How Hard Could It Be?

The following is a conversation with Jensen Huang, CEO of NVIDIA, one of the most important and influential companies in the history of human civilization.

下面这场对话的嘉宾是 NVIDIA CEO 黄仁勋——NVIDIA 是人类文明史上最重要、最有影响力的公司之一。

NVIDIA is the engine powering the AI revolution, and a lot of its success can be directly attributed to Jensen's sheer force of will and his many brilliant bets and decisions as a leader, engineer, and innovator.

NVIDIA 是驱动这轮 AI 革命的引擎,而它的成功很大一部分可以直接归因于黄仁勋那股纯粹的意志力,以及他作为领导者、工程师和创新者做出的一连串精彩下注和决定。

This is Lex Fridman Podcast. And now dear friends, here's Jensen Huang.

这里是 Lex Fridman Podcast。各位朋友,有请黄仁勋。

You've propelled NVIDIA into a new era in AI, moving beyond his focus on chip scale design to now rack scale design.

你已经把 NVIDIA 推进了 AI 的一个新时代——从原来只盯着芯片级设计,走到了今天的机架级设计。

And I think it's fair to say that winning for NVIDIA for a long time used to be about building the best GPU possible, and you still do, but now you've expanded that to extreme co-design of GPU, CPU memory, networking, storage, power cooling, software, the rack itself, the pod that you've announced, and even the data center.

可以说在很长一段时间里,NVIDIA 的胜负就在于造出尽可能好的 GPU——你们现在依然在做这件事,但已经把它扩展成了极限协同设计:GPU、CPU 内存、网络、存储、供电散热、软件、机架本身、你们发布的那个 pod,甚至整个数据中心。

So let's talk about extreme co-design. What is the hardest part of co-designing a system with that many complex components and design variables?

我们就从极限协同设计聊起。当一个系统里有这么多复杂部件和设计变量,协同设计最难的部分是什么?

Yeah, thanks for that question. So first of all, the reason why extreme co-design is necessary is because the problem no longer fits inside one computer to be accelerated by one GPU.

这个问题问得好。首先,极限协同设计之所以必要,是因为问题已经装不进一台计算机、也不可能靠一块 GPU 来加速了。

The problem that you're trying to solve is you would like to go faster than the number of computers that you add.

你要解的问题是:让性能提升的幅度超过你增加的计算机数量。

So you added 10,000 computers, but you would like it to go a million times faster.

你加了一万台计算机,但你希望它快一百万倍。

Then all of a sudden you have to take the algorithm, you have to break up the algorithm, you have to refactor it, you have to shard the pipeline, you have to shard the data, you have to shard the model.

于是你突然就得动算法了——把算法拆开、重构,把流水线切分、把数据切分、把模型切分。

Now all of a sudden when you distribute the problem this way, not just scaling up the problem, but you're distributing the problem, then everything gets in the way.

而当你这样把问题分布出去——不只是把问题放大,而是把它分散开——所有东西都开始挡路。

This is the Amdahl's law problem where the amount of speed up you have for something depends on how much of the total workload it is.

这就是阿姆达尔定律的问题:某一部分能带来多少整体加速,取决于它占总负载的比例。

And so if computation represents 50% of the problem, and I sped up computation infinitely like a million times, you know, I only sped up the total workload by a factor of two.

如果计算只占问题的 50%,那我就算把计算加速到无穷、快一百万倍,总负载也只快了两倍。

Now all of a sudden, not only do you have to distribute a computation, you have to shard the pipeline somehow.

所以你不只要把计算分布出去,还得想办法把流水线切分开。

You also have to solve the networking problem because you've got all of these computers are all connected together.

你还得解决网络问题,因为这些计算机全都是连在一起的。

And so distributed computing at the scale that we do, the CPU is a problem, the GPU is a problem, the networking is a problem, the switching is a problem.

所以在我们这个量级做分布式计算,CPU 是问题,GPU 是问题,网络是问题,交换是问题。

And distributing the workload across all these computers is a problem.

把负载分配到所有这些计算机上,也是问题。

It's just a massively complex computer science problem. And so we just gotta bring every technology to bear.

这就是一个极其庞杂的计算机科学问题。所以我们必须把手里每一项技术都调动起来。

Otherwise, we scale up linearly or we scale up based on the capabilities of Moore's Law, which has largely slowed because Dennard scaling has slowed.

