，经营一家大公司——都需要思考变化速度和增长速度的加快，以及你将如何在新的世界中实现“协同智能”。

未来可见的AI应用：机器人与数字孪生

Well, there there are a lot of things that that that many people have already said and and um and they're all they're all very sensible. I think the the um uh in the next 5 years, one of the things that that is really cool that's going to get solved is the fusion of artificial intelligence and megatronics, robotics. And so we're going to have we're going to have, you know, AIS that are that are going to be wandering around us. And uh we all and that everybody knows uh we all all know that we're going to all grow up with our own R2-D2. Yeah. And that R2-D2 remember everything about us and coach us along the way and be our companion. We already know that. Okay. and and so the idea and and and the idea that every human will have their own GPUs associated with them in the cloud um and that they're 8 billion people 8 billion GPUs that's a you know viable outcome. Yeah. you know, and so and each having their own model that's fine-tuned for them, fine-tuned for them. And they that that AI is in the cloud is also embodied in a whole bunch of it's embodied in your car. It's embodied in your own robot. It's everywhere with you. And so that I think that that future is a very sensible thing. the idea that uh we're we're going to understand the the infinite complexity of biology and understanding the system of biology uh and how it how how to predict it and have digital twins of every everybody um our own digital twin for health care like we have a digital twin for shopping at Amazon why wouldn't we have our digital twin at health care of course we would and so uh you know a dig a system that that predicts how our how we're going to age what disease will likely going to have and uh anything that's that's about to happen maybe even next week or you know tomorrow afternoon and predict it early. Of course, we're going to have all that.

嗯，很多人已经说了很多事情，而且……而且它们都是非常合理的。我认为在未来五年内，一件非常酷的事情将被解决，那就是人工智能和机械电子学、机器人技术的融合。所以我们将拥有……我们将拥有，你知道，在我们周围游荡的AI。而且我们都知道，我们将都伴随着我们自己的R2-D2一起长大。是的。而那个R2-D2会记住关于我们的一切，一路指导我们，成为我们的伙伴。我们已经知道了。好的。那么，每个人都将在云端拥有与自己关联的GPU的想法，而且有80亿人，就有80亿个GPU，这是一个可行的结果。是的。你知道，所以，每个人都有为他们量身定制的模型，为他们量身定制。而那个云端的AI也体现在很多方面，它体现在你的车里。它体现在你自己的机器人身上。它无处不在。所以我认为那个未来是非常合理的。那个关于我们将理解生物学的无限复杂性，理解生物学系统，以及如何预测它，拥有每个人的数字孪生……我们自己的医疗保健数字孪生，就像我们有亚马逊购物的数字孪生一样，我们当然也会有医疗保健的数字孪生。所以，一个可以预测我们的衰老方式、可能患上什么疾病，以及接下来可能发生什么事情的系统，比如下周或明天下午，并提前预测它。当然，我们会有所有这些。

抓住列车：立即参与的重要性

And so I think all of that is a given. Um I think the the the the part that that I'm asked a lot by uh CEOs that I work with about now given all of that, what happens? What do you do? And this is this is the this is a common sense of of things that move fast, right? If you if you have a if you have a train that's about to get faster and faster and go exponential, the only thing that you really need to do is get on it. Yeah. And once you get on it, you'll figure everything else out along the way. Right. And so to predict where that train's going to be, right, and try to shoot a bullet at it. Yeah. or predict where that train's going to be and it's going exponentially faster every second and go figure out what intersection to wait for it, right? That's impossible. Just get on it while it's going kind of slowly, right? And go exponential along the way.

