黄仁勋：英伟达 AI 革命背后的愿景

在某个时刻，你必须相信一些东西。我们重新定义了计算。你对接下来会发生什么有什么愿景？我们问自己：如果它能做到这一点，它能走多远？我们如何从现在的机器人走向你看到的未来世界？克莱奥：总有一天，所有会动的东西都将是机器人，而且很快就会实现。我们在它真正发生之前投资了数百亿美元。不，那很好，你做了一些研究。但我想说，最大的突破是当我们……那是黄仁勋，无论你是否意识到，他的决定正在塑造你的未来。他是英伟达的 CEO，这家公司在过去几年里飙升，成为世界上最有价值的公司之一，因为他们领导了计算机工作方式的根本性转变，释放了当前技术可能性的爆炸。英伟达又做到了。我们发现自己成为世界上最重要的科技公司之一，甚至可能是有史以来最重要的。你听到的关于人工智能、机器人、游戏、自动驾驶汽车和突破性医学研究的大量最前沿技术，都依赖于他和他的公司设计的新芯片和软件。在我为准备这次采访而进行的数十次背景访谈中，最让我印象深刻的是黄仁勋在过去 30 年里已经影响了我们所有人的生活，而且很多人说这只是更大事情的开始。我们都需要知道他在建造什么以及为什么，最重要的是，他接下来要建造什么。欢迎来到《宏大对话》。非常感谢你来做这个节目。很高兴这样做。在我们深入之前，我想告诉你这次采访将与你最近看到的其他采访有点不同。好的。我不会问你任何关于公司财务的问题。谢谢。我不会问你关于你的管理风格或你为什么不喜欢一对一会议的问题。我不会问你关于法规或政治的问题。我认为所有这些都很重要，但我认为我们的观众可以在其他地方得到很好的覆盖。好的。我们在《宏大如果为真》中所做的是制作乐观的解释性视频，我们已经覆盖了……我是最不适合做解释性视频的人。我认为你可能最适合。而我真的希望我们能一起做的是制作一个联合解释性视频，关于我们如何实际利用技术让未来更美好。是的。我们这样做是因为我们相信当人们看到那些更美好的未来时，他们会帮助建造它们。所以你要交谈的人都很棒。他们是想要建造更美好未来的乐观主义者。但因为我们的主题如此多样——我们覆盖了超音速飞机、量子计算机和粒子对撞机——这意味着数百万人在没有任何先验知识的情况下进入每一集。你可能在和一个领域的专家交谈，他分不清 CPU 和 GPU 的区别，或者一个 12 岁的孩子，他有一天可能会成长为像你一样的人，但才刚刚开始学习。就我而言，我已经为这次采访准备了几个月，包括与你的团队许多成员进行背景对话，但我不是工程师。所以我的目标是帮助那些观众看到你所看到的未来。所以我将问三个领域。第一个是：我们是如何走到这一步的？导致我们现在所处的计算根本性转变的关键洞察是什么？第二个是：现在实际发生了什么？这些洞察如何导致了我们现在生活的这个世界，似乎一切同时发生？第三个是：你对接下来会发生什么有什么愿景？

At some point you have to believe something. We've reinvented computing as we know it. What is the vision for what you see coming next? We asked ourselves: if it can do this, how far can it go? How do we get from the robots that we have now to the future world that you see? Cleo: everything that moves will be robotic someday, and it will be soon. We invested tens of billions of dollars before it really happened. No, that's very good, you did some research. But the big breakthrough, I would say, is when we... That's Jensen Huang, and whether you know it or not, his decisions are shaping your future. He's the CEO of Nvidia, the company that skyrocketed over the past few years to become one of the most valuable companies in the world because they led a fundamental shift in how computers work, unleashing this current explosion of what's possible with technology. Nvidia's done it again. We found ourselves being one of the most important technology companies in the world, and potentially ever. A huge amount of the most futuristic tech that you're hearing about in AI and robotics and gaming and self-driving cars and breakthrough medical research relies on new chips and software designed by him and his company. During the dozens of background interviews that I did to prepare for this, what struck me most was how much Jensen Huang has already influenced all of our lives over the last 30 years, and how many said it's just the beginning of something even bigger. We all need to know what he's building and why, and most importantly, what he's trying to build next. Welcome to Huge Conversations. Thank you so much for doing this. So happy to do it. Before we dive in, I wanted to tell you how this interview is going to be a little bit different than other interviews I've seen you do recently. Okay. I'm not going to ask you any questions about company finances. Thank you. I'm not going to ask you questions about your management style or why you don't like one-on-ones. I'm not going to ask you about regulations or politics. I think all of those things are important, but I think that our audience can get them well covered elsewhere. Okay. What we do on Huge If True is we make optimistic explainer videos, and we've covered... I'm the worst person to be an explainer video. I think you might be the best. And I think that's what I'm really hoping that we can do together is make a joint explainer video about how can we actually use technology to make the future better. Yeah. And we do it because we believe that when people see those better futures, they help build them. So the people that you're going to be talking to are awesome. They are optimists who want to build those better futures. But because we cover so many different topics—we've covered supersonic planes and quantum computers and particle colliders—it means that millions of people come into every episode without any prior knowledge whatsoever. You might be talking to an expert in their field who doesn't know the difference between a CPU and a GPU, or a 12-year-old who might grow up one day to be you but is just starting to learn. For my part, I've now been preparing for this interview for several months, including doing background conversations with many members of your team, but I'm not an engineer. So my goal is to help that audience see the future that you see. So I'm going to ask about three areas. The first is: how did we get here? What were the key insights that led to this big fundamental shift in computing that we're in now? The second is: what's actually happening right now? How did those insights lead to the world that we're now living in that seems like so much is going on all at once? And the third is: what is the vision for what you see coming next?

为了谈论我们正处于的人工智能这一重大时刻，我认为我们需要回到 90 年代的电子游戏。当时，我知道游戏开发者想要创建更逼真的图形，但硬件无法跟上所有必要的数学计算。人手不够。英伟达提出了一个解决方案，它不仅改变了游戏，还改变了计算本身。你能带我们回到那个时候，解释当时发生了什么，以及是什么洞察导致你和英伟达团队创造了第一个现代 GPU 吗？

In order to talk about this big moment we're in with AI, I think we need to go back to video games in the 90s. At the time, I know game developers wanted to create more realistic looking graphics, but the hardware couldn't keep up with all of that necessary math. Not enough man. Nvidia came up with a solution that would change not just games but computing itself. Could you take us back there and explain what was happening and what were the insights that led you and the Nvidia team to create the first modern GPU?

所以在 90 年代初我们刚创办公司时，我们观察到在一个软件程序中，只有几行代码——可能 10%的代码——完成了 99%的处理，而那 99%的处理可以并行完成。然而，另外 90%的代码必须顺序执行。事实证明，合适的计算机，完美的计算机，是能够同时进行顺序处理和并行处理的，而不仅仅是其中之一。那是重大的观察，我们着手建立一家公司来解决普通计算机无法解决的计算机问题。这真的是英伟达的开始。

So in the early '90s when we first started the company, we observed that in a software program, inside it, there are just a few lines of code—maybe 10% of the code—that does 99% of the processing, and that 99% of the processing could be done in parallel. However, the other 90% of the code has to be done sequentially. It turns out that the proper computer, the perfect computer, is one that could do sequential processing and parallel processing, not just one or the other. That was the big observation, and we set out to build a company to solve computer problems that normal computers can't. And that's really the beginning of Nvidia.

我最喜欢的关于 CPU 与 GPU 为何如此重要的视觉展示是英伟达 YouTube 频道上一个 15 年前的视频，其中《流言终结者》使用一个小机器人一个一个地发射彩弹，展示在 CPU 上一次解决一个问题或顺序处理，然后他们推出一个巨大的机器人，一次发射所有彩弹，同时处理较小的问题或在 GPU 上进行并行处理。3, 2, 1。所以英伟达为电子游戏解锁了所有这些新力量。为什么先做游戏？

My favorite visual of why a CPU versus a GPU really matters so much is a 15-year-old video on the Nvidia YouTube channel where the Mythbusters use a little robot shooting paintballs one by one to show solving problems one at a time or sequential processing on a CPU, but then they roll out this huge robot that shoots all of the paintballs at once, doing smaller problems all at the same time or parallel processing on a GPU. 3, 2, 1. So Nvidia unlocks all of this new power for video games. Why gaming first?

电子游戏需要并行处理来处理 3D 图形，我们选择电子游戏是因为，第一，我们喜欢这个应用——它是虚拟世界的模拟，谁不想去虚拟世界呢？而且我们有一个很好的观察：电子游戏有潜力成为有史以来最大的娱乐市场，事实证明这是真的。拥有一个巨大的市场很重要，因为技术很复杂，如果我们有一个大市场，我们的研发预算就可以很大。我们可以创造新技术，而技术、市场和更先进技术之间的飞轮正是让英伟达成为世界上最重要的科技公司之一的飞轮。这一切都归功于电子游戏。

The video games require parallel processing for processing 3D graphics, and we chose video games because one, we loved the application—it's a simulation of virtual worlds, and who doesn't want to go to virtual worlds? And we had the good observation that video games have potential to be the largest market for entertainment ever, and it turned out to be true. And having it being a large market is important because the technology is complicated, and if we had a large market, our R&D budget could be large. We could create new technology, and that flywheel between technology and market and greater technology was really the flywheel that got Nvidia to become one of the most important technology companies in the world. It was all because of video games.

我听过你说 GPU 是一台时间机器。你能告诉我更多关于你是什么意思吗？

I've heard you say that GPUs were a time machine. Could you tell me more about what you mean by that?

