英伟达 CEO 黄仁勋：为何我不认同 AI 末日警告

Jensen，感谢你今天接受采访。

Jensen, thank you for doing this today.

我很高兴。很高兴见到你。

I'm happy to do it. Nice to see you.

也很高兴见到你。感谢你的时间。我的第一个问题，我们听到一些顶级 CEO，如 Dario Amodei、Sam Altman、Elon Musk 说我们需要放缓 AI 的发展。还有举报人称人类可能在 10 年内终结。你同意这种说法吗？

It's nice to see you, too. Thank you for your time. My first question, we've heard from some of the top CEOs, Dario Amodei, Sam Altman, Elon Musk saying that we need to pace the development of AI. You've got whistleblowers saying that the end of humanity could be within 10 years. Do you agree with this?

嗯，首先，AI 技术极其强大。它当然可以带来巨大的好处，也可能被用于可怕的应用。因此，我们必须确保在开发这项技术时，安全地进行开发。整个行业正在从开发变得有能力的 AI 技术转型，也就是说，今年上半年我们经历了一个转型，AI 变得有用了，而且需求惊人。这些 AI 产品的 token 生成现在是有利可图的，而且利润丰厚。许多公司都在采用它。OpenAI 最新的 Astra 非常出色。我们到处都在使用它。我们用 Claude Code，我们用 Cursor，SpaceX AI 的产品，我们用 Cognition。所以这项技术的能力现在已经达到了极其有用的状态。因此，这些公司正在从试图让技术运转的实验室转变为产品公司。事实上，即使称其为实验室也说明了一些问题。他们在前 10-15 年处于研究阶段，而现在突然在过去的六个月里，整个行业几乎都转向了高产量生产模式。所以我认为，所有正在发生的反应，其核心都是这一转型的一部分。

Well, first of all, AI technology is incredibly powerful. It could of course do incredible good and it could be applied to terrible applications as well. And so we have to make sure that when we're developing this technology, we develop it safely. And the industry is transitioning from developing an AI technology that was becoming capable, meaning that this first part of this year we went through a transition where AI became useful and it became useful, the demand for it is incredible. Token generation of these AI products are now profitable and actually incredibly profitable. So many companies are adopting it. OpenAI's latest Astra is incredible. We're using it all over the place. We use Claude Code. We use Cursor. SpaceX AI's product. We use Cognition. And so the capability of the technology has now reached extremely useful state. And so these companies are transitioning from being a lab trying to get the technology to work to now a products company. And so in fact even calling it a lab kind of tells you something. They were in research for the first 10-15 years and now all of a sudden in the last six months practically the industry has shifted to high production ramp mode. And so I think that all of the reactions that are happening are part of that transition at its core.

这些评估正确吗？用一些人的话说，我们都会在十年内消失吗？

Are they correct assessments of what could happen? Are we all going to be, in the words of some, gone in a decade?

不，那完全是错误的。那种叙事很戏剧化，能吸引很多关注。当然，它显然引起了媒体的注意。但我认为，其核心是他们深切关心产品是否被安全地构建。

No, that's completely false. And that kind of narrative is dramatic. It gets a lot of attention. Of course, obviously, it's gotten the attention of the press. And I think at its core, they care deeply that the products are built safely.

他们正在从实验室转向工程。他们必须投入更多的研究人员、更多的算力用于安全。

They're transitioning from a lab to engineering. They have to dedicate a lot more of their researchers, a lot more of their compute to safety.

在开发产品的情况下，有引擎部分、设计部分、能力部分，还有验证部分。在任何行业的初期，大部分工程都用于让产品运转起来。然后随着产品成熟并推向社会，大部分工程实际上用于验证、测试、评估、基准测试、符合标准，确保他们交付的产品可靠、按预期运行、兑现产品的承诺。因此，在当前这个转型期，他们必须将大量能力转向 AI 安全，这是非常明智的。我认为这才是关键所在。所有围绕戏剧性的信息都是不必要的。坦率地说，太多戏剧性了，不负责任。它现在不是基于事实，也不是基于科学。

That's in the case of developing products, there's the engine, there's the design part of it, the capabilities part of it, and there's the verification part of it. In the beginning of any industry, most of the engineering goes towards making the product work. And then as the product becomes mature and it rolls out into society, most of the engineering is actually dedicated to verification, testing, evaluation, verification, benchmarking, attesting to standards, making sure that the product that they ship is reliable, it's functioning as expected, delivers on the promise of the product. And so it's very sensible that right now during this transition they have to move a lot of their capabilities to AI safety. I think that that's really where the crux of it. Anything, all the messaging around the drama of it is unnecessary. It's quite frankly too much drama, irresponsible. It is now grounded on facts. It's not grounded on science.

所以这是错误的。

So it's wrong.

嗯，这是错误的，因为它不是基于科学。他们的担忧没有错。我只是想确保，其核心的担忧没有错。这和我刚下飞机没什么不同。我真的很高兴整个航空旅行的生态系统和行业都认真对待安全，所有那些昼夜工作的人，从事航空安全、飞机安全和航空旅行安全的人比制造新型飞机的人还多。所以这是合理的，我们正在经历同样的事情。

Well, it's wrong because it's not grounded on science. Their concern is not wrong. I just want to make sure that at the core of it, the concern is not wrong. This is no different than I just landed from a plane. I'm really glad that the entire ecosystem and industry of air travel takes safety seriously and all the people who are working around the clock, they're more people working on airline safety and aircraft safety and air travel safety than they are working on making new types of aircrafts. And so it makes sense and we're just going through the same thing.

你说过我们不需要任何新法律。我们不需要任何新法规。但你刚才说支持主要行业的保障措施。为什么不支持为 AI 设立相当于 FDA 或 FAA 的机构？

You've said we don't need any new laws. We don't need any new regulations. But you just said here that you support safeguards in major industries. Why not support the equivalent of an FDA or of FAA for AI?

嗯，在我们提出新法律和新法规之前，让我们先应用现有的法律和法规。有各种各样的法律和法规。你知道我们生活在法治之地，这就是为什么我们感到安全、有保障、受到治理，我们有各种各样的法律和法规应用于每一个应用，无论是食品还是航空旅行或任何

Well, before we come up with new laws and new regulations, let's apply the current laws and the current regulations. There are all kinds of laws and regulations. You know we live in the land of laws and is the reason why we feel safe and secure and governed and we have all kinds of laws and regulations applied to each and every one of those applications whether it's food or air travel or any

特别是针对 AI。

specifically for AI.

是的。那么缺失的部分是什么？告诉我缺失的部分。我很好奇你的观点，为什么你认为没有缺失的部分，当像 Dario Amodei 或 Sam Altman 这样的人说我们希望看到监管时。

Yeah. So what's the missing part? Tell me the missing part. I'm curious to know your opinion why you think there are no missing parts when someone like a Dario Amodei or a Sam Altman is saying we'd like to see regulation.