否则我们只能线性扩展,或者只能按摩尔定律的能力扩展——而摩尔定律已经大幅放缓,因为 Dennard scaling 放缓了。

I'm sure there's trade-offs there. Plus you have a complete disparate disciplines here.

这里面肯定有权衡。而且你面对的是完全不同的学科。

I'm sure you have specialists in each one of these high bandwidth memory, the network and the NVLink, the NICs, the optics and the copper that you're doing, the power delivery, the cooling, all of that.

你肯定在每个方向上都有专家:高带宽内存、网络和 NVLink、网卡、你们在做的光互连和铜互连、供电、散热,所有这些。

I mean, there's like world experts in each of those. How do you get 'em in a room together to figure out-

每一个方向上都有世界级专家。你怎么把他们弄到一个房间里,一起把问题想清楚——

That's why my staff is so large. Yeah.

这就是为什么我的直属团队这么大。

What's the pro- can you take me through the process of the specialists and the generalists?

这个流——你能带我过一遍,专家和通才是怎么配合的吗?

Like how do you put together the rack when you know the s- the set of things you have to shove into a rack together? Like what does that process look like of designing it all together?

当你已经知道有哪些东西必须塞进同一个机架,你是怎么把这个机架拼起来的?把这一切放在一起设计,过程大概是什么样?

Yeah. There's the first question, which is: what is extreme co-design?

先说第一个问题:什么是极限协同设计?

We're optimizing across the entire stack of software from architectures to chips, to systems, to system software, to the algorithms, to the applications. That's one layer.

我们是在整个栈上做优化——从架构到芯片、到系统、到系统软件、到算法、再到应用。这是一层。

The second thing that you and I just talked about goes beyond CPUs and GPUs and networking chips and scale up switches and scale out switches.

第二件事就是你和我刚才聊的那个——它超出了 CPU、GPU、网络芯片,以及 scale-up 交换机和 scale-out 交换机的范围。

And then of course, you gotta include power and cooling and all of that because all these computers are extremely power hungry.

然后当然还得把供电和散热这些都算进去,因为这些计算机极其耗电。

They do a lot of work and they're very energy efficient, but they in aggregate still consume a lot of power.

它们干的活很多,能效也很高,但加总起来功耗还是很大。

And so that's one. The first question is, what is it?

这是第一点。第一个问题是:它是什么?

The second question is, why is it, and we just spoke about the reason, you know you want to distribute the workload so that you can exceed the benefit of just increasing the number of computers.

第二个问题是:它为什么存在?原因我们刚才讲过了——你要把负载分布出去,好让收益超过单纯增加计算机数量所能带来的那点提升。

And the, and then the third question is, how is it, how do you do it?

第三个问题是:它是怎么做的?你具体怎么做?

And, and that's the, that's kind of the miracle of this company.

而这个,大概就是这家公司神奇的地方。

You know, when you're designing a computer, you have to have an operating system of computers.

你设计一台计算机,就得有一套管理这些计算机的操作系统。

When you're designing a company, you should first think about what is it that you want the company to produce.

你设计一家公司,首先该想的是:你想让这家公司产出什么。

You know, I see a lot of companies' organization charts, and they all look the same.

我看过很多公司的组织架构图,长得都一个样。

Hamburger organization charts, soft organization charts, and car company organization charts. They all look the same. And it doesn't make any sense to me.

汉堡店的组织图、软件公司的组织图、汽车公司的组织图,全都一个样。这在我看来毫无道理。

You know, the goal of a comp- of a company is to be the machinery, the mechanism, the system that produces the output.

一家公司的目标,是成为那台机器、那套机制、那个能产出结果的系统。

And that output is the product that we like to create.

而那个产出,就是我们想创造的产品。

It is also designed, the architecture of the company should reflect the environment by which it exists.

公司的架构同样是设计出来的——它应该反映它所处的环境。

It almost directly says what you should do with the organization.

环境几乎是直接告诉了你,组织该怎么搭。

My direct staff is 60 people. You know, I don't have one-on-ones with 'em because it's impossible.