GPU 就像一台时间机器，因为它让你更早地看到未来。有人对我说过的最令人惊叹的事情之一是一位量子化学科学家。他说：‘黄仁勋，因为英伟达的工作，我可以在我的有生之年完成我毕生的工作。’那就是时间旅行。他能够在自己的有生之年完成超越其一生的工作。这是因为我们让应用程序运行得更快，你就能看到未来。所以当你做天气预报时，例如，你就在看到未来。当你进行模拟——一个带有虚拟交通的虚拟城市——并且我们通过那个虚拟城市模拟我们的自动驾驶汽车时，我们就在进行时间旅行。所以并行处理在游戏中起飞，它让我们能够在计算机中创造世界……

A GPU is like a time machine because it lets you see the future sooner. One of the most amazing things anybody's ever said to me was a quantum chemistry scientist. He said: 'Jensen, because of Nvidia's work, I can do my life's work in my lifetime.' That's time travel. He was able to do something that was beyond his lifetime within his lifetime. And this because we make applications run so much faster, and you get to see the future. So when you're doing weather prediction, for example, you're seeing the future. When you're doing a simulation—a virtual city with virtual traffic—and we're simulating our self-driving car through that virtual city, we're doing time travel. So parallel processing takes off in gaming and it's allowing us to create worlds in computers that we...

以前从未有过。游戏是并行处理释放更多算力的第一个惊人案例。然后如你所说，人们开始将这种算力用于许多不同行业。那位量子化学研究员的故事：我听你讲过，他运行分子模拟的方式是，即使在当时，在英伟达 GPU 上并行运行也比在他以前使用的 CPU 超级计算机上快得多。

Never could have before. And gaming is sort of this first incredible case of parallel processing unlocking a lot more power. And then as you said, people began to use that power across many different industries. The case of the quantum chemistry researcher: when I've heard you tell that story, it's that he was running molecular simulations in a way where it was much faster to run in parallel on Nvidia GPUs, even then, than it was to run them on the supercomputer with the CPU that he had been using before.

是的，没错。

Yeah, that's true.

所以，天哪，它也在彻底改变所有这些其他行业。它开始改变我们对计算机可能性的看法。据我所知，在 21 世纪初，你看到了这一点，并意识到实际上这样做有点困难，因为那位研究员必须想办法欺骗 GPU，让它们以为他的问题是一个图形问题。

So, oh my God, it's revolutionizing all of these other industries as well. It's beginning to change how we see what's possible with computers. And my understanding is that in the early 2000s, you see this and you realize that actually doing that is a little bit difficult because what that researcher had to do is he had to sort of trick the GPUs into thinking that his problem was a graphics problem.

完全正确。不，你说得很好。嗯，你做过一些研究。

That's exactly right. No, that's very good. Well, you did some research.

所以你创造了一种方法，让这变得容易得多。没错。具体来说，它是一个名为 CUDA 的平台，它让程序员可以使用他们已经知道的编程语言（如 C 语言）来告诉 GPU 做什么。这很重要，因为它让更多人更容易获得所有这些计算能力。你能解释一下是什么愿景促使你创建 CUDA 的吗？

So you create a way to make that a lot easier. That's right. Specifically, it's a platform called CUDA, which lets programmers tell the GPU what to do using programming languages that they already know, like C. And that's a big deal because it gives way more people easier access to all of this computing power. Could you explain what the vision was that led you to create CUDA?

部分原因是研究人员的发现，部分原因是内部的灵感，还有部分原因是解决问题。很多有趣的想法都来自这种混合。有些是抱负和灵感，有些纯粹是绝望。CUDA 的情况也差不多。最早使用我们的 GPU 进行并行处理的外部想法可能来自医学影像领域的一些有趣工作。麻省总医院的几位研究人员用它来做 CT 重建。他们出于这个原因使用了我们的图形处理器，这启发了我们。与此同时，我们公司内部试图解决的问题是，当你为电子游戏创建虚拟世界时，你希望它既美观又动态。水应该像水一样流动，爆炸应该像爆炸一样。所以有粒子物理，你想做流体动力学，如果你的流水线只能做计算机图形，那就难多了。因此，在我们服务的市场中，我们有自然的理由去做这件事。研究人员也在尝试将我们的 GPU 用于通用加速。所以多种因素汇聚在一起。当时机成熟时，我们决定做一件正经事，并由此创建了 CUDA。从根本上说，我确信 CUDA 会成功，并且我们把整个公司都押在它上面，是因为我们的 GPU 将成为世界上产量最高的并行处理器，因为电子游戏市场非常大。所以这个架构有很大机会惠及许多人。

Partly researchers discovering it, partly internal inspiration, and partly solving a problem. A lot of interesting ideas come out of that soup. Some of it is aspiration and inspiration, some of it is just desperation. In the case of CUDA, it's very much the same way. Probably the first external ideas of using our GPUs for parallel processing emerged out of some interesting work in medical imaging. A couple of researchers at Mass General were using it to do CT reconstruction. They were using our graphics processor for that reason, and it inspired us. Meanwhile, the problem that we're trying to solve inside our company has to do with the fact that when you're trying to create these virtual worlds for video games, you would like it to be beautiful but also dynamic. Water should flow like water, and explosions should be like explosions. So there's particle physics, you want to do fluid dynamics, and that is much harder to do if your pipeline is only able to do computer graphics. So we have a natural reason to want to do it in the market that we were serving. Researchers were also horsing around with using our GPUs for general-purpose acceleration. So there were multiple factors coming together in that soup. When the time came, we decided to do something proper and create CUDA as a result of that. Fundamentally, the reason why I was certain that CUDA was going to be successful and we put the whole company behind it was because our GPU was going to be the highest volume parallel processor built in the world, because the market of video games was so large. So this architecture has a good chance of reaching many people.

在我看来，创建 CUDA 是一件极其乐观、如果成真就意义重大的事情，你在说：如果我们为更多人创造一种使用更多算力的方式，他们可能会创造出不可思议的东西。然后当然它成真了。他们做到了。2012 年，三位研究人员向一个著名的竞赛提交了参赛作品，目标是创建能够识别图像并按类别标记的计算机系统。他们的参赛作品彻底击败了对手。它犯的错误少得多。太不可思议了。它让所有人震惊。它叫 AlexNet，是一种称为神经网络的 AI。据我所知，它如此出色的一个原因是他们使用了大量数据来训练该系统，并且是在英伟达 GPU 上完成的。突然间，GPU 不再只是让计算机更快更高效的方式。它们正在成为一种全新计算方式的引擎。我们正在从用逐步指令指导计算机，转向通过向计算机展示大量示例来训练它们学习。2012 年的这一刻真正开启了我们现在都看到的 AI 领域的巨大转变。你能从你的角度描述一下那一刻是什么样的吗？你当时认为这对我们所有人的未来意味着什么？

It has seemed to me like creating CUDA was this incredibly optimistic, huge-if-true thing to do, where you were saying: if we create a way for many more people to use much more computing power, they might create incredible things. And then of course it came true. They did. In 2012, a group of three researchers submits an entry to a famous competition where the goal is to create computer systems that could recognize images and label them with categories. And their entry just crushes the competition. It gets way fewer answers wrong. It was incredible. It blows everyone away. It's called AlexNet, and it's a kind of AI called the neural network. My understanding is one reason it was so good is that they used a huge amount of data to train that system, and they did it on Nvidia GPUs. All of a sudden, GPUs weren't just a way to make computers faster and more efficient. They're becoming the engines of a whole new way of computing. We're moving from instructing computers with step-by-step directions to training computers to learn by showing them a huge number of examples. This moment in 2012 really kicked off this truly seismic shift that we're all seeing with AI right now. Could you describe what that moment was like from your perspective, and what did you see it would mean for all of our futures?

当你创造像 CUDA 这样的新东西时，如果你建好了，他们可能不会来。这总是愤世嫉俗者的观点。然而，乐观者的观点会说：但如果你不建，他们就不能来。我们通常就是这样看待世界的。我们必须凭直觉推理为什么这会非常有用。事实上，在 2012 年，多伦多大学的 Alex Krizhevsky、Ilya Sutskever 和 Geoff Hinton，他们所在的实验室，他们联系了 GeForce GTX 580，因为他们了解了 CUDA，并且 CUDA 可能可以用作训练 AlexNet 的并行处理器。所以我们的灵感——GeForce 可以成为将这种并行架构推向世界的载体，并且研究人员总有一天会找到它——是一个好策略。这是一个基于希望的策略，但也是有理有据的希望。真正引起我们注意的是，与此同时，我们正在公司内部试图解决计算机视觉问题，并且我们试图让 CUDA 成为一个好的计算机视觉处理器。我们在内部早期的计算机视觉工作以及让 CUDA 能够做到这一点方面遇到了很多挫折。突然间，我们看到了 AlexNet，这种全新的算法，与之前的计算机视觉算法完全不同，在计算机视觉能力上实现了巨大飞跃。当我们看到它时，部分是出于兴趣，但部分是因为我们自己也在挣扎。所以我们非常感兴趣，想看到它成功。当我们看到 AlexNet 时，我们受到了启发。但我要说，最大的突破是，当我们看到 AlexNet 时，我们问自己：AlexNet 能走多远？如果它能在计算机视觉上做到这一点，它能走多远？如果它能达到我们认为它能达到的极限，它能解决的那种问题，对计算机行业意味着什么？对计算机架构意味着什么？我们合理地推断，如果机器学习，如果深度学习架构可以扩展，绝大多数机器学习问题都可以用深度神经网络表示。而我们可以用机器学习解决的问题种类如此之多，以至于它有潜力彻底重塑整个计算机行业。这促使我们重新设计……