我认为我回应的是，在两个实验室的情况下，一个实验室发生了四起事件，网络安全事件由他们触发。另一个刚刚又发生了几起。所以在这两起事件之前，有大量的法律可以处理。你有网络安全未经授权进入的法律。你有损害赔偿责任。你有各种与网络安全相关的责任。你有各种与产品相关的责任。所以如果你的产品造成伤害，如果你的产品没有按承诺运行，如果你与公司签订了服务级别协议合同，有大量的法律与之相关。我的观点是，先应用这些法律，不要让这种末日叙事导致某人免除他们现有的法律。现在，去字里行间读一读。实际上，他们并不是在要求更多法律。他们是在要求免除我们现有的法律。我认为这是一个问题。我们有大量的法律。所以，让我们先应用这些。

I think what I was responding to is in the case of both of the labs, one lab had four incidents, cyber security incidents triggered by them. Another one just had a couple more. And so before, in both of those incidents has plenty of laws to deal with. You have cyber security unauthorized entry laws. You have damage liabilities. You have all kinds of liabilities associated with cyber security. You have all kinds of liabilities associated with products. And so if your product does harm, if your product doesn't perform as you promised, and if you entered into contracts with companies for service level agreements, there are plenty of laws associated with that. My point is apply those first and don't let this doomsday narrative cause somebody to relieve them of the laws that currently exist. Now, go and read between the lines. There's a lot of, they're actually not asking for more laws. They're asking to be relieved of the laws we do have. And I think that that's a problem. We have plenty of laws. So, let's apply those first.

那么，你认为这些公司要求被监管，是因为他们不想自己处理这些问题，不想为自己正在构建的东西承担责任吗？

So, do you think then these companies are asking to be regulated in a sense because they don't want to have to handle this themselves and take their own responsibility for what they're building?

嗯，他们正在从实验室转变为产品公司，从非客户收入业务转变为客户收入业务。这一转型必定非同寻常，因为他们做得如此之快。而这是历史上最重要的两家科技公司。那里有极其才华横溢的研究人员。他们做着出色的工作，但这并不能改变事实。事实是他们正在从实验室转变为产品服务公司，产生数百亿美元的收入，并期望达到数千亿美元的收入。

Well, they're moving from a lab to a products company, from a non-customer revenue producing to a customer revenue producing business. And that transition must be extraordinary because they're doing it so fast. And these are two of the most consequential technology companies in history. They're incredibly talented researchers there. They're doing excellent work, but it doesn't change the fact. The fact is they're transitioning from a laboratory into a product service company generating tens of billions of dollars of revenues, expecting it to be hundreds of billions of dollars of revenues.

两家公司都在以人类历史上任何公司都未曾见过的速度扩充算力。这理所应当，因为他们的产品好、有用，人们想要更多，他们想创造更多营收，投资者想看到增长。这一切都是好事。但法律是存在的。有产品责任法，这些产品责任法应当针对当前的事件进行检验。然后我们再看接下来会发生什么。到时候再看。

Both of them increasing their compute capacity at a rate no companies have ever seen in human history. And deservedly so, because their products are good and they're useful and people want more of it, and they like to generate more revenues and their investors want to see their growth. And all of that is good. But there are laws. There are product liability laws, and those product liability laws ought to be tested against the current incidents. And then we'll see what happens next. Then we'll see.

对。对。对。对。先做显而易见的事。先做实际的事，然后我们之后再谈理论。但不要不做实际的事就空谈理论。

Yeah. Yeah. Yeah. Yeah. Do the obvious first. Do the practical first, and then we can talk about the theoretical later. But don't talk about the theoretical without doing the practical.

在提出不需要额外政府监管的论点时，我很好奇。

In making the argument for no additional government regulation, I'm curious to know.

所以我是主张运用现有的（法律）。

So I'm arguing for exercising the current.

对。但你是说，你目前不主张增加政府监管。

Right. So but you are saying you're not, you don't advocate at this moment for additional government regulation.

监管来做什么？

To do what?

对。所以如果那是前提的话。

Right. So if that's the premise.

直到我们知道缺什么。直到我们知道缺什么，为什么要增加更多法律？

Until we know what's missing. Until we know what's missing, why add more laws?

那么说到这个，我很好奇，比起额外的监管，你是否更信任掌控这项技术的四五位 CEO？你相信他们会做正确的事，并缓解这里的一些担忧吗？

So on that note, curious, do you have more faith in the four or five CEOs who control this technology than in additional regulation? Do you trust them to do the right thing and to allay some of the fears here?

他们当然有能力做正确的事。他们正在经历的事情，他们正在从什么过渡到什么，是可以理解的。首先，他们都是年轻的 CEO，第一次当 CEO。我做这行已经 33 年了，我能受益于后见之明，坐在这里的后座上对他们的业务夸夸其谈。但事实是，他们正在经历这个过渡。建立公司、建立组织、建立产品、建立行业、建立业务、上市，都不容易。所有这些都是新的，而且同时发生。

They're surely capable of doing the right things. What they're going through, what they're transitioning from and to, is an understandable thing. First of all, they're all young CEOs, first-time CEOs. I've been doing this for 33 years, and I get the benefit of hindsight and to sit here in the back seat and pontificate about their business. But the fact of the matter is they're going through this transition. It's not easy building a company, building organizations, building products, building industries, building a business, going public. All of that is new and it's all happening at the same time.

所以我对他们正在应对的巨大挑战感同身受。我也对他们团队的能力非常有信心。这些人非常出色。

And so I'm empathetic to the enormity of the challenges that they're dealing with. And I also have a lot of confidence in the capabilities of the teams that they have. These are extraordinary people.

所以你信任，你信任。

So you trust, you trust.

我相信他们。我相信这些人。我也对他们的处境感同身受。我不赞同用世界末日之类的说法来转移注意力和分散精力。我更希望他们专注于我知道他们正在专注的事情。他们的工程师完全能够理解，他们遇到的事件要求他们为实验创建更安全的沙箱、隔离和遏制措施。我确信他们已经认识到，需要更密切地监控所有软件在测试和评估期间的运行。我确信，如果他们的软件要与外部云和其他服务交互，他们应该让人们知道；如果发生事件，要确保公开分享，以便整个行业都能从中学习，我们都能改进自己。网络安全行业就是这样运作的。

I believe in them. I believe in the people. I also am empathetic of the circumstances. I don't approve of deflections and distractions about the end of the world. And I'd rather them focus on the things that I know they're focused on. Their engineers are absolutely able to understand that the incidents they had require them to create more secure sandboxes and isolation and containment of their experiments. I am certain that they have learned that they need to monitor much more closely all of their software as it's operating during testing and eval. I am certain that if their software were to interact with external clouds and other services, that they should let people know, and if incidents were to happen, to make sure that they share it publicly so that the whole industry can learn from it and we can all improve ourselves. This is how the cyber security industry works.