我的直属下属有 60 个人。我不跟他们做一对一,因为根本不可能。

You can't have 60 people on your staff if you're, you know, gonna get work done and-

如果你还想把活干完,你的直属团队不可能有 60 个人还搞一对一——

So you still have 60 reports. You still have across-

所以你现在还是有 60 个直接汇报的人。你还是横跨——

More, yeah.

还更多。

More. And most stars at least have a foot in engineering.

更多。而且这些人大部分至少有一只脚踩在工程里。

Almost all of them. There's experts in memory, there's experts in CPUs, there's experts in optical. All-

几乎全部都是。有内存专家、有 CPU 专家、有光学专家。全都——

That's incredible.

这太惊人了。

Yeah, GPUs and- Architecture, algorithms, design-

对,还有 GPU——架构、算法、设计——

So, you constantly have an eye on the entire stack, and you're having to have, like, intense discussions about the design of the entire stack?

所以你是一直盯着整个技术栈,而且必须就整个栈的设计做非常密集的讨论?

And no conversation is ever one person. That's why I don't do one-on-ones.

而且没有任何一场对话只有一个人。这就是我不做一对一的原因。

We present a problem and all of us attack it. You know, because we're doing extreme co-design.

我们把一个问题摆出来,所有人一起攻。因为我们做的是极限协同设计。

And literally, the company is doing extreme co-design all the time.

这家公司确实是在一直做极限协同设计。

So, even if you're talking about a particular component, like cooling, networking, everybody's listening in?

所以哪怕你们在谈某个具体部件,比如散热、网络,所有人都在旁听?

Yeah, exactly.

对,正是。

And they can contribute, "Well, this doesn't work for the power distribution. This doesn't-"

而他们可以插话:"这样配电就不行了。这样——"

Exactly.

正是。

"… This doesn't work for the memory. This doesn't work for this."

"……这样内存就不行了。这样这个也不行。"

Exactly. And whoever wants to tune out, tune out. You know what I'm saying?

正是。谁想走神就走神。你明白我的意思吧?

And the reason for that is because the people who are on the staff, they know when to pay attention.

原因是这些直属团队的人,自己知道什么时候该专注。

There's supposed… You know, it's something they could have contributed to, they didn't contribute to, "I'm going to call them out." You know?

如果有件事他本来能贡献却没贡献,我会当场点他。

And so, "Hey, come on, let's get in here."

就是那句:"喂,别躲了,进来说话。"

So, as you mentioned, NVIDIA is this company that's adapting to the environment.

你刚才说,NVIDIA 是一家会适应环境的公司。

So, which point can you say, did the environment change and began adapting sort of secretly- … in the early days from GPU for gaming, maybe the early deep learning revolution to we're now going to start thinking of it as an AI factory?

那你能说出是哪个节点,环境变了、你们开始悄悄适应——从早期的游戏 GPU,也许是深度学习革命的早期,一路走到"我们现在要把它当成一座 AI 工厂"?

What does NVIDIA do? It produces AI; let's build a factory that makes AI.

NVIDIA 做的是什么?它生产 AI;那就建一座生产 AI 的工厂。

I could reason through that systematically. We started out as an accelerator company.

我可以系统地推一遍。我们一开始是一家加速器公司。

But the problem with accelerators is that the application domain's too narrow.

但加速器的问题是,应用领域太窄。

It has the benefit of being incredibly optimized for the job. You know, any specialist has that benefit.

它的好处是为那份特定工作做到了极致优化——任何专才都有这个好处。

The problem with intense specialization is that, of course, your market reach is narrower, but that's even fine.

极度专精的问题当然是市场覆盖更窄,但这一点其实还能接受。

The problem is, the market size also dictates your R&D capacity.

真正的问题是,市场规模同时决定了你的研发能力。

And your R&D capacity ultimately dictates the influence and impact that you can possibly have in computing.

而你的研发能力最终决定了你在计算领域能有多大的影响力和冲击力。

And so, when we first started out as an accelerator, very specific accelerator, we always knew that was going to be our first step.

所以我们最早做加速器、做非常特定的加速器时,一直清楚那只是我们的第一步。

We had to find a way to become accelerated computing.

我们必须找到一条路,变成"加速计算"。

But the problem is, when you become a computing company, it's too general purpose and it takes away from your specialization.

但问题是,一旦你变成一家计算公司,就太通用了,而通用会削掉你的专精。

The tur- I connected two words that actually have fundamental tension.