When you create something new like CUDA, if you build it, they might not come. That's always the cynic perspective. However, the optimist perspective would say: but if you don't build it, they can't come. And that's usually how we look at the world. We have to reason about intuitively why this would be very useful. In fact, in 2012, Alex Krizhevsky, Ilya Sutskever, and Geoff Hinton at the University of Toronto, the lab they were at, they reached out to a GeForce GTX 580 because they learned about CUDA and that CUDA might be able to be used as a parallel processor for training AlexNet. So our inspiration that GeForce could be the vehicle to bring out this parallel architecture into the world, and that researchers would somehow find it someday, was a good strategy. It was a strategy based on hope, but it was also reasoned hope. The thing that really caught our attention was simultaneously we were trying to solve the computer vision problem inside the company, and we were trying to get CUDA to be a good computer vision processor. We were frustrated by a whole bunch of early developments internally with respect to our computer vision effort and getting CUDA to be able to do it. All of a sudden, we saw AlexNet, this new algorithm that is completely different than computer vision algorithms before it, take a giant leap in terms of capability for computer vision. When we saw that, it was partly out of interest, but partly because we were struggling with something ourselves. So we were highly interested to want to see it work. When we looked at AlexNet, we were inspired by that. But the big breakthrough, I would say, is when we saw AlexNet, we asked ourselves: how far can AlexNet go? If it can do this with computer vision, how far can it go? And if it could go to the limits of what we think it could go, the type of problems it could solve, what would it mean for the computer industry? And what would it mean for computer architecture? We rightfully reasoned that if machine learning, if the deep learning architecture can scale, the vast majority of machine learning problems could be represented with deep neural networks. And the type of problems we could solve with machine learning is so vast that it has the potential of reshaping the computer industry altogether. Which prompted us to re-engineer the...

整个计算堆栈——DGX 就源于此，还有这个小家伙 DGX 坐在这里——全都源于一个观察：我们应该一层一层地重新发明整个计算堆栈。自从 IBM System/360 引入现代通用计算以来，65 年过去了，我们重新定义了计算。这是一个完整的故事：并行处理重新定义了现代游戏，并彻底改变了一个行业。然后这种并行计算方式开始在不同行业中使用。你通过构建 CUDA 来投资它，而 CUDA 和 GPU 的使用使得神经网络和机器学习实现了阶跃式变化，并开启了一场革命，其重要性如今只增不减。突然间，计算机视觉被解决了，语音识别被解决了，语言理解被解决了——这些与智能相关的不可思议的问题，一个接一个，过去我们渴望解决却毫无办法，现在每隔几年就被解决了。真是不可思议。

The entire computing stack, which is where DGX came from, and this little baby DGX sitting here—all of this came from the observation that we ought to reinvent the entire computing stack layer by layer. Computers, after 65 years since IBM System/360 introduced modern general-purpose computing, we've reinvented computing as we know it. This is a whole story: parallel processing reinvents modern gaming and revolutionizes an entire industry. Then that way of computing, parallel processing, begins to be used across different industries. You invest in that by building CUDA, and then CUDA and the use of GPUs allows for a step change in neural networks and machine learning, and begins a revolution that we're now seeing only increase in importance today. All of a sudden, computer vision is solved, speech recognition is solved, language understanding is solved—these incredible problems associated with intelligence, one by one, where we had no solutions in the past, desperate desire to have solutions for, all of a sudden get solved every couple of years. It's incredible.

所以你在 2012 年看到了这一点，你展望未来，相信那就是你将要生活的未来，并且你正在下注去实现它——非常大的赌注，风险极高。然后我作为一个外行的感觉是，这需要相当长的时间才能实现。你下这些赌注，八年，十年。所以我的问题是：如果 AlexNet 发生在 2012 年，而 10 年后这个观众可能才看到和听到更多关于 AI 和英伟达的消息，为什么花了十年？而且，因为你下了那些赌注，那十年的中间阶段对你来说是什么感觉？

So you're seeing that in 2012, you're looking ahead and believing that that's the future you're going to be living in now, and you're making bets that get you there—really big bets that have very high stakes. And then my perception as a lay person is that it takes a pretty long time to get there. You make these bets, eight years, ten years. So my question is: if AlexNet happened in 2012, and this audience is probably seeing and hearing so much more about AI and Nvidia specifically 10 years later, why did it take a decade? And also, because you would place those bets, what did the middle of that decade feel like for you?

嗯，这是个好问题。可能感觉就像今天一样。对我来说，总是有一些问题，有一些理由让人不耐烦，也总是有一些理由让你对现状感到满意，并且总有无数理由继续前进。所以我想，就像我刚才反思的那样，听起来就像今天早上。但我要说的是，在我们追求的所有事情中，首先你必须拥有核心信念。你必须从你最好的原则出发进行推理，理想情况下，你是从物理学原理或对行业的深刻理解或对科学的深刻理解出发进行推理。你从第一性原理推理，在某个时刻你必须相信一些东西。如果那些原则没有改变，假设也没有改变，那么就没有理由改变你的核心信念。在这个过程中，总会有一些成功的证据表明你正朝着正确的方向前进。有时你会长时间没有成功的证据，你可能需要稍微调整方向，但证据会来的。如果你觉得自己走在正确的方向上，我们就继续前进。为什么我们如此长期地坚持？答案实际上是相反的：没有理由不坚持，因为我们相信它。我相信英伟达已经超过 30 年了，我仍然每天在这里工作。没有根本性的理由让我改变我的信念体系。我从根本上相信，我们在革新计算方面所做的工作今天仍然正确，甚至比以往更加正确。所以我们会坚持下去，直到情况改变。当然，一路上会有非常困难的时期，当你投资某样东西而其他人都不相信它时，它花费了很多钱，也许投资者或其他人宁愿你保留利润或提高股价等等。但你必须相信你的未来，你必须投资于自己。我们如此深信这一点，以至于在它真正实现之前我们就投资了数百亿美元。是的，那是漫长的十年，但一路走来也很有趣。

Well, that's a good question. It probably felt like today, you know. To me, there's always some problem and there's some reason to be impatient, there's always some reason to be happy about where you are, and there's always many reasons to carry on. So I think, as I was reflecting a second ago, that sounds like this morning. But I would say that in all things we pursue, first you have to have core beliefs. You have to reason from your best principles, and ideally you're reasoning from principles of either physics or deep understanding of the industry or deep understanding of the science, wherever you're reasoning from. You reason from first principles, and at some point you have to believe something. If those principles don't change and the assumptions don't change, then there's no reason to change your core beliefs. And along the way, there's always some evidence of success and that you're leading in the right direction. Sometimes you go a long time without evidence of success, and you might have to course correct a little, but the evidence comes. And if you feel like you're going in the right direction, we just keep on going. The question of why did we stay so committed for so long—the answer is actually the opposite: there was no reason to not be committed, because we believed it. I've believed in Nvidia for 30 plus years, and I'm still here working every single day. There's no fundamental reason for me to change my belief system. I fundamentally believe that the work we're doing in revolutionizing computing is as true today, even more true today than it was before. And so we'll stick with it until otherwise. There are, of course, very difficult times along the way, when you're investing in something and nobody else believes in it, and it costs a lot of money, and maybe investors or others would rather you just keep the profit or improve the share price or whatever. But you have to believe in your future, you have to invest in yourself. We believed this so deeply that we invested tens of billions of dollars before it really happened. And yeah, it was 10 long years, but it was fun along the way.

你如何总结那些核心信念？你相信计算机应该如何工作以及它们能为我们做什么，这不仅让你度过了那十年，还让你现在做着你正在做的事情——我相信你正在为未来几十年下注？

How would you summarize those core beliefs? What is it that you believe about the way computers should work and what they can do for us that keeps you not only coming through that decade but also doing what you're doing now, making bets I'm sure you're making for the next few decades?

第一个核心信念是关于加速计算、并行计算与通用计算。我们会把两个这样的处理器加在一起做加速计算，我今天仍然相信这一点。第二个是认识到这些深度学习网络，这些在 2012 年进入公众视野的 DNN，有能力从大量不同类型的数据中学习模式和关系，并且如果它们变得越来越大，就能学习越来越细微的特征。让它们变得越来越大、越来越深或越来越宽是更容易的。所以架构的可扩展性在经验上是成立的。模型大小和数据大小越大，能学到的知识越多，这也是经验上成立的。如果是这样，那么限制在哪里？除非存在物理限制、架构限制或数学限制，但从未发现过。所以我们相信你可以扩展它。那么唯一剩下的问题是：你能从数据中学到什么？数据基本上是人类经验的数字版本。那么你能学到什么？你显然可以从图像中学习物体识别，可以从听声音中学习语音，甚至可以通过研究大量的字母和单词来学习语言、词汇、句法和语法。所以我们现在已经证明，AI 或深度学习有能力学习几乎任何模态的数据，并且可以转换到任何模态的数据。这意味着什么？你可以从文本到文本——比如总结一段话。你可以从文本到文本，进行语言翻译。你可以从文本到图像——那是图像生成。你可以从图像到文本——那是字幕生成。你甚至可以从氨基酸序列到蛋白质结构。未来，你可以从蛋白质到文字：这种蛋白质有什么功能？或者给我一个具有这些特性的蛋白质的例子，识别药物靶点。所以你可以看到，所有这些问题都即将被解决。你可以从文字到视频。为什么不能从文字到机器人的动作标记？从计算机的角度来看，这有什么不同？