所以我想所有这些都是理所当然的。嗯，我认为我经常被我合作的首席执行官问到的部分是：现在有了所有这些，会发生什么？你该怎么做？这是……这是常识，关于快速发展的事物，对吧？如果你有一列火车，它将变得越来越快，并呈指数级发展，你真正需要做的唯一一件事就是登上它。是的。一旦你登上了它，你就会在路上解决所有其他问题。对吧？所以去预测那列火车会在哪里，对吧？然后试图向它射击一颗子弹。是的。或者预测那列火车会在哪里，而它每秒都在呈指数级加速，然后去弄清楚在哪个路口等待它，对吧？那是不可能的。趁它还在慢慢行驶时就登上它，对吧？然后一路呈指数级增长。

历史先驱与对未来的展望

A lot of people think this just happened overnight. You know, you've been you've been at this for 35 years. I remember hearing Larry Pageige say probably around 2005 or 2006 that the end state of Google will be when the machine can predict uh the question before you even answer it before you even ask it and give you the answer without having to look right. I heard Bill Gates say in 2016 because contextually you must be asking about you must be wondering about that. I heard Bill Gates say in 2016 when somebody said hasn't all the things been done we've had the internet we've had cloud we've had mobile social etc. He said, "We haven't even started." He said, "What do you think? Why would you say that?" He said, "We won't even begin until machines go from being dumb calculators to beginning to think for themselves, to think with us, right?" Um, kind of that is the moment, right, that we're in. I think to have leaders like you, leaders like Sam and Elon, Satcha, etc., it's such an extraordinary advantage for this country, right? and to have the cooperation that we see between a system of risk capital that I take, you know, that I'm part of which can provide the risk capital for people to do. We don't we're not relying on government having a Manhattan project. We can actually do this ourselves and together for the benefit of the country. It's an extraordinary time and at a scale that's unimaginable.

很多人认为这一切是一夜之间发生的。你知道，你已经在这个领域工作了35年。我记得拉里·佩奇（Larry Page）在2005年或2006年左右说过，谷歌的最终状态将是机器能够在你回答问题之前，甚至在你提问之前就能预测出问题并给你答案，而无需查找，对吧？我记得2016年比尔·盖茨说过……因为从背景来看，你一定是在问……你一定是在想……我记得比尔·盖茨在2016年说，当有人问他，难道所有的事情都已经做完了吗？我们有了互联网，有了云，有了移动社交等等。他说：“我们才刚刚开始。”他说：“你觉得呢？你为什么这么说？”他说：“直到机器开始自己思考，与我们一起思考，而不是做愚蠢的计算器时，我们才会开始，对吧？”嗯，我们现在所处的正是那个时刻。我认为拥有像你、萨姆、埃隆、萨蒂亚等领导人，对这个国家来说是一个非凡的优势，对吧？而且我们看到了风险资本体系——我参与其中，可以为人们提供风险资本——与我们之间的合作。我们不是依赖政府拥有曼哈顿计划。我们可以自己一起为国家的利益做这件事。这是一个非凡的时代，其规模是难以想象的。

责任感与包容性增长

Right. Right. It's an extraordinary time. But I also think, you know, one of the things that I'm just grateful is that we have leaders who also understand their responsibility to the fact that we are creating change at an accelerating rate. And we know while it will most likely be great for the vast majority, there'll be challenges along the way. And we'll deal with those as they come and raise the floor for everybody and make sure that this is an a win, not just for some elite bureaucrats at the top hanging out in Silicon Valley. And don't scare them. Bring them along. Don't scare them. Bring them along. And we will. Yeah. So, thank you for that. Exactly.

对。对。这是一个非凡的时代。但我还认为，你知道，我所感激的一件事是，我们有领导人也明白他们的责任，我们正在以加速的速度创造变革。我们知道，虽然这对绝大多数人来说很可能是好事，但一路上也会有挑战。我们会应对这些挑战，为每个人提高底线，确保这是一场胜利，而不仅仅是为那些在硅谷闲逛的顶层精英官僚带来的胜利。不要吓到他们。把他们带上。不要吓到他们。把他们带上。我们会做到的。是的。所以，谢谢你。完全正确。

📌 文中提及的人物和组织

公司/组织: openai

媒体/书籍: bg2-pod

关键字: llm 计算扩展定律