我把两个本身就存在根本张力的词拼在了一起。

The better computing company we become, the worse we became as a specialist.

我们越是一家好的计算公司,作为专才就越差。

The more of a specialist, the less capacity we have to do overall computing.

越专精,我们能做通用计算的余地就越小。

And so, that… And I connected those two words together on purpose, that the company has to find that really narrow path, step by step by step, to expand our aperture of computing, but not give up on the most important specialization that we had.

我是故意把这两个词拼在一起的:公司必须一步一步找到那条极窄的路——把计算的口径扩大,同时不放弃我们手里最重要的那份专精。

Okay, so the first step that we took beyond acceleration was we invented a programmable pixel shader.

所以,我们跨出加速器的第一步,是发明了可编程像素着色器。

So, that was the first step towards programmability. It was our first journey towards moving into the world of computing.

那是走向可编程的第一步,也是我们迈进计算世界的第一段路。

The second thing that we did was we created, we put FP32 into our shaders.

第二件事,是我们在着色器里放进了 FP32。

That FP32 step, IEEE-compatible FP32, was a huge step in the direction of computing.

FP32 这一步——符合 IEEE 标准的 FP32——是朝计算方向迈出的一大步。

It was the reason why all of the people who were working on stream processors and, you know, other types of data flow processors discovered us.

正因为这个,所有在做流处理器、以及其他各类数据流处理器的人发现了我们。

And they said, "Hey, all of a sudden, you know, we might be able to use this GPU that's incredibly computationally intensive, and it's now, you know, compliant with IEEE."

他们说:"嘿,这下我们说不定能用上这块算力极强的 GPU 了,而且它现在符合 IEEE 标准。"

I can take my software that I was writing, you know, previously on CPUs, and I can see about using the GPU for that.

我可以把原来写在 CPU 上的软件拿过来,试试用 GPU 来跑。

And which led us to create, put C on top of FP32, what's called, we call Cg.

这就促使我们在 FP32 之上加了 C 语言,我们叫它 Cg。

The Cg path took us to eventually CUDA.

Cg 这条路最终把我们带到了 CUDA。

CUDA, step by step by step we… Well, the putting CUDA on GeForce, that was a strategic decision that was very, very hard to do, because it cost the company enormous amounts of our profits, and we couldn't afford it at the time.

CUDA 我们一步一步来……而把 CUDA 放进 GeForce,那是一个极其艰难的战略决定,因为它耗掉了公司大量的利润,而我们当时根本负担不起。

But we did it anyway because we wanted to be a computing company.

但我们还是做了,因为我们想成为一家计算公司。

A computing company has a computing architecture. A computing architecture has to be compatible across all of the chips that we build.

计算公司要有计算架构。而计算架构必须在我们造的所有芯片上都兼容。

Can you take me through that decision? So, putting CUDA on GeForce, could not afford to do?

能带我过一遍这个决定吗?把 CUDA 放进 GeForce,当时根本负担不起?

Can you explain that decision? Why boldly choose to do that anyway? Can you explain that decision?

你能解释一下这个决定吗?为什么还是大胆做了?

Yeah, excellent. That was… I would say that that was the first strategic decision that is as close to an existential threat.

好问题。我会说,那是第一个最接近生存威胁的战略决定。

For people who don't know, it turned out to be, spoiler alert, one of the most incredibly brilliant decisions ever made by a company.

先给不了解的人剧透一下结局:这后来成了一家公司做过的最精彩的决定之一。

So, CUDA turned out to be an incredible foundation for computation in this AI infrastructure world. So, so- … just setting the context. It turned out to be a good decision.

CUDA 最后成了这个 AI 基础设施世界里做计算的绝佳地基。先把背景铺一下——事后看,这是个好决定。

Yeah, it turned out to have been a good decision. I think the… So, here's the way it went.

对,事后看是个好决定。过程是这样的。

So, we invented this thing called CUDA, and it expanded the aperture of applications that we can accelerate with our accelerator.

我们发明了 CUDA 这个东西,它把我们的加速器能加速的应用口径扩大了。

The question is, how do we attract developers to CUDA? Because a computing platform is all about developers.

问题是:怎么把开发者吸引到 CUDA 上来?因为计算平台的核心就是开发者。

And developers don't come to a computing platform just because, you know, it could perform something interesting.