The first core belief was about accelerated computing, parallel computing versus general-purpose computing. We would add two of those processors together and do accelerated computing, and I continue to believe that today. The second was the recognition that these deep learning networks, these DNNs that came to the public during 2012, have the ability to learn patterns and relationships from a whole bunch of different types of data, and that they can learn more and more nuanced features if they could be larger and larger. It's easier to make them larger and larger, make them deeper and deeper or wider and wider. So the scalability of the architecture is empirically true. The fact that model size and data size being larger and larger can learn more knowledge is also empirically true. So if that's the case, what are the limits? Not unless there's a physical limit or an architectural limit or mathematical limit, and it was never found. So we believe that you could scale it. Then the only other question is: what can you learn from data? Data is basically digital versions of human experience. So what can you learn? You obviously can learn object recognition from images, you can learn speech from just listening to sound, you can learn even languages and vocabulary and syntax and grammar all just by studying a whole bunch of letters and words. So we've now demonstrated that AI or deep learning has the ability to learn almost any modality of data and can translate to any modality of data. So what does that mean? You can go from text to text—right, summarize a paragraph. You can go from text to text, translate from language to language. You can go from text to images—that's image generation. You can go from images to text—that's captioning. You can even go from amino acid sequences to protein structures. In the future, you'll go from protein to words: what does this protein do? Or give me an example of a protein that has these properties, identifying a drug target. And so you could just see that all of these problems are around the corner to be solved. You can go from words to video. Why can't you go from words to action tokens for a robot? From the computer's perspective, how is it any different?

不同之处在于，它开启了一个充满机遇和问题的宇宙，我们可以去解决。这让我们非常兴奋。感觉我们正处于一场真正巨大变革的前夜。当我思考未来 10 年时，与过去 10 年不同，我知道我们已经经历了很多变化，但我认为我再也无法预测我将如何使用目前正在开发的技术了。

Different and so it opened up this universe of opportunities and universe of problems that we can go solve. And that gets us quite excited. It feels like we are on the cusp of this truly enormous change. When I think about the next 10 years, unlike the last 10 years, I know we've gone through a lot of change already, but I don't think I can predict anymore how I will be using the technology that is currently being developed.

完全正确。我认为过去 10 年实际上是关于 AI 的科学。未来 10 年，我们仍会有很多 AI 的科学，但未来 10 年将是 AI 的应用科学。基础科学与应用科学。因此，应用研究、AI 的应用方面现在变成了：我如何将 AI 应用于数字生物学？如何将 AI 应用于气候技术？如何将 AI 应用于农业、渔业、机器人、交通、物流优化？如何将 AI 应用于教学？如何将 AI 应用于播客？

That's exactly right. I think the last 10 years was really about the science of AI. The next 10 years, we're going to have plenty of science of AI, but the next 10 years is going to be the application science of AI. The fundamental science versus the application science. And so the applied research, the application side of AI now becomes: how can I apply AI to digital biology? How can I apply AI to climate technology? How can I apply AI to agriculture, to fishery, to robotics, to transportation, optimizing logistics? How can I apply AI to teaching? How do I apply AI to podcasting?

我想挑选其中几个来帮助人们看到，我们一直在谈论的计算领域的根本性变化将如何改变他们的生活体验，他们将如何实际使用基于我们刚才讨论的一切的技术。我经常听你谈到的一个话题，也是我特别感兴趣的，就是物理 AI，换句话说，机器人。我的朋友们，意思是人形机器人，但也包括自动驾驶汽车、智能建筑、自主仓库或自主割草机等机器人。据我所知，我们可能即将看到所有这些机器人能力的巨大飞跃，因为我们正在改变训练它们的方式。直到最近，你要么必须在现实世界中训练你的机器人，它可能会损坏或磨损，要么你可以从相当有限的来源获取数据，比如穿着动作捕捉服的人类。但这意味着机器人无法获得足够多的样本来更快地学习。但现在，我们开始在世界中训练机器人，这意味着每天更多的重复次数、更多的条件、更快的学习。所以我们现在可能正处于机器人的大爆炸时刻，而 Nvidia 正在构建实现这一目标的工具。你有 Omniverse，据我所知，这是帮助训练机器人系统的 3D 世界，这样它们就不需要在物理世界中训练了。

I'd love to choose a couple of those to help people see how this fundamental change in computing that we've been talking about is actually going to change their experience of their lives, how they're actually going to use technology that is based on everything we just talked about. One of the things that I've now heard you talk a lot about, and I have a particular interest in, is physical AI, or in other words, robots. My friends, meaning humanoid robots, but also robots like self-driving cars and smart buildings or autonomous warehouses or autonomous lawnmowers. From what I understand, we might be about to see a huge leap in what all of these robots are capable of because we're changing how we train them. Up until recently, you've either had to train your robot in the real world where it could get damaged or wear down, or you could get data from fairly limited sources like humans in motion capture suits. But that means that robots aren't getting as many examples as they'd need to learn more quickly. But now we're starting to train robots in digital worlds, which means way more repetitions a day, way more conditions, learning way faster. So we could be in a big bang moment for robots right now, and Nvidia is building tools to make that happen. You have Omniverse, and my understanding is this is 3D worlds that help train robotic systems so that they don't need to train in the physical world.

完全正确。你刚刚宣布了 Cosmos，这是一种让那个 3D 宇宙更加逼真的方法，这样你就可以获得各种不同的...如果我们在桌子上训练某样东西，桌子上有各种不同的光照，一天中不同的时间，机器人经历的各种不同的体验，这样它就能从 Omniverse 中获得更多。

That's exactly right. You just announced Cosmos, which is ways to make that 3D universe much more realistic, so you can get all kinds of different... if we're training something on this table, many different kinds of lighting on the table, many different times of day, many different experiences for the robot to go through, so that it can get even more out of Omniverse.

作为一个从小热爱《星际迷航》中的 Data、艾萨克·阿西莫夫的书，并梦想着机器人未来的孩子，我们如何从现在的机器人走向你所看到的机器人未来世界？

As a kid who grew up loving Data on Star Trek, Isaac Asimov's books, and just dreaming about a future with robots, how do we get from the robots that we have now to the future world that you see of robotics?

让我用语言模型，也许 ChatGPT，作为理解 Omniverse 和 Cosmos 的参考。首先，当 ChatGPT 刚出现时，它非常出色，能够根据你的提示生成文本。然而，尽管它很神奇，但它有产生幻觉的倾向，如果它说得太长，或者它大谈特谈一个它不了解的话题，它仍然会很好地生成看似合理的答案，只是没有基于事实。人们称之为幻觉。所以下一代，不久之后，它有了根据上下文进行调整的能力。你可以上传你的 PDF，现在它基于 PDF。PDF 成为事实依据。它可以查找搜索，然后搜索成为它的事实依据。在这之间，它可以推理如何产生你要求的答案。所以第一部分是生成式 AI，第二部分是事实依据。现在让我们进入物理世界，世界模型。我们需要一个基础模型，就像 ChatGPT 有一个核心基础模型，那是突破。为了让机器人对物理世界有智能，它必须理解重力、摩擦、惯性、几何和空间意识等。它必须理解一个物体即使在我移开视线时也放在那里，当我回来时它仍然在那里——物体恒存性。它必须理解因果关系：如果我把它倾斜，它会倒下。所以这些物理常识，如果你愿意，必须被捕获或编码到世界基础模型中，这样 AI 就有了世界常识。必须有人去创造它，这就是我们用 Cosmos 所做的。我们创建了一个世界语言模型。就像 GPT 是语言模型一样，这是一个世界模型。我们必须做的第二件事是，就像我们用 PDF 和上下文以及用事实依据来接地一样。我们用物理模拟来增强 Cosmos 的事实依据，因为 Omniverse 使用基于原理求解器的物理模拟。数学是牛顿物理学，是我们知道的数学。所有基本物理定律我们已经理解了很长时间，它们被编码到 Omniverse 中。这就是为什么 Omniverse 是一个模拟器。使用模拟器来接地或调整 Cosmos，我们现在可以生成无限数量的未来故事，它们基于物理事实。就像 PDF 或搜索加上 ChatGPT，我们可以生成无限数量的有趣事物，回答一大堆有趣的问题。Omniverse 加 Cosmos 的组合，你可以为物理世界做到这一点。

Let me use language models, maybe ChatGPT, as a reference for understanding Omniverse and Cosmos. So first of all, when ChatGPT first came out, it was extraordinary and it has the ability to generate text from your prompt. However, as amazing as it was, it has the tendency to hallucinate if it goes on too long or if it pontificates about a topic it is not informed about. It'll still do a good job generating plausible answers, it just wasn't grounded in the truth. People called it hallucination. And so the next generation, shortly after, it had the ability to be conditioned by context. So you could upload your PDF, and now it's grounded by the PDF. The PDF becomes the ground truth. It could look up search, and then the search becomes its ground truth. And between that, it could reason about how to produce the answer you're asking for. So the first part is a generative AI, and the second part is ground truth. Now let's come into the physical world, the world model. We need a foundation model just like ChatGPT had a core foundation model that was the breakthrough. In order for robotics to be smart about the physical world, it has to understand things like gravity, friction, inertia, geometric and spatial awareness. It has to understand that an object is sitting there even when I looked away, when I come back it's still sitting there—object permanence. It has to understand cause and effect: if I tip it, it'll fall over. So these kinds of physical common sense, if you will, has to be captured or encoded into a world foundation model so that the AI has world common sense. Somebody has to go create that, and that's what we did with Cosmos. We created a world language model. Just like GPT was a language model, this is a world model. The second thing we have to do is the same thing that we did with PDFs and context and grounding it with ground truth. The way we augment Cosmos with ground truth is with physical simulations, because Omniverse uses physics simulation which is based on principled solvers. The mathematics is Newtonian physics, it's the math we know. All of the fundamental laws of physics we've understood for a very long time, and it's encoded into Omniverse. That's why Omniverse is a simulator. Using the simulator to ground or condition Cosmos, we can now generate an infinite number of stories of the future, and they're grounded on physical truth. Just like between PDF or search plus ChatGPT, we can generate an infinite amount of interesting things, answer a whole bunch of interesting questions. The combination of Omniverse plus Cosmos, you could do that for the physical world.