你知道，就在此时此刻，全世界都有网络攻击，对吧？全世界的工程师都在拼命修补软件、改进软件、修复漏洞。这些是极其复杂的系统。我们拥有开放软件和开源软件，让所有公司都能自我防御和升级软件，这是这个行业能够运作的原因之一，而且就在我们说话这会儿它正在运作，我们互相照应。有数百家公司作为一个社区一起合作。我们拥有不对称的优势，即防御者众多而攻击者少。所以关于他们遇到的事件有很多东西可以剖析，但这并不改变一个事实：他们目前正在做的是——他们回归到以工程为基础。他们会改进工程。他们会在评估、基准测试和测试上投入更多。这对他们来说是一个好的转变。

You know, I just so you know, right now all over the world there's cyber attacks, right? And there's all over the world engineers are working feverishly to patch their software, improve their software, improve vulnerability. These are incredibly complex systems. The fact that we have open software and we have open-source software so that all the companies could defend themselves and upgrade their software is one of the reasons why the industry works, and it's working right now as we speak, and we watch each other's back. There are hundreds of companies that work together as a community. And we have the asymmetric benefit of having a lot of defenders against attackers. And so there's a lot of stuff to unpack about the incidents they had, but it doesn't change the fact that they're currently doing—they're back to grounding themselves in engineering. They're going to improve their engineering. They're going to invest a lot more in evaluations and benchmarking and testing, and that's just a good transition for them.

我很好奇，你知道，当美国人看着你，你是英伟达的 CEO 和创始人，这家极其——世界上最有价值的公司。为什么在这个特定的安全问题上，为什么美国人应该信任你的观点，而你公司的成功、你的净资产都直接与销售芯片挂钩？

I'm curious, you know, as Americans look at you, you're the CEO and founder of Nvidia, this incredibly—the most valuable company in the world. Why should, on this specific issue of safety, why should Americans trust your perspective when your company's success, your net worth is directly tied to selling chips?

我们公司的成功与产品和服务的安全部署直接相关。如果计算机行业不能安全、高效地为世界提供价值——像我们迄今为止那样提供巨大的积极影响——如果我们不继续这样做，我们的价值就会减损。因此，让 AI 这一新类别的应用能够安全地生产并交付给世界，符合我们个人和行业的利益。我们必须这样做。我们必须平衡这种负责任的乐观主义。我们之所以从事这个职业，是因为我们相信它对世界有益。我们相信我们希望自己的孩子参与其中。我的孩子就是其中一部分。我们希望我们所有的家庭都参与其中。我们希望这样一个未来：人类苦难更少，公司生产力更高，我们能更快地发现知识。我们希望未来拥有这一切。我们希望应对人口老龄化。我们希望拥有可持续能源，并利用 AI 的市场力量推动世界走向可持续能源。我们希望所有这些，但我们必须负责任地、安全地去做。我们必须成熟地去做这一切，以免吓到别人——用戏剧性和历史性的未来预测吓唬人们。那根本不是真的。

Our company's success is directly connected to the safe deployment of products and services. If the computer industry doesn't provide value to the world safely and productively—provide enormously positive impact as we have so far—if we don't continue to do that, our value would be diminished. And so it is in our very personal and industrial benefit that this new class of applications called AI is able to be produced and delivered to the world safely. We have to do this. We have to balance this responsible optimism. The reason why we do what we do for a living is because we believe it's great for the world. We believe that we want our children to be part of it. My children are part of it. We want all of our families to be part of it. We want this future where we have less human suffering, where companies could be more productive, where we could discover knowledge more quickly. We want all of this in the future. We want to deal with aging populations. We want to have sustainable energy and use AI market forces to drive the world into sustainable energy. We want all of these things, but we have to do it responsibly and we have to do it safely. And we have to do this all with maturity so that we don't scare other people—scare people with dramatic and historical predictions of the future. That is just not true.

这里显而易见的问题是与中国竞赛。一直有一种论调说，我们不能放慢 AI 进展，否则就会输给中国。你认为这个论点成立吗？

The elephant in the room here is the race with China. The argument has been we can't slow down AI progress lest we lose to China. Do you think that's a valid argument?

作为对每个行业的每一位 CEO、每一位学生、每一位教授、美国社会每一位成员的建议，我们希望美国尽可能繁荣。我们越繁荣，作为经济体越富裕，我们就能为社会保障提供更多，为国家军事安全提供更多，我们就能更好地成为世界公民，支持其他人的挑战、考验、磨难、痛苦和苦难，我们作为一个国家就能更加慷慨。所以我建议每一位美国公民，美国的每一个人，都朝着同一个方向竭尽全力。

As an advice to every single CEO in every industry, every student, every professor, every member of the American society, we want America to be as prosperous as we can. The more prosperous we are, the richer we are as an economy, the more we could provide for social security, the more we could provide for national military security, the better we could be a citizen of the world and support other people's challenges and trials and tribulation and pain and suffering, the more philanthropic we can be as a nation. And so I would advise that every single American citizen, everybody in America pulls in the same direction as hard as we can.

那么我们应该放慢速度吗？这会影响与中国的竞争吗？美国的地位……

So should we slow down and would that affect the race with China? The American standing in...

我们应该尽可能快地前进，不管别人怎么样。这不是……如果中国发现了治愈癌症的方法，对我们来说是好事。如果中国发明了可持续能源的新技术，对我们来说是好事。如果中国在科学和数学的许多领域领先，我们能够向他们学习并从中受益，对我们来说是好事。但听着，我们作为一个国家是强大的。这是一个拥有非凡能力和非凡创新者的国家。我很自豪能成为美国科技产业的一部分，我认为这是我们国家财富的三大支柱之一。当然，我们的金融体系是其中之一。当然，我们的军队是其中之一。你知道，我所服务的、我所参与的产业，科技产业绝对是我们国家财富之一。我们将尽可能快地前进，但……

We should go as fast as we can irrespective of anybody else. It's not... if China were to discover a cure for cancer, it's great for us. If China were to invent a new technology for sustainable energy, it's great for us. If China were to lead in many areas of science and math and we were able to learn from them and benefit from it, it's great for us. But listen, we're packing heat as a nation. This is a nation with extraordinary capability, extraordinary innovators. I'm proud to be part of the American technology industry that I would consider one of the three pillars of our national treasures. Of course, our financial system is one of them. Of course, our military is one of them. You know, the industry that I serve, that I'm part of, the technology industry is absolutely one of our national treasures. We are going to go as fast as we can, but...

这有多重要？

How important is it?

我们将尽可能快地前进，但我们永远不会、也绝不应该在产品准备好之前就发货，交付不安全的产品，没有人期望我们这样做。但每个人都期望我们成功，帮助美国尽可能繁荣。

We're going to go as fast as we can, but we would never ever and never should ship products before they're ready, deliver products that are unsafe, and nobody's expecting us to do that. But everybody's expecting us to succeed and help America be as prosperous as possible.

David Sacks 本周早些时候告诉我，他认为如果我们放慢 AI 发展的速度，那可能会对美国经济、美国军队造成巨大损害，并使美国处于不利地位。所以我只是想清楚地了解，鉴于我们与中国的情况，你认为我们现在应该调整 AI 的发展速度吗？

David Sacks told me earlier this week that he believes that if we slow down the rate of AI development, then that could do huge damage to the American economy, the American military and put the United States at a disadvantage. So I just want to clearly understand like do you think that we should be pacing the development of AI right now given where we are vis-à-vis China?