为了向观众说明这一点：如果你在工厂里有一个机器人，你想让它学习它可以走的每一条路线，而不是手动走完所有那些可能需要几天时间并且对机器人造成很多磨损的路线，我们现在能够在数字世界中模拟所有路线，只需一小部分时间，并且模拟机器人可能面临的许多不同情况——黑暗、堵塞等。所以机器人现在学习得越来越快。在我看来，如果这样发展 10 年，未来可能与今天大不相同。你认为在不久的将来，人们将如何实际与这项技术互动？

To illustrate this for the audience: if you had a robot in a factory and you wanted to make it learn every route that it could take, instead of manually going through all of those routes which could take days and could be a lot of wear and tear on the robot, we're now able to simulate all of them digitally in a fraction of the time and in many different situations that the robot might face—it's dark, it's blocked, etc. So the robot is now learning much, much faster. It seems to me like the future might look very different than today if you play this out 10 years. How do you see people actually interacting with this technology in the near future?

Cleo，所有移动的东西总有一天都会变成机器人，而且很快就会实现。我们还要推着割草机到处走的想法已经有点愚蠢了。也许人们这样做是因为好玩，但没有必要。每辆车都将是机器人。人形机器人...实现这一目标所需的技术即将到来。

Cleo, everything that moves will be robotic someday, and it will be soon. The idea that we'll be pushing around a lawn mower is already kind of silly. Maybe people do it because it's fun, but there's no need to. And every car is going to be robotic. Human robots... the technology necessary to make it possible is just around the corner.

所以一切会动的东西都将是机器人。它们将在 Omniverse 和 Cosmos 中学习如何成为机器人，并生成所有这些合理的、物理上可行的未来。机器人将从这些未来中学习，然后进入物理世界。这完全一样。一个你被机器人包围的未来是确定的。我只是很兴奋能拥有自己的 R2-D2。当然，R2-D2 不会完全是那个罐子形状并滚来滚去。它可能会有不同的物理形态，但它始终是 R2。所以我的 R2 会一直跟着我。有时它在我的智能眼镜里，有时在我的手机里，有时在我的电脑里，它还在我的车里。所以 R2 一直和我在一起，包括我回家的时候，那里有一个物理版本的 R2。无论那个版本是什么，我们都会与 R2 互动。我认为我们一生都会拥有自己的 R2-D2，它和我们一起成长，这现在是确定的。

And so everything that moves will be robotic. They'll learn how to be a robot in Omniverse and Cosmos, and will generate all these plausible physically plausible futures. The robots will learn from them, and then they'll come into the physical world. It's exactly the same. A future where you're just surrounded by robots is for certain. I'm just excited about having my own R2-D2. Of course, R2-D2 wouldn't be quite the can that it is and roll around. It'll probably be a different physical embodiment, but it's always R2. So my R2 is going to go around with me. Sometimes it's in my smart glasses, sometimes it's in my phone, sometimes it's in my PC, it's in my car. So R2 is with me all the time, including when I get home, where I left a physical version of R2. Whatever that version happens to be, we'll interact with R2. I think the idea that we'll have our own R2-D2 for our entire life and it grows up with us, that's a certainty now.

是的。我认为很多新闻媒体在谈论这样的未来时，都会关注可能出错的地方。这有道理。确实有很多可能出错的地方。我们应该讨论可能出错的地方，这样我们才能防止它出错。

Yeah. I think a lot of news media, when they talk about futures like this, they focus on what could go wrong. And that makes sense. There is a lot that could go wrong. We should talk about what could go wrong so we could keep it from going wrong.

是的，这就是我们在节目中喜欢采取的方法：找出重大挑战，以便我们能够克服它们。

Yeah, that's the approach that we like to take on the show: what are the big challenges so that we can overcome them.

是的。当你担心这个未来时，你会考虑哪些方面？

Yeah. What buckets do you think about when you're worrying about this future?

嗯，有很多大家常谈的问题：偏见、有害性、幻觉。对自己一无所知的事情充满信心地发言，结果我们依赖了这些信息。这是一种生成虚假信息、假新闻或假图像的方式。当然还有冒充。它模仿人类做得太好了。它可以非常出色地模仿一个特定的人。所以我们需要关注的领域范围相当明确。有很多人在研究这个问题。一些与 AI 安全相关的问题需要深入的研究和工程。简单来说就是：它想做正确的事，但执行得不对，结果伤害了别人。例如，一辆自动驾驶汽车想好好驾驶，但传感器坏了，或者没检测到什么东西，或者转弯太急。它做得不好，做错了。所以需要做大量的工程来确保 AI 安全，通过确保产品正常运行。最后，如果 AI 想做好事但系统失败了怎么办？意思是 AI 想阻止某事发生，但就在它想做的时候，机器坏了。这和飞机上的飞行计算机有三个版本没什么不同。自动驾驶系统内部有三重冗余，然后还有两名飞行员，还有空中交通管制，还有其他飞行员在监视这些飞行员。所以 AI 安全系统必须作为一个社区来架构，使得这些 AI 正常工作，当它们不能正常工作时，不会让人处于危险之中，并且它们周围有足够的安全和安保系统来确保 AI 安全。所以这个讨论的范围非常巨大。我们必须把各个部分拆开，像工程师一样构建它们。

Well, there's a whole bunch of the stuff that everybody talks about: bias, toxicity, hallucination. Speaking with great confidence about something it knows nothing about, and as a result we rely on that information. That's a version of generating fake information, fake news, or fake images, whatever it is. Of course, impersonation. It does such a good job pretending to be a human. It could do an incredibly good job pretending to be a specific human. So the spectrum of areas we have to be concerned about is fairly clear. There's a lot of people who are working on it. Some of the stuff related to AI safety requires deep research and deep engineering. That's simply: it wants to do the right thing, it just didn't perform it right, and as a result hurt somebody. For example, a self-driving car that wants to drive nicely and properly, but somehow the sensor broke down, or it didn't detect something, or made too aggressive a turn. It did it poorly, it did it wrongly. So that's a whole bunch of engineering that has to be done to make sure that AI safety is upheld by making sure that the product functions properly. And then lastly, whatever happens if the AI wants to do a good job but the system failed. Meaning the AI wanted to stop something from happening, and it turned out just when it wanted to do it, the machine broke down. This is no different than a flight computer inside a plane having three versions of them. There's triple redundancy inside the system inside autopilots, and then you have two pilots, and then you have air traffic control, and then you have other pilots watching out for these pilots. So the AI safety systems have to be architected as a community such that these AIs work function properly, and when they don't function properly, they don't put people in harm's way, and that they have sufficient safety and security systems all around them to make sure that we keep AI safe. So this spectrum of conversation is gigantic. We have to take the parts apart and build them as engineers.

我们现在所处的这个时刻令人难以置信的一点是，我们不再像在 CPU 和顺序处理的世界中那样面临许多技术限制。我们不仅解锁了一种新的计算方式，还找到了一种持续改进的方法。并行处理具有与我们在 CPU 上所做的改进不同的物理特性。我很好奇：在当前世界中，你正在思考哪些科学或技术限制？

One of the incredible things about this moment that we're in right now is that we no longer have a lot of the technological limits that we had in a world of CPUs and sequential processing. We've unlocked not only a new way to do computing, but also a way to continue to improve. Parallel processing has a different kind of physics to it than the improvements that we were able to make on CPUs. I'm curious: what are the scientific or technological limitations that we face now in the current world that you're thinking a lot about?

嗯，归根结底，一切都取决于在能量限制下你能完成多少工作。所以这是一个物理极限。关于传输信息、翻转比特和传输比特的物理定律。最终，完成这些工作所需的能量限制了我们能做什么，而我们拥有的能量也限制了我们能做什么。我们远未达到阻碍进步的根本极限。与此同时，我们致力于建造更好、更节能的计算机。这台小电脑，大版本要 25 万美元。是的，这是小宝贝。是的，这是一台 AI 超级计算机。我交付的版本，这只是个原型，所以是个模型。第一个版本是 DGX-1。我在 2016 年交付给 OpenAI，当时是 25 万美元。比这个版本多需要 10,000 倍的功率和能量。而这个版本每瓦性能提高了六倍。我知道这很不可思议。我们进入了一个全新的世界，而这只是从 2016 年开始。所以八年后，我们将计算的能效提高了 10,000 倍。想象一下，如果我们能效提高 10,000 倍，或者汽车能效提高 10,000 倍，或者电灯泡能效提高 10,000 倍。我们的灯泡现在不是 100 瓦，而是用 10,000 分之一的能量产生同样的亮度。所以计算的能效，特别是我们一直在研究的 AI 计算，已经取得了令人难以置信的进步。这至关重要，因为我们想创造更智能的系统，并且想使用更多的计算来变得更聪明。所以完成工作的能效是我们的首要任务。

Well, everything in the end is about how much work you can get done within the limitations of the energy that you have. So that's a physical limit. The laws of physics about transporting information, flipping bits, and transporting bits. At the end of the day, the energy it takes to do that limits what we can get done, and the amount of energy that we have limits what we can get done. We're far from having any fundamental limits that keep us from advancing. In the meantime, we seek to build better and more energy-efficient computers. This little computer, the big version of it was $250,000. Yeah, this is little baby digits. Yeah, this is an AI supercomputer. The version that I delivered, this is just a prototype, so it's a mockup. The very first version was DGX-1. I delivered it to OpenAI in 2016, and that was $250,000. 10,000 times more power, more energy necessary than this version. And this version has six times more performance per watt. I know it's incredible. We're in a whole new world, and it's only since 2016. So eight years later, we've increased the energy efficiency of computing by 10,000 times. Imagine if we became 10,000 times more energy efficient, or if a car was 10,000 times more energy efficient, or an electric light bulb was 10,000 times more energy efficient. Our light bulb would be right now instead of 100 watts, 10,000 times less producing the same illumination. So the energy efficiency of computing, particularly for AI computing that we've been working on, has advanced incredibly. That's essential because we want to create more intelligent systems, and we want to use more computation to be smarter. So energy efficiency to do the work is our number one priority.

当我准备这次采访时，我和很多工程师朋友聊过，这个问题是他们非常想让我问的。所以你确实是在和你的同行说话。你展示了通过 CUDA 提高可访问性和抽象化的价值，让更多人能够以各种其他方式使用更多的计算能力。随着技术应用变得更加具体，我想到了 AI 中的 Transformer。对观众来说，Transformer 是一种非常流行的较新的 AI 结构，现在被用于你见过的很多工具中。它们之所以流行，是因为 Transformer 的结构...

When I was preparing for this interview, I spoke to a lot of my engineering friends, and this is a question that they really wanted me to ask. So you're really speaking to your people here. You've shown a value of increasing accessibility and abstraction with CUDA, and allowing more people to use more computing power in all kinds of other ways. As applications of technology get more specific, I'm thinking of Transformers in AI, for example. For the audience, a Transformer is a very popular more recent structure of AI that's now used in a huge number of the tools that you've seen. The reason that they're popular is because Transformers are structured...

这种方式能帮助它们关注关键信息，并给出更好的结果。你可以构建完全针对某一种 AI 模型的芯片，但那样做会让它们做其他事情的能力变弱。所以，随着这些特定的 AI 结构或架构越来越流行，我的理解是，存在一场争论：你是把这些赌注押在将特定任务固化到芯片上，还是设计更通用的硬件？我的问题是，你如何下这些赌注？你是认为解决方案应该是一辆能去任何地方的汽车，还是真正优化一列从 A 到 B 的火车？你下的赌注风险巨大，我很好奇你是怎么想的。

In a way that helps them pay attention to key bits of information and give much better results. You could build chips that are perfectly suited for just one kind of AI model, but if you do that, then you're making them less able to do other things. So as these specific structures or architectures of AI get more popular, my understanding is there's a debate between how much you place these bets on burning them into the chip or designing hardware that is very specific to a certain task versus staying more general. So my question is: how do you make those bets? How do you think about whether the solution is a car that could go anywhere, or it's really optimizing a train to go from A to B? You're making bets with huge stakes, and I'm curious how you think about that.

这又回到了你的问题：你的核心信念是什么？核心信念要么是 Transformer 是最后一个 AI 算法，最后一个研究者会再次发现的 AI 架构；要么是 Transformer 是迈向未来几乎无法辨认的 Transformer 演变的垫脚石。我们相信后者。原因是你只需回顾历史，问问自己：在计算机算法、软件、工程和创新的世界里，有哪一个想法能持续那么久？答案是否定的。这正是计算机的本质之美：它今天能做一些 10 年前没人能想象到的事情。如果你 10 年前把计算机变成了微波炉，那么为什么应用会不断涌现？所以我们相信创新和发明的丰富性。我们希望创建一个架构，让发明家、创新者、软件程序员和 AI 研究者能在其中畅游，提出惊人的想法。看看 Transformer：其基本特征是注意力机制。它说 Transformer 会理解每个词与其他每个词的含义和相关性。如果你有 10 个词，它必须找出它们之间的关系。但如果你有 10 万个词，或者你的上下文大到阅读一个 PDF 或一堆 PDF，上下文窗口达到一百万个 token，那么处理所有这些关系是不可能的。解决这个问题的方法是各种新想法：flash attention、hierarchical attention、wave attention——我前几天刚读到。自 Transformer 以来发明的注意力机制种类之多令人惊叹。所以我认为这还会继续。我们相信计算机科学没有终结，AI 研究没有放弃，我们也没有放弃。拥有一台能够实现研究、创新和新想法灵活性的计算机，从根本上来说是最重要的。

That now comes back to exactly your question: what are your core beliefs? The core belief is either that Transformer is the last AI algorithm, the last AI architecture that any researcher will ever discover again, or that Transformers is a stepping stone towards evolutions of Transformers that are barely recognizable as a Transformer years from now. We believe the latter. The reason is that you just have to go back in history and ask yourself: in the world of computer algorithms, software, engineering, and innovation, has one idea stayed that long? The answer is no. That's the essential beauty of a computer: it's able to do something today that no one even imagined possible 10 years ago. If you had turned that computer 10 years ago into a microwave, then why would the applications keep coming? So we believe in the richness of innovation and invention. We want to create an architecture that lets inventors, innovators, software programmers, and AI researchers swim in the soup and come up with amazing ideas. Look at Transformers: the fundamental characteristic is the attention mechanism. It says the Transformer will understand the meaning and relevance of every single word with every other word. If you had 10 words, it has to figure out the relationship across 10 of them. But if you have 100,000 words, or your context is now as large as reading a PDF or a whole bunch of PDFs, and the context window is like a million tokens, processing all of it across all of it is just impossible. The way you solve that problem is with all kinds of new ideas: flash attention, hierarchical attention, wave attention—I just read about that the other day. The number of different types of attention mechanisms invented since the Transformer is quite extraordinary. So I think that's going to continue. We believe computer science hasn't ended, AI research hasn't given up, and we haven't given up. Having a computer that enables the flexibility of research, innovation, and new ideas is fundamentally the most important thing.

我特别好奇的一点是：你设计芯片，有公司组装芯片，有公司设计硬件使其能在纳米尺度工作。当你设计这样的工具时，你如何考虑当前物理上可能的设计？你在推动极限时考虑哪些事情？

One of the things I am just so curious about: you design the chips, there are companies that assemble the chips, there are companies that design hardware to make it possible to work at nanometer scale. When you're designing tools like this, how do you think about design in the context of what's physically possible right now? What are the things that you're thinking about with sort of pushing that limit today?

我们的做法是：即使我们有东西被制造出来——例如，我们的芯片由台积电制造——我们假设我们需要拥有台积电那样的深厚专业知识。所以我们公司里有非常擅长半导体物理的人，这样我们就能对当今半导体物理的极限有感觉、有直觉。然后我们与他们紧密合作，共同发现极限，因为我们试图推动极限。所以我们一起发现极限。我们在系统工程和冷却系统方面也做同样的事情。事实证明，管道对我们来说非常重要，因为液冷；风扇可能也很重要，因为风冷。我们试图以近乎空气动力学合理的方式设计这些风扇，以便能以最低噪音通过最大风量。所以我们公司里有空气动力学工程师。即使我们不制造它们，我们设计它们，并且我们必须拥有知道如何制造它们的深厚专业知识。由此，我们努力推动极限。

The way we do it is: even though we have things made—for example, our chips are made by TSMC—we assume that we need to have the deep expertise that TSMC has. So we have people in our company who are incredibly good at semiconductor physics, so that we have a feeling for, an intuition for, what are the limits of today's semiconductor physics. Then we work very closely with them to discover the limits, because we're trying to push the limits. So we discover the limits together. Now we do the same thing in system engineering and cooling systems. It turns out plumbing is really important to us because of liquid cooling, and maybe fans are really important to us because of air cooling. We're trying to design these fans in a way almost like they're aerodynamically sound, so that we could pass the highest volume of air with the least amount of noise. So we have aerodynamics engineers in our company. Even though we don't make them, we design them, and we have to have deep expertise of knowing how to have them made. From that, we try to push the limits.

这次对话的一个主题是，你是一个对未来下大赌注的人，而且一次又一次地赌对了。我们谈到了 GPU、CUDA、你在 AI、自动驾驶汽车上的赌注，而且我们将在机器人领域再次正确。这就是我的问题：我们会是对的。最新的赌注，当然，我们刚刚在 CES 上描述过，我对此非常自豪和兴奋，那就是 Omniverse 和 Cosmos 的融合，这样我们就有了这种新型的生成式世界生成系统，这个多元宇宙生成系统。我认为这在机器人和物理系统的未来将极其重要。当然，我们在人形机器人方面的工作，开发工具系统、训练系统和人类演示系统，所有这些你已经提到过的东西——我们才刚刚开始看到这些工作的开端。我认为未来 5 年，人形机器人领域将非常有趣。当然，我们在数字生物学方面的工作，以便我们能理解分子的语言和细胞的语言，就像我们理解物理和物理世界的语言一样，我们想理解人体的语言和生物学的语言。所以如果我们能学会并预测它，那么突然之间，我们拥有人类数字孪生的能力就变得可行了。所以我对这项工作非常兴奋。我喜欢我们在气候科学方面的工作，能够从天气预报中理解和预测高分辨率区域气候，你头顶一公里内的天气模式，我们能够以很高的精度预测它——其影响确实非常深远。所以我们正在做的很多事情都非常酷。你知道，我们很幸运，我们创造了这个仪器，它是一台时间机器。

One of the themes of this conversation is that you are a person who makes big bets on the future, and time and time again you've been right about those bets. We've talked about GPUs, we've talked about CUDA, we've talked about bets you've made in AI, self-driving cars, and we're going to be right on robotics. And this is my question: we're going to be right. The latest bet, of course, we just described at CES, and I'm very proud of it and very excited about it, is the fusion of Omniverse and Cosmos, so that we have this new type of generative world generation system, this multiverse generation system. I think that's going to be profoundly important in the future of robotics and physical systems. Of course, the work that we're doing with human robots, developing the tooling systems and the training systems and the human demonstration systems, all of this stuff that you've already mentioned—we're just seeing the beginnings of that work. I think the next 5 years are going to be very interesting in the world of human robotics. Of course, the work that we're doing in digital biology, so that we can understand the language of molecules and understand the language of cells, and just as we understand the language of physics and the physical world, we'd like to understand the language of the human body and understand the language of biology. So if we can learn that and we can predict it, then all of a sudden our ability to have a digital twin of the human is plausible. So I'm very excited about that work. I love the work that we're doing in climate science, and being able to from weather predictions understand and predict the high-resolution regional climates, the weather patterns within a kilometer above your head, that we can somehow predict that with great accuracy—its implications are really quite profound. So the number of things that we're working on is really cool. You know, we're fortunate that we've created this instrument that is a time machine.