首先，David 非常聪明。他深思熟虑，有深度，务实。他是伟大的美国爱国者之一。认识他是我的荣幸，我是在他担任白宫顾问时认识他的，他是个了不起的人。我不太确定他说话的上下文是什么，但我相当肯定他希望美国尽可能快地前进，但不要超过我们应该的速度。我们应该尽可能快地前进，但不要超过我们应该的速度。如果我们牺牲安全，那不能发生。如果你牺牲安全，交付不安全的产品，将人们置于危险之中，我们有各种法律来追究你。因此，开发这些技术的公司完全有动力做正确的事情。我完全有信心。我完全有信心，因为他们有动力这样做。他们有能力这样做。我们将交付，我们将创造 AI，以对社会安全的方式推进 AI。

First of all, David is incredibly smart. This is thoughtful. He's deep. He's pragmatic. He is one of the great American patriots. It's been a privilege to know him and I met him when he was an adviser at the White House and incredible human. I'm not exactly what the context of what he said was, but I'm fairly certain he wants America to go as fast as we can, but not faster than we should. We should go as fast as we can, but not faster than we should. And if we're compromising safety, that can't happen. If you're compromising safety and you deliver products that are unsafe, that put people in harm's way, we have laws of all kinds to come after you. And so the companies that are developing these technologies are perfectly incentivized to do the right thing. And I have every confidence. I have every confidence because they're incentivized to do so. They have the capability to do so. That we will deliver, we will create AI, advance AI in a way that's safe to society.

Dario Amodei 说我们应该停止向中国出售芯片。原话是：不要向中国出售强大的 AI 芯片或半导体制造设备。你对此有何回应？

Dario Amodei says we should stop selling chips to China. The quote is do not sell powerful AI chips or semiconductor manufacturing equipment to China. What's your response to that?

嗯，我们必须确保美国创造的技术首先让美国受益。每个国家都会这样做。每个公司都这样做。每个父母都这样做。每个公司都这样做。每个行业都这样做。每个国家都这样做。就美国而言，美国已经决定，美国已经决定，我非常支持我们首先向美国公司提供 Nvidia 最好的技术。所以我们正处于我们称之为 Vera Rubin 的时代。Vera Rubin 正在批量生产。我们在这项技术上领先中国几年，它正在提供给 Dario 和 Sam 以及你知道的美国实验室和美国行业，这正是我们正在做的，我们很高兴这样做。另一方面，记住 AI 是一个五层蛋糕。不仅仅是那个模型。AI 行业由能源部门、芯片部门、模型部门代表，模型层就在那里。你可以说有五层或六层，但上面基本上是数据和应用层。我们需要每一层，美国的每个人都赢。我们希望我们在能源领域获胜。我们希望我们在芯片领域领先世界。我们希望确保每一个应用公司——我喜欢 FedEx 将成为一家 AI 公司。我喜欢沃尔玛将成为一家 AI 公司。我们需要美国的每一家公司都成为 AI 公司。不仅仅是一家公司，不仅仅是两家公司。记住，你必须退一步，记住 AI 是什么。AI 是数字智能。为什么美国每一家公司都没有得到支持并拥有数字智能的能力？所以当然，每家公司都将成为未来的公司，数字智能或人工智能作为公司的一部分。所以我希望每个人都赢，包括芯片行业。所以，

Well, we have to make sure that technology that's created by America advantage America first. Every country would do that. Every company does that. Every parent does that. Every company does that. Every industry does that. Every nation does that. And in the case of America, America has decided, the United States has decided and I'm very supportive that we provide Nvidia's best technology to American companies first. And so we're in the generation of what we call Vera Rubin. Vera Rubin is shipping in volume production. We are several years ahead of China with this technology and it's being made available to Dario and Sam and you know the American labs and American industries and that's exactly what we're doing and we're delighted to do that. On the other hand, remember that AI is a five layer cake. It is not just that model. The AI industry is represented by the energy sector, the chips sector, the model sector where they are, the model layer where they are. You could argue that there's six layers or five, but above that is basically the data and the application layer. And we need every single one of these layers, everybody in America to win. We want us to win in the energy sector. We'd like us to lead the world in the chip sector. We like to make sure that every single application company—I love the fact that FedEx is going to be an AI company. I love the fact that Walmart's going to be an AI company. We need every single company in America to be an AI company. Not just one company, not just two companies. Remember, you got to take a step back and remember what AI is. AI is digital intelligence. Why is it how is it possible that every single company in America is not supported and have capabilities of digital intelligence? And so of course every company is going to be a company of the future with digital intelligence or artificial intelligence as part of the company. And so I want everybody to win including the chips industry. And so,

所以你不同意 Dario 的立场，即我们不应该出售？

So you disagree with Dario's position that we should not sell?

我完全不同意我们应该禁止中国获得所有美国产品。美国公司应该拥有先进的能力。但在此之后，美国 AI 技术栈的每一层，包括传统公司和每个领域的 AI 模型公司以及每家芯片公司，都应该去服务世界，为世界竞争。服务世界，拥有市场——获得市场份额，获得经济繁荣，这都是关于竞争。美国是关于竞争的，我们应该去竞争，而不是放弃整个市场。我们的市场不是他能放弃的。我们的市场不是他能放弃的。我们应该去竞争世界市场。

I completely disagree that we should ban China of all American products. The American company should have advanced capabilities. But then after that, every single layer of the American AI tech stack, including the traditional companies and the AI model companies of every single field and every single chip company should go and serve the world, compete for the world. Serving the world, having market—have capturing market share, capturing economic prosperity, that's all about competing. America is about competition and we ought to go and compete and not cede entire markets. Our market is not his to cede. Our market is not his to cede. We ought to go compete for the world's market.

我想问你关于特朗普总统和你与他的关系。你同意特朗普总统吗？你是他在科技和 AI 方面最亲密的顾问之一。

I want to ask you about President Trump and your relationship with him. Do you agree with President Trump? You're one of his closest advisers in tech and AI.

他有很多顾问。他有很多建议，而且他是一位出色的倾听者。他与广泛的人交谈。

He has a lot of advisers. He has a lot of advice and he's an excellent listener. He has a broad range of people he talks to.

我很好奇

I'm curious

而我——这就是他能看透这些复杂情况的原因之一。

and I—that's one of the reasons why he sees through these complicated situations.

嗯，说到这个，是的，你知道，他本周说他认为对 AI 危险的担忧、恐惧是一个骗局。用 AI 毁灭人类的可能性是一个骗局。你同意特朗普总统的这个说法吗？

Well, speaking of that, yeah, you know, he said this week that he believes that the concerns, the fears about AI's dangers are a hoax. The potential to destroy humanity with AI is a hoax. Do you agree with President Trump on this?

哦，是的。现在你和我刚才的对话正是从这个开始的。我完全不同意。我完全不同意 AI 将在 2030 年毁灭世界——成为世界末日。

Oh, yeah. Now you and I started the conversation just now with exactly this. I completely disagree with that. I completely disagree that AI is going to destroy—is going to be the end of the world in 2030.

你同意吗？总统称这种恐惧为骗局。你同意他的描述吗？

Do you agree? The president is calling this fear a hoax. Do you agree with his characterization of it?

嗯，我只是——我在 All-In 峰会上听你在电话里和他交谈。

Um, I just—I was at the All-In Summit listening to you talk to him on the phone.

嗯，我不想代表他说话。我问的是你。

Well, I don't want to speak for him. I'm asking you about you.

我相信那些关于世界末日的说法在全美煽动恐惧，而且是由那些正在这样做的人做的，这对我来说毫无道理。所以他们一定有不可告人的动机。也许是政治原因，也许是别的，也许只是为了吸引眼球，也许是为了能跟你做一次精彩的采访。我不知道原因是什么，但不管怎么描述，2030 年不会是世界末日。那是世界末日的可能性是 0%。所以我认为吓唬人是不必要的，是不负责任的，我不知道他们的动机是什么。然而，他确实说了一些话和其他一些事情——所以这是他说的一件事。他说的另一件事是，我们不需要更多监管。我们需要执行现有的法规。这是非常实际、非常明智的事情，我们刚才也谈到了，我同意这一点。

I believe the claims of the end of the world stirring fear across America, and doing it by the people who are doing it, makes no sense to me. So they must be doing it for ulterior reasons. Maybe it's political, maybe it's otherwise, maybe just attention grabbing, maybe so that they could get a great interview with you. I don't know what the reason is, but however it's characterized, 2030 is not going to be the end of the world. There is 0% chance that's going to be the end of the world. And so I think that scaring people is unnecessary, it is irresponsible, and I don't know what their motives are. However, what he did say and some of the other things—so that's one thing that he said. The other thing that he said was that we don't need more regulations. We need to apply the current regulations we have. That's a very practical, very sensible thing, and we just spoke about that as well, and I agree with that.

他还说了一些有趣的话，关于他有能力确保 AI 不会对美国人产生这种影响。你同意特朗普总统有能力在任期内抵御和防止 AI 危机吗？你认识他，你跟他谈过。他有这个能力吗？

He also said something interesting about his ability to make sure that AI does not have this impact on Americans. Do you agree that President Trump has the ability to fend off and prevent an AI crisis while he's in office? You know him, you talk to him. Does he have that ability?

作为美国的领导人，凭借我们目前拥有的所有法律和法规，并认识到我们现在面临的事件，我相信我们有必要的法律和法规来处理这些事情。

As the leader of America, with all of the laws and regulations we currently have, and recognizing the incidents that we now have in front of us, I believe we have the necessary laws and regulations to handle those things.

所以在你看来，他能抵御它？

So he can fend it off in your view?

是的，因为美国有——我们的体系已经有很多法律法规，适用于许多不同的产品和服务领域。所以我认为他说的是实际的事情，就像我们之前已经说过的，我认为我们有很多法律法规，它们应该被执行。

Yeah, because America has—our system already has a lot of laws and regulations that apply to many different sectors of products and services. And so I think he's saying practical things, and like we've already spoken earlier, I think we have lots of laws and regulations and they should be applied.

我这里有两个简短的问题。你将出席与中国国家主席习近平的国宴。对吗？下周。

I have two just quick questions for you here. You will be attending the state dinner with President Xi Jinping of China. Is that correct? Next week.

下周。是的。

Next week. Yes.

如果你有机会跟他谈谈我们现在所处的这种情况，美国人非常担心的那种情况，你会对中国主席说什么？

If you get a chance to speak to him about this situation that we're in right now, the one that Americans are so concerned about, what will you say to the president of China?

我会说，我很高兴看到 AI 技术从实验室过渡到现在真正有用的能力和产品及服务。因为它正在经历这种转变，所有开发和推进 AI 的国家都以分享最佳实践的方式来做这件事会很好。也许就产品期望设定全球标准。

I would say that I'm delighted to see AI's technology transitioning from laboratories to now really useful capabilities and products and services. And because it's making that transition, it would be good for all countries that are developing and advancing AI to do so in a way where we share best practices. Maybe set global standards about product expectations.

更紧密地合作。

Work closer together.

更紧密地合作。是的，当然。是的，当然。更紧密地合作。如果我们想——看，他们希望中国安全。他们希望中国受益于 AI。他们希望中国繁荣，就像我们希望美国繁荣一样。我很高兴他们有雄心勃勃的期望和抱负。我希望每个国家——你不希望每个国家都对更美好的明天有抱负和雄心吗？我喜欢他们这一点。随着他们推进技术，我们应该确保尽可能多地相互合作，尽可能多。当然，有些领域我们因某种原因无法合作。也许是军事上——嗯，可能是军事原因。可能是其他国家安全原因。那些是政府要解决的问题。除此之外，我认为我们在每一个可能的方面合作都会很好。

Work closer together. Yeah, of course. Yeah, of course. Work closer together. If we want—look, they want China to be safe. They want China to have the benefit of AI. They want China to prosper just as we want United States to prosper. I'm glad they have ambitious expectations and aspirations. I wish every country—don't you wish every country has aspirations and ambitions for a better tomorrow? And I love that about them. As they advance the technology, we ought to make sure that we work with each other as much as we can, as much as we can. There are, of course, areas where we can't work together for whatever reason. Maybe it's militarily—well, it could be military reasons. It could be other national security reasons that could arise. And those are for the governments to sort out. Other than that, I think that it would be great for us to work in every possible way.

我能问你一些关于你个人状况的问题吗？我的意思是，我如果不问就是失职——我们都知道黄仁勋是英伟达非常成功的 CEO。你是亿万富翁。你说过你在加州的一些亿万富翁同行正在花费数百万美元来反对第 40 号提案，也就是所谓的亿万富翁税。你说过如果加州通过它，你会完全愿意支付。为什么你与这么多同行的看法不同？

Can I ask you about something about your personal status here? I mean, I would be remiss to—we all know that Jensen Huang is the very successful CEO of Nvidia. You are a billionaire. You have said some of your fellow billionaires here in California are spending millions to fight Proposition 40, the billionaire tax as it were. You have said that if California passes it, you'd be perfectly fine paying it. Why do you see it differently than so many of your peers?

我不介意交税。

I don't mind paying taxes.

为什么不呢？

Why not?

看，我不喜欢税款被浪费。我喜欢我们的国家高效运转。但我也支持这个国家所代表的一切。我支持我们所有的政府官员履行他们的公民职责，让美国变得伟大和更好。我喜欢我们为所有公民提供的基础设施。我喜欢它所创造的东西。对我来说，你知道，这就是——我是终极的美国梦。如果不是因为这个国家，教育，创建新公司的能力，追求你的梦想，让功绩成为你成功的原因——你的努力、你的毅力、你的创新，让这些成为你成功的原因。如果不是因为这个，你知道，我为什么在这里。我不会遇到所有让英伟达成为可能的人。我不会在学校遇到我的妻子。我的意思是，名单还在继续。我以各种可能的方式爱这个国家。它并不完美，但它令人难以置信。所以我有幸环游世界，有时回头看看自己的国家，我们所有的考验和磨难和讨论，你知道，我们国家在很多方面的混乱，那正是它美丽的原因，也是它运转的原因。所以我很高兴成为它的一部分。

Look, I don't like tax dollars wasted. I like that our country is efficient and runs efficiently. But I also support all that this country represents. I support all of our government officials in their civil duties to make America great and better. I love the infrastructure that we've provided for all of our citizens. I love what it's created. For me, you know, this is—I am the ultimate American dream. And if not for this country, the education, the ability to create a new company, pursue your dreams, and let merits be the reason why you succeed—and your endeavor, your grit, your innovation, let that be the reason why you succeed. If not for, you know, why I'm here. I wouldn't have met all the people that made Nvidia possible. I wouldn't have met my wife in school. I mean, the list goes on. I love this country in every possible way. And it's not perfect, but it's incredible. And so I've had the benefit of traveling around the world and sometimes looking back in on your own country and all of our trials and tribulations and discourse, you know, the messiness of our country in a lot of ways, that's what makes it beautiful and it's what makes it work. And so I'm delighted to be part of it.