如果有人正在看这个视频，也许他们知道英伟达是一家极其重要的公司，但并不完全理解为什么或它如何影响他们的生活。现在他们可能更好地理解了过去几十年我们在计算领域经历的巨大转变，以及我们目前所处的这个非常激动人心、有点奇怪的时刻，我们正处在许多不同事物的边缘。如果他们想稍微展望一下未来，你会建议他们如何准备或思考他们个人所处的这个时刻，以及这些工具将如何真正影响他们？

If someone is watching this and maybe they came into this video knowing that Nvidia is an incredibly important company but not fully understanding why or how it might affect their life, and they're now hopefully better understanding a big shift that we've gone through over the last few decades in computing, this very exciting, very sort of strange moment that we're in right now where we're sort of on the precipice of so many different things. If they would like to be able to look into the future a little bit, how would you advise them to prepare or to think about this moment that they're in personally with respect to how these tools are actually going to affect them?

嗯，有几种方式可以思考我们正在创造的未来。一种思考方式是：假设你从事的工作仍然重要，但完成它所付出的努力从一周缩短到几乎瞬间完成。辛苦的工作基本上降为零。这意味着什么？这很像如果我们突然在这个国家有了高速公路会发生的变化。上次工业革命就发生了类似的事情。突然之间我们有了州际高速公路。当你有州际高速公路时会发生什么？郊区开始出现，货物从东到西的运输不再是问题，高速公路上开始出现加油站、快餐店、汽车旅馆，因为人们跨州或跨国旅行时想找个地方休息几小时或过夜。所以突然之间，新的经济体和新的能力出现了。如果视频会议让我们无需旅行就能见面，会发生什么？突然之间，在离家很远的地方工作和生活变得可行。所以你问自己这些问题：如果我身边一直有一个软件程序员，无论我想出什么，他都能为我编写，那会怎样？那会带来什么？如果我只有一个想法的种子，我粗略地勾勒出来，然后突然一个产品原型就摆在我面前，那会怎样？那会如何改变我的生活和机会？它会让我能够做什么？等等等等。所以我认为，在未来十年，智能——不是所有方面，但某些方面——基本上会变得超人类。我可以确切告诉你那是什么感觉。我身边都是超人类的人，从我的角度看是超级智能，因为他们是各自领域的世界顶尖，做得比我好得多。我身边有成千上万个这样的人。但从来没有一天让我觉得自己不再必要。这实际上赋予我力量，给我信心去应对更雄心勃勃的事情。所以假设现在每个人身边都有这些超级 AI，它们非常擅长特定事情或某些事情。那会让你感觉如何？它会赋予你力量，让你感到自信。我敢肯定你可能用过 ChatGPT 和 AI，我今天感到更有力量，更有信心去学习。几乎所有领域的知识，理解它的障碍都降低了。我身边一直有一个私人导师。所以我认为这种感觉应该是普遍的。如果有一件事我鼓励大家去做，那就是立刻给自己找一个 AI 导师。那个 AI 导师当然可以教你任何你喜欢的东西，帮助你编程、写作、分析、思考、推理。所有这些都会让你真正感到有力量。我认为那将是我们的未来。我们会成为超人，不是因为我们有超能力，而是因为我们有超级 AI。

Well, there are several ways to reason about the future that we're creating. One way to reason about it is: suppose the work that you do continues to be important, but the effort by which you do it went from being a week long to almost instantaneous. The effort of drudgery basically goes to zero. What is the implication of that? This is very similar to what would change if all of a sudden we had highways in this country. That kind of happened in the last Industrial Revolution. All of a sudden we have interstate highways. When you have interstate highways, what happens? Suburbs start to be created, distribution of goods from east to west is no longer a concern, gas stations start cropping up on highways, fast food restaurants show up, motels show up because people traveling across the state or country want to stay somewhere for a few hours or overnight. So all of a sudden, new economies and new capabilities emerge. What would happen if a video conference made it possible for us to see each other without having to travel anymore? That all of a sudden it's okay to work far away from home and live further away. So you ask yourself these questions: what would happen if I have a software programmer with me all the time, and whatever I can dream up, the software programmer could write for me? What would that do? What would happen if I just had a seed of an idea, and I rough it out, and all of a sudden a prototype of a product was put in front of me? How would that change my life and my opportunity? What would it free me to be able to do? And so on and so forth. So I think that in the next decade, intelligence—not for everything, but for some things—would basically become superhuman. I can tell you exactly what that feels like. I'm surrounded by superhuman people, super intelligence from my perspective, because they're the best in the world at what they do, and they do what they do way better than I can do it. I'm surrounded by thousands of them. Yet it never once caused me to think that I'm no longer necessary. It actually empowers me and gives me the confidence to go tackle more and more ambitious things. So suppose now everybody is surrounded by these super AIs that are very good at specific things or good at some things. What would that make you feel? It's going to empower you, it's going to make you feel confident. I'm pretty sure you probably use ChatGPT and AI, and I feel more empowered today, more confident to learn something. The knowledge of almost any particular field, the barriers to understanding it, have been reduced. I have a personal tutor with me all the time. So I think that feeling should be universal. If there's one thing that I would encourage everybody to do, it's to go get yourself an AI tutor right away. That AI tutor could of course just teach you things, anything you like, help you program, help you write, help you analyze, help you think, help you reason. All of those things are going to really make you feel empowered. I think that's going to be our future. We're going to become super humans, not because we have super powers, but because we have super AIs.

你能给我们介绍一下这些物品吗？这是一款新的 GeForce 显卡。

Could you tell us a little bit about each of these objects? This is a new GeForce graphics card.

是的，这是 RTX 50 系列。它本质上是一台可以放进你 PC 的超级计算机。我们当然用它来玩游戏。如今人们也用它进行设计和创意艺术，它还能做惊人的 AI。真正的突破——这确实是一件了不起的事情——GeForce 实现了 AI。它让杰夫·辛顿、亚历克斯·克里热夫斯基和伊利亚·苏茨克弗能够训练 AlexNet。我们发现了 AI 并推动了 AI 的发展。然后 AI 又回到 GeForce 来帮助计算机图形学。神奇之处在于：在 4K 显示屏的大约 800 万个像素中，我们只计算了其中的 50 万个。其余的我们用 AI 来预测——AI 猜出来了——但图像是完美的。我们用计算出的 50 万个像素来告知 AI，并且我们对每一个像素都进行了光线追踪，一切都非常漂亮，完美无缺。然后我们告诉 AI：如果屏幕上有这 50 万个完美像素，那么其他 800 万个像素是什么？它就去填充屏幕的其余部分，而且完美无缺。如果你只需要处理更少的像素，你就可以投入更多资源，因为你要处理的更少，所以质量更好。AI 所做的外推正是因为无论你拥有多少计算注意力、多少资源，你现在都可以将它们投入到 50 万个像素中。这是一个完美的例子，说明为什么 AI 会让我们都成为超人：因为它能做的所有其他事情，它都会为我们做，让我们把时间和精力集中在真正有价值的事情上。所以我们把自己的资源——精力密集、注意力密集、专注于那几十万个像素——用 AI 来超分辨率、提升分辨率到其他所有事物。所以这款显卡现在主要由 AI 驱动，内部的计算机图形技术也非常出色。

Yes, and this is the RTX 50 Series. It is essentially a supercomputer that you put into your PC. We use it for gaming, of course. People today use it for design and creative arts, and it does amazing AI. The real breakthrough here—and this is truly an amazing thing—GeForce enabled AI. It enabled Jeff Hinton, Alex Krizhevsky, and Ilya Sutskever to train AlexNet. We discovered AI and we advanced AI. Then AI came back to GeForce to help computer graphics. Here's the amazing thing: out of 8 million pixels or so in a 4K display, we are computing only 500,000 of them. The rest we use AI to predict—the AI guessed it—and yet the image is perfect. We inform it by the 500,000 pixels that we computed, and we ray traced every single one, and it's all beautiful, it's perfect. Then we tell the AI: if these are the 500,000 perfect pixels in this screen, what are the other 8 million? And it goes and fills in the rest of the screen, and it's perfect. If you only have to do fewer pixels, you can invest more in doing that because you have fewer to do, so the quality is better. The extrapolation that the AI does exactly because whatever computing attention you have, whatever resources you have, you can place it into 500,000 pixels now. This is a perfect example of why AI is going to make us all superhuman: because all of the other things that it can do, it'll do for us, allowing us to take our time and energy and focus it on the really, really valuable things that we do. So we'll take our own resource, which is energy-intensive, attention-intensive, and well-dedicated to the few hundred thousand pixels, and use AI to super-res it, up-res it, to everything else. So this graphics card is now powered mostly by AI, and the computer graphics technology inside is incredible as well.

研究人员和我们把第一台交给了 OpenAI，Elon 当时在场接收。这个版本，我做了个迷你版。原因是 AI 已经从 AI 研究人员扩展到每个工程师、每个学生、每个 AI 科学家，AI 将无处不在。所以我们不再做 25 万美元的版本，而是做 3000 美元的版本。学校可以有，学生可以有。你把它放在你的 PC 或 Mac 旁边，突然之间你就有了自己的 AI 超级计算机。你可以开发和构建 AI，打造你自己的 AI，打造你自己的 R2-D2。

Researchers and we delivered the first one to OpenAI and Elon was there to receive it. This version, I built a mini version. The reason for that is because AI has now gone from AI researchers to every engineer, every student, every AI scientist, and AI is going to be everywhere. So instead of these $250,000 versions, we're going to make these $3,000 versions. Schools can have them, students can have them. You set it next to your PC or Mac, and all of a sudden you have your own AI supercomputer. You can develop and build AIs, build your own AI, build your own R2-D2.

你觉得对这个观众来说，有什么重要的事情我还没问到？

What do you feel is important for this audience to know that I haven't asked?

我建议最重要的事情之一是，比如如果我现在是个学生，第一件事就是学习 AI。如何学习与 ChatGPT 交互？如何学习与 Gemini Pro 交互？如何学习与 Grok 交互？学习与 AI 交互就像成为一个非常擅长提问的人。你非常擅长提问，而提示 AI 非常相似。你不能随便问一堆问题。让 AI 成为你的助手需要一些专业知识和艺术性来提示它。所以如果我现在是个学生，无论我学的是数学、科学、化学、生物学，还是任何科学领域，或者任何职业，我都会问自己：我如何用 AI 把工作做得更好？如果我想成为律师，我如何用 AI 成为更好的律师？如果我想成为医生，我如何用 AI 成为更好的医生？如果我想成为化学家，我如何用 AI 成为更好的化学家？如果我想成为生物学家，我如何用 AI 成为更好的生物学家？这个问题应该贯穿每个人。就像我们这一代是必须问自己如何用电脑把工作做得更好的第一代。我们之前的一代没有电脑。我们这一代是第一代必须问这个问题的人：我如何用电脑把工作做得更好？我在 Windows 95 之前进入这个行业，1984 年办公室还没有电脑。之后不久电脑开始出现，所以我们不得不问自己如何用电脑把工作做得更好。下一代不需要问那个问题，但必须问下一个问题：我如何用 AI 把工作做得更好？这就是开始和结束。我认为对每个人来说，这是一个既令人兴奋又可怕，因此值得一问的问题。我认为这会非常有趣。

One of the most important things I would advise is, for example, if I were a student today, the first thing I would do is to learn AI. How do I learn to interact with ChatGPT? How do I learn to interact with Gemini Pro? And how do I learn to interact with Grok? Learning how to interact with AI is not unlike being someone who is really good at asking questions. You're incredibly good at asking questions, and prompting AI is very similar. You can't just randomly ask a bunch of questions. Asking an AI to be an assistant to you requires some expertise and artistry in how to prompt it. So if I were a student today, irrespective of whether it's for math, science, chemistry, biology, or any field of science I'm going into, or any profession, I'm going to ask myself: how can I use AI to do my job better? If I want to be a lawyer, how can I use AI to be a better lawyer? If I want to be a doctor, how can I use AI to be a better doctor? If I want to be a chemist, how do I use AI to be a better chemist? If I want to be a biologist, how do I use AI to be a better biologist? That question should be persistent across everybody. Just as my generation grew up as the first generation that had to ask ourselves how we can use computers to do our jobs better. The generation before us had no computers. My generation was the first generation that had to ask the question: how do I use computers to do my job better? I came into the industry before Windows 95, right? 1984, there were no computers in offices. Shortly after that, computers started to emerge, so we had to ask ourselves how we use computers to do our jobs better. The next generation doesn't have to ask that question, but it has to ask the next question: how can I use AI to do my job better? That is start and finish. I think for everybody, it's a really exciting and scary and therefore worthwhile question. I think it's going to be incredibly fun.

AI 显然是一个人们现在才刚刚学习的词，但它让你的电脑变得更容易使用。提示 ChatGPT 问任何问题比你自己去做研究要容易得多。所以我们降低了理解的门槛，降低了知识的门槛，降低了智能的门槛。每个人都真的需要去尝试一下。非常疯狂的是，如果我把一台电脑放在一个从未用过电脑的人面前，他们不可能在一天内学会使用电脑。必须有人教他们。然而对于 ChatGPT，如果你不知道如何使用它，你只需要输入“我不知道如何使用 ChatGPT，告诉我”，它就会回来给你一些例子。所以这就是神奇之处。智能的奇妙之处在于它会一路帮助你，让你变得超级。

AI is obviously a word that people are just learning now, but it has made your computer so much more accessible. It is easier to prompt ChatGPT to ask it anything you like than to go do the research yourself. So we've lowered a barrier of understanding, we've lowered a barrier of knowledge, we've lowered a barrier of intelligence. Everybody really has to just go try it. The thing that's really crazy is if I put a computer in front of somebody who has never used a computer, there is no chance they're going to learn that computer in a day. Somebody really has to show it to you. Yet with ChatGPT, if you don't know how to use it, all you have to do is type in 'I don't know how to use ChatGPT, tell me' and it will come back and give you some examples. So that's the amazing thing. The amazing thing about intelligence is it'll help you along the way and make you super.

如果你有时间，我还有一个问题。这不是我计划问你的，但在来的路上，我有点害怕飞机，这不是我最合理的品质。这趟航班有点颠簸，非常颠簸。我坐在那里，飞机在晃动，我在想我的葬礼上他们会说什么。‘她问了好问题’——墓碑上会这么写。我希望如此。然后‘我爱我的丈夫、朋友和家人’，我希望他们会谈论的是乐观。我希望他们能认识到我在这里努力做的事情。我很好奇你：你从事这个行业很久了，感觉你在前面的愿景中描述了很多。你希望人们如何评价你努力做的事情？

I have one more question if you have a second. This is not something I planned to ask you, but on the way here, I'm a little bit afraid of planes, which is not my most reasonable quality. The flight here was a little bit bumpy, very bumpy. I'm sitting there and it's moving, and I'm thinking about what they're going to say at my funeral. 'After she asked good questions' — that's what the tombstone's going to say. I hope so. And after 'I loved my husband and my friends and my family', the thing that I hoped they would talk about was optimism. I hope they would recognize what I'm trying to do here. I'm very curious for you: you've been doing this a long time, it feels like there's so much that you've described in this vision ahead. What would the theme be that you would want people to say about what you're trying to do?

很简单，他们产生了非凡的影响。我认为我们很幸运，因为很久以前的一些核心信念，坚持这些信念并在此基础上发展。今天我们发现自己成为世界上最重要、最具影响力的科技公司之一，甚至可能是历史上最重要的。我们非常认真地对待这个责任。我们努力确保我们创造的能力能够提供给大公司以及各个科学领域的个人研究人员和开发者，无论是否盈利、大小、知名与否。正是因为我们理解我们所做工作的重要性及其对许多人的潜在影响，我们才希望尽可能广泛地普及这种能力。我确实认为，当我们几年后回顾时，我希望下一代意识到的是，首先，他们会因为我们创造的所有游戏技术而认识我们。但我确实认为我们会回顾，整个数字生物学和生命科学领域已经被改变，我们对材料科学的整个理解已经被革命，机器人正在帮助我们做危险和单调的事情，如果我们想开车就可以开车，否则你可以小睡或像家庭影院一样享受你的车，从工作地读到回家。那时你希望自己住得远，这样你就能在车里待更长时间。你回顾并意识到有一家公司几乎处于这一切的中心，而且恰好是你从小玩游戏时认识的公司。我希望这就是下一代学到的。

Very simply, they made an extraordinary impact. I think we're fortunate because of some core beliefs a long time ago and sticking with those core beliefs and building upon them. We found ourselves today being one of the most important and consequential technology companies in the world, and potentially ever. We take that responsibility very seriously. We work hard to make sure that the capabilities we've created are available to large companies as well as individual researchers and developers across every field of science, no matter profitable or not, big or small, famous or otherwise. It's because of this understanding of the consequential work we're doing and the potential impact it has on so many people that we want to make this capability as pervasive as possible. I do think that when we look back in a few years, and I do hope that what the next generation realized is, well first of all, they're going to know us because of all the gaming technology we create. But I do think we'll look back and the whole field of digital biology and life sciences has been transformed, our whole understanding of material sciences has been revolutionized, robots are helping us do dangerous and mundane things all over the place, if we wanted to drive we can drive but otherwise you know take a nap or enjoy your car like it's a home theater, read from work to home. At that point you're hoping you live far away so you could be in a car for longer. You look back and realize there's this company almost at the epicenter of all of that, and it happens to be the company that you grew up playing games with. I hope that to be what the next generation learns.

非常感谢你的时间。我很享受。

Thank you so much for your time. I enjoyed it.

谢谢。我很高兴。

Thank you. I'm glad.