即使你可能最终在五年内支付大约 80 亿美元，仍然值得吗？

Even if you might end up paying about perhaps eight billion dollars over five years, still worth it?

嗯，我认为首先，我很高兴你这么说。我能在五年内支付 80 亿美元的税款，这是一种特权。不是那 80 亿美元，而是我能支付 80 亿美元这个事实。所以我不怕交税。我只是怕穷。而且我不介意为他人做贡献。我喜欢支付——我以说我想尽可能多地支付所有员工而闻名，我确实这样做，而不是尽可能少地付钱来获得他们的工作。我尽量多付钱给人。这是成功的特权。坦率地说，我很高兴这样做。我觉得这是一种特权。这是一种责任。没有什么比能够支付更多税款更让我高兴的了。

Well, I think that first of all, I'm glad you said that. The fact that I can afford to pay $8 billion in taxes over five years is a privilege. It's not the $8 billion. It's the fact that I can afford to pay $8 billion. And so I'm not afraid of paying taxes. I'm just afraid of being poor. And I have no trouble contributing to others. I love paying—I'm famed for saying that I want to pay all of our employees as much as I can, and I do, instead of paying people as little as possible to get their work. I try to pay people as much as possible. That's a privilege of being successful. And frankly, I'm delighted to do it. I feel it's a privilege. It's a responsibility. And nothing would give me more joy than to be able to pay even more taxes.

等等，你是在要求交更多税吗？

Wait, are you asking to be paying more in taxes?

不，你赚得越多，交的税就越多。所以我不介意。

No, the more you make, the more taxes you pay. So I'm okay with that.

那是真的。那是真的。

That's true. That's true.

我不介意。

I'm okay with that.

好的，Jensen，谢谢你。我们还有时间再问几个问题吗？或者我想尊重你的时间。

Well, Jensen, thank you. Do we have time for a couple more questions or I want to be respectful of your time.

我有的是时间。

I have all the time in the world.

哦，是吗？

Oh, you do?

是我的团队。

It's my team.

太好了。嗯，我想问你一件现在让很多美国人非常困扰的事。有很多人不希望数据中心建在他们家后院。这是罕见的两党一致。反对声是两党共同的。NVIDIA 为很多数据中心提供动力。你对那些公开反对、说我们不希望这些东西建在我们社区的人有什么想说的？

Excellent. Well, I think I want to ask you about something that's really bothering a lot of Americans right now. There's a lot of people who don't want data centers in their backyards. This is rare bipartisan unity. The backlash is bipartisan. NVIDIA powers a lot of data centers. What do you say to the people out there who are speaking out and saying we do not want these in our neighborhoods?

嗯，首先，我想说很抱歉我们没有早点来和你们沟通。我们应该早点来和他们谈，在你们甚至还没在某个社区、某个农村地区、某个小镇建起一个巨大的数据中心之前。你知道，我在一个只有 600 人的小镇长大，上的学校只有 300 个孩子，那 300 个孩子，那就是肯塔基州的奥尼塔，就那样。如果突然之间，一个数据中心出现在那唯一一家杂货店旁边，我想我会感到惊慌。所以我认为，首先，我们需要在建立数据中心之前很久就去和社区接触。我认为我所有的同事今天都认识到了这一点。你知道，这一切发生得太快了，在全美各地都发生得太快了，有时我们只是没有意识到，我们没有机会表现得礼貌，没有机会成为某个新小镇的访客。然而，我还要说，数据中心也在极大地、极快地改进。水，冷却系统现在是循环用水。我们用热水来冷却数据中心，所以所需的能量、所需的冷却量都大幅减少了。我们消耗的能源，我们应该为发电厂提供。事实上，我们甚至应该为当地电网做贡献，增强其可靠性。而美国越来越理解的一件事是，利用 AI 的市场动态和市场力量，以及这个不可思议的投资阶段，100 年来我们第一次有机会投资各种可持续能源，如果某个小镇想要太阳能，如果你喜欢核能，无论是聚变还是裂变，水电，今天每一种发电形式都在考虑之中，每个人都在被投资。所以这是一个奇妙的时刻。如果我们想要一个可持续的未来，未来五年，毫无疑问我们会使用更多的化石燃料。但在那之后的几十年，由于人工智能，我们将看到各种可持续能源的应用。所以我想说，我还想说，有很多小镇从数据中心建设者的投资中受益的精彩故事。你知道，他们的社区中心改善了，新学校、新棒球场、新道路、更低的财产税，所有这些都在考虑之中。所以我认为数据中心建设者应该去社区，让他们知道他们正在考虑在该社区建设，让他们知道对话包括升级他们小镇、降低税收、降低能源账单的机会，而且应该有一些最低标准。应该对这些数据中心的美观有一些期望。我认为它们应该设计得更好。它们可以设计得更好。

Well, first of all, I would say we're sorry we didn't come talk to you sooner. We should have come talk to them sooner before you even stand up a giant data center in some community, some rural area, some small town. You know, I grew up in a town of 600 people and I went to a school with 300 kids, and those 300 kids, that was Onita, Kentucky, that was it. And if all of a sudden a data center were to crop up right next to the one grocery store, I think I would be alarmed. And so I think that first of all, we need to go engage the communities long before we set up the data centers. And I think all of my colleagues recognize this today. You know, it's just all of this happened so quickly and it happened all over America so quickly that sometimes we just didn't realize, and we didn't get a chance to be polite, to be a visitor in some new town. However, I will also say the data centers are also improving tremendously and tremendously quickly. The water, the cooling system is now recirculating water. We use hot water to cool the data centers, and so the amount of energy necessary, the amount of cooling necessary is dramatically reduced. The energy that we consume, we should be providing for the power generation plants. We should contribute, in fact, even to the local power grid, enhancing its reliability. And one of the things that the United States is increasingly understanding, using the market dynamics and the market forces of AI and just this incredible investment phase, the first time in a 100 years we have an opportunity to invest in all kinds of sustainable energies, where if a particular town would like solar, if you like nuclear, whether it's fusion or fission, hydro, every single form of power generation is on the table today and everybody is being invested in. And so that's a fantastic moment. If we want a sustainable future, the next five years, recognize no doubt we'll use much more fossil fuel. But the decades after that, we're going to see all kinds of applications of sustainable energy because of artificial intelligence. And so I would say, I would also say that there are so many great stories of towns that have been at the benefit of investments from the data center builders. You know, their community centers improved, new schools, new baseball fields, new roads, lower property taxes, all of that is on the table. And so I think that data center builders should go to the communities, let them know that they're considering building in the community, let them know that the conversation includes an opportunity to upgrade their town, lower their taxes, lower their energy bill, and there should be some minimums. There should be some expectation about, you know, the aesthetics of these data centers. I think they ought to be designed better. They could be designed better.

一种设计数据中心的新方法。尽管主要的担忧真的是关于……让东西变得美观只是等式的一部分。我认为今天没有人认为数据中心的美观是其功能的一部分。嗯，人们考虑很多的是能源账单、水以及投入的资源。这似乎是抗议这些数据中心的人的主要担忧。

A new approach to designing the data center. Although the main concern really is about making something beautiful is just part of the equation. I think that today nobody considers the aesthetics of a data center to be part of its functionality. Well, people think a lot about the energy bills and the water and the resources that go into it. That seems to be the primary concern from people who are protesting these data centers.

所以那些东西，水消耗，现在是个迷思。这些数据中心使用更少的蒸发水。它们不再使用蒸发。它使用的水比一个游泳池全年蒸发的水还少。所以我认为所有这些都是历史性的，可以理解，但现在是个迷思。能源账单上涨实际上恰恰相反。大多数这些数据中心，所有这些都是数据中心应该为电网做贡献，如果它们做贡献，当它们为电网做贡献时，会使电网更有韧性。它们降低电力成本。

And so those things, water consumption, that's a myth now. These data centers use less water evaporative. They don't use evaporation anymore. It uses less water than a swimming pool evaporating all year long. And so I think all of that is kind of historic, understandable, but it's a myth now. Energy bills going up is actually the opposite. Most of these data centers, all these data centers should contribute to the grid, and if they contribute, when they contribute to the grid, it makes the grid more resilient. They reduce the power cost.

听起来你是在说人们应该重新考虑让这些建在他们社区，但我们应该，你仍然应该做得漂亮。你仍然应该让它更美观。

It sounds like you're trying to say that people should reconsider having these in their neighborhood, but that we should you should still do it nicely. You should still make it more aesthetic.

退距应该离城镇更远。因为你知道这些数据中心，有很多车辆进出。有噪音等等。所以你只要把它，你必须对这些事情保持理智。我们，你知道，我们应该，我们应该很早就去和社区对话，在我们出现之前。它发生得太快了，所以对此很不幸。但展望未来，我们可以做得更好。

The setback should be further away from the towns. Because you know these data centers, there's a lot of traffic in and out. There's noise and so on and so forth. And so you just got to put it, you have to be sensible about these things. We have, you know, we should have, we should have very early on gone and had dialogue with the communities before we showed up. It just happened so quickly and so for that it's unfortunate. But going forward we can do a much better job.

你创办了这家公司，你创办 Nvidia 时是一家游戏芯片公司。30 年前你坐在一家 Denny's 里。

You started this company, you started Nvidia as a gaming chip company. You were sitting in a Denny's 30 years ago.

我当时是 Denny's 的洗碗工。

I was a dishwasher at the Denny's.

是的。那就是我喜欢 Denny's 的地方。我喜欢 Denny's。美国的餐馆。谁？是的。

Yeah. That's where I love Denny's. I love Denny's. America's Diner. Who? Yeah.

当你在 Denny's 当洗碗工时，你有没有想过你会最终站在关于 AI 和人类终结的辩论前线？

Did you ever think when you were a dishwasher at Denny's that you would end up on the front lines about a debate regarding AI and the end of humanity?

呃，没有。我只是非常非常专注于洗碗，像没有洗碗工那样洗碗，而且我是最棒的服务生助手，比任何收拾桌子的服务生助手都棒。你知道，我对我做过的几乎每一份工作都充满热情。不管是什么工作。

Uh, no. I was just really really focused on doing dishes like no dishwasher ever did dishes and um I was the best bus boy of any bus boy that ever bust tables. You know I brought enthusiasm to almost every job that I had. It didn't matter what it was.

看起来你今天把同样的热情带到了你的工作中。

It seems like you're bringing that same level of enthusiasm to your job today.

这很重要。这很关键。这是一场重要的对话。我很高兴我们正在进行，我想确保美国公众明白，这个行业将会，我们的责任是安全地构建技术，就像制药行业安全地制药，汽车行业安全地造车，飞机制造业安全地造飞机一样。这个行业将安全地构建 AI。我们必须对此负责，这是第一。第二，不要让迷思和荒谬的戏剧性故事导致我们对这种不可思议的能力不乐观。

This is consequential. This is important. It's an important conversation. I'm happy we're having it and I want to make sure that the American public understands that this industry will, it is our responsibility to build the technology safely, just as it is the drug industry to build the drug safely, the automotive, the car industry to build the cars safely, the plane industry to build the plane safely. This industry will build AI safely. We have to hold ourselves accountable to that, number one. Number two, don't let the myths and the ridiculous dramatic stories cause us not to be optimistic about this incredible capability.

我们不想成为发明了这次工业革命技术的国家，却让自己的国家被落在后面。别让这种事发生。我不希望我的孩子遇到这种情况。我也不希望你的孩子遇到这种情况。

We do not want to be the country that invented the technologies of this industrial revolution but have this country be left behind. Don't let that happen. I don't want that happening for my kids. I don't want that happening for your kids.

你能解释一下数据中心实际上做什么，以及为什么 AGI（通用人工智能）可能是件好事吗？

Can you explain what data centers actually do and why AGI could be a good thing?

好的。这个问题问得真好。首先，如你所知，几百年前，大约 350 年前，现代工业革命开始了。电力被制造出来。它是一件神奇的事，几乎是一件神话般的事：你可以让水进入这台机器，然后一种看不见的技术——叫做电——从中出来，我们可以把能量输送到世界各地，取代蜡烛、炉子、烧木头的烤箱等等，还有蒸汽机。所以我们让这个新产业成为可能，整个产业都建立在它之上。然后下一次工业革命，也就是上一次工业革命，实际上是关于信息时代的，我们通过建造这些数据中心让互联网连接在一起，从而让我们可以找到任何东西。

Okay. So really great question. First of all, as you know, several hundred years ago, about 350 years ago, the modern industrial revolution started. Electricity was being manufactured. It was a magical thing, almost a mythical thing, where you could have water come into this machinery and this invisible technology comes out of it called electricity, and we can transport energy all over the world, replacing candles and stoves and wood-burning ovens and so on, steam engines. So we made it possible for this new industry to happen, and a whole industry was built on top of it. Then the next industrial revolution, the last industrial revolution, was really about the information age, and we made it possible by building these data centers for the internet to be connected together so that we can find anything.

所以在之前的工业秩序中，我们可以为任何东西提供动力。现在我们可以找到任何东西。而今天我们创造了一种叫做人工智能的技术：给定信息、给定上下文，它就能对此进行推理，对你给它的信息进行推理，从而能够回答并做任何事情。你甚至可以使用工具，比如，你和我不用再使用网页浏览器，或者你和我不用再使用 PowerPoint、Excel，或者叫做 AutoCAD 或 Blender 之类的 3D 设计工具。这个软件可以推理你要求它做的事情，获取上下文、你需要它完成你意图的信息，然后它为你完成工作。

So the previous industrial order we could power anything. Now we can find anything. And where we are today, we created this technology called artificial intelligence where, given information, given context, it can reason about it, it can reason about the information that you gave it so that it can answer and do anything. You can even use tools like, for example, instead of you and I using a web browser or you and I using PowerPoint, Excel, or a 3D design tool called AutoCAD or Blender or something like that. This software can reason about what you asked it to do, take the context, the information that you need it to do your intentions, and it does the work for you.

而 OpenAI 的最新一代叫做 Astra，它令人难以置信。你只要告诉 Astra：“这是我后院的照片。这里还有一些我想要的例子，也许我想在里面放一个游泳池或一个花园。我现在要你把它变成一个设计文件，给我的园丁和木匠，这样他们就能为我做那个后院，或者为我设计那栋房子，为我建造那栋房子。”于是 Astra 使用所有必要的工具和你给它的图像，为房子、厨房或后院创建 CAD 文件。太不可思议了。这就是 AGI（通用人工智能）所做的。

And the latest generation from OpenAI called Astra is incredible. You just tell Astra, "Here's a picture of my backyard. Here's some other examples that I would like, maybe a pool or a garden I like to place in it. I want you to now turn it into a design file for my gardeners and my carpenters so that they could do that backyard for me or design that house for me, build that house for me." And so Astra uses all the necessary tools and the images that you gave it and it creates the CAD files for the house or the kitchen or the backyard. It's incredible. And so that's what AGI does.

然而，现在必要的是我们需要一种新型基础设施。不再是发电机，不再是互联网的数据中心，我们现在需要一个人工智能 AI 工厂。它是一个 AI 工厂。能量输入其中，数据输入其中，这些我们称为 token 的数字，基本上被重新组合成智能。我知道这听起来很复杂，但基本上能量输入，智能输出，它基本上就是一个 AI 工厂。

Now what is necessary, however, is we now need a new type of infrastructure. Instead of a power generator, instead of a data center for the internet, we now need an artificial intelligence AI factory. It's an AI factory. Energy comes into it, data comes into it, and these numbers we call tokens, but basically reformulated into intelligence. I know that sounds complicated, but basically energy comes in, intelligence comes out and it basically is an AI factory.

输入输出。

Input output.

输入输出。就这么简单。当然，这些工厂里面有很多 Nvidia 制造的计算机。

Input output. It's as simple as that. And of course, these factories have a lot of computers inside that Nvidia makes.

所以现在的问题是，这个行业会有多大？事实证明，几乎经济的每个部分和每个人，我们都喜欢了解事物。我们喜欢做事。也许我们喜欢自动化购物。也许我们想自动化规划假期或旅行，并为我们找到最优惠的价格。等等等等。我刚才举了例子，也许我们想改善后院或设计新厨房。所以每个行业都有信息工作要做，而全世界的产业，称之为 100 万亿美元的经济，其中大约 15 万亿美元将受益于比目前更多的智能。所以这些工厂是环绕每个国家的一层。每家公司都会使用它。这就是为什么有这么多数据中心的原因。这将是一个数万亿美元的产业。

And so now the question is how big would this industry be? Now it turns out that just about every part of the economy and every human, we like to know things. We like to do things. Maybe we like to automate our shopping. Maybe we want to automate the planning of a vacation or a trip and find us the best fares. So on and so forth. I just gave you the example of maybe we want to improve our backyard or design a new kitchen. And so every single industry has information work to do, and with the world's industries, call it a hundred trillion dollars of economy, and out of that some 15 trillion dollars of it would benefit from more intelligence than it currently has. And so these factories is a layer that goes around every single country. Every company will use it. And that's the reason why there's so many of these data centers. This is going to be a multi-trillion dollar industry.

那么问题就变成了，从美国的角度来看，我们如何看待这些数据中心。需要意识到的是，这是 100 年来第一次，至少是 50 年来第一次，我们有机会重新工业化美国，意味着把制造业带回美国。当然，这是制造某种看不见的东西，就像发电一样。

Now the question then becomes how do we think about these data centers from America's perspective. And the thing to realize is that this is the first time in a 100 years, well at least first time in 50 years, that we have an opportunity to reindustrialize the United States, meaning bring manufacturing back to United States. Now, of course, this is manufacturing of something that's invisible, just like power generation was.

不完全是你的汽车制造。

Not exactly your car manufacturing.

没错。传统的。它不是制造原子。它是制造电子。好吧。如果你愿意，它是看不见的。但它仍然是制造业。而且它正在美国各地创造各种制造业工作岗位，如你所知。建筑工人、电工、装配工、设计师、建造者，每个人都忙得不可开交，因为他们正在美国各地建造价值数万亿美元的 AI 工厂。

That's right. Traditional. It's not manufacturing atoms. It's manufacturing electrons. Okay. If you will, it's invisible. But it's manufacturing nonetheless. And it's creating all kinds of manufacturing jobs across the United States as you know. The construction workers, electricians, the fitters, the designers, the builders, everybody is so incredibly busy because they're building out the several trillion dollars worth of AI factories across America.

所以这部分让我非常兴奋，这是几十年来第一次，我们国家的蓝领工人、建造者、手工艺人，这是他们的巨大机会。他们现在都享受着这样的繁荣。我喜欢看到我们真的缺少电工和建筑工人。但这对美国来说是一个绝妙的机会，可以重新工业化，再次成为一个制造业国家，为我们国家的蓝领工人和所有建造者创造机会，并再次带回建筑和制造业。所以我认为这太不可思议了。所以它正在创造各种就业机会。

And so this is the part that is really exciting to me is that this is the first time in several decades that the blue-collar workers, the builders, the crafts people of our country, this is their great opportunity. And all of them are enjoying such prosperity now. And I love seeing that we have a real shortage of electricians and construction workers. But this is just a wonderful opportunity for United States to re-industrialize, be a manufacturing country again, for us to create opportunities for the blue-collar workers and all the builders of our country and to bring back building and manufacturing again. And so I think it's incredible. So it's creating all kinds of jobs.

很多人不把数据中心看作一个制造业的故事。所以我觉得这很有趣。

A lot of people don't see data centers as a manufacturing story. So I think it's interesting.

没错。是的。正如我之前所说，这只是制造业的新版本。发电是工业。发电是工业化的，你有工人在那里建造这些工厂，它必须被建造。它是由建造者建造的，而数据中心只是制造业的新版本。而这个新版本的制造产品是看不见的。它是数字。我知道制造数字很奇怪。但本质上就是这样。

That's right. Yeah. It's just a new version of manufacturing just as I said earlier. Power generation is industry. Power generation is industrial and you have people in there that's building these factories and it has to be built. It's built by builders and data centers are just manufacturing the new version. And this new version of manufacturing products is it's invisible. It's numbers. I know it's weird to be manufacturing numbers. But that's essentially what it is.