而开发者不会因为一个平台"能干点有意思的事"就跑过来。

They come to a computing platform because the install base is large.

他们来一个计算平台,是因为 install base 大。

Because a developer, like anybody else, wants to develop software that reaches a lot of people.

因为开发者跟所有人一样,想写能触达很多人的软件。

So, the install base is, in fact, the single most important part of an architecture. The architecture could attract enormous amounts of criticism.

所以装机量其实是一个架构里最重要的那一件事。架构本身可以招来铺天盖地的批评。

For example, no architecture has ever attracted more criticism than the x86… you know, as a less than elegant architecture, but yet it is the defining architecture of today.

举个例子,没有哪个架构招来的批评比 x86 更多——大家都说它不够优雅,可它偏偏是当今定义性的架构。

It gives you an example that in fact so many RISC architectures which were beautifully architected, incredibly well-designed by some of the brightest computer scientists in the world, largely failed.

这就说明:那么多由世界上最聪明的计算机科学家精心设计、架构漂亮的 RISC 架构,基本上都失败了。

And so I've given you two examples where one is, you know, one is elegant, the other one's barely aesthetic, and so yet x86 survived and the reason for-

所以我给了你两个例子:一个优雅,另一个几乎谈不上美感,结果活下来的是 x86,原因就是——

Install base is everything.

装机量就是一切。

Install base defines an architecture. Not… Everything else is secondary, okay?

装机量定义架构。其他一切都是次要的。

And so there were other architectures at the time. CUDA came out, OpenCL was here. There were… You know, there's several other competing architectures.

当时也有别的架构。CUDA 出来的时候,OpenCL 在那儿,还有另外几个竞争架构。

But the thing that… The decision that we made that was good was we said, "Hey, look, ultimately it's about install base and what is the best way we could get a new computing architecture into the world?"

但我们做对的那个决定是:"归根结底这是装机量的问题——把一个新计算架构推向世界,最好的办法是什么?"

By that timeframe, GeForce had become successful.

到那个时间点,GeForce 已经做成了。

We were already selling millions and millions of GeForce GPUs a year, and we said, "You know, we, we ought to put CUDA on GeForce and put it into every single PC whether customers use it or not, and use it as a starting point of cultivating our install base."

我们那时一年已经能卖出好几百万块 GeForce GPU,于是我们说:"我们应该把 CUDA 放进 GeForce,塞进每一台 PC,不管客户用不用,就把它当成培育装机量的起点。"

Meanwhile, we'll go and attract developers, and we went to universities and wrote books and taught classes and put CUDA everywhere.

同时我们去吸引开发者:跑大学、写书、开课,把 CUDA 铺到所有地方。

And eventually people discover… And at the time, the PC was the primary computing vehicle.

最终会有人发现它……而且当时 PC 是主要的计算载体。

There was no cloud, and we could put a supercomputer in the hands of every researcher in school, every scientist, you know, every engineering school, every… or every student in school, and eventually something amazing will happen.

那时没有云,而我们可以把一台超级计算机送到每个校内研究者、每个科学家、每所工学院、每个在校学生手里——最终一定会有惊人的事情发生。

Well, the problem was CUDA increased our cost of that GPU, which is a consumer product, so tremendously, it completely consumed all of the company's gross profit dollars.

问题是,CUDA 把那块 GPU——一个消费级产品——的成本抬得太高了,直接吞掉了公司全部的毛利。

And so at the time, the company was probably, you know, worth, I don't know, at the time, eight… Was it like $8 billion or something? Like six, $7 billion or something like that.

当时公司大概值……八十亿?还是六七十亿美元左右?

After we launched CUDA, I recognized that it was going to add so much cost, but it was something we believed in.

CUDA 发布之后,我意识到它会带来巨额成本,但这是我们相信的事。

You know, our market cap went down to like one and a half billion dollars.

我们的市值掉到了大概 15 亿美元。

And so we were down there for a while and we clawed our way back slowly, but we carried CUDA on GeForce.

我们在那个位置趴了一段时间,然后一点一点爬回来,但我们一直让 GeForce 背着 CUDA。

I always say that NVIDIA is the house that GeForce built, because it was GeForce that took CUDA out to everybody.

我总说,NVIDIA 是 GeForce 盖起来的房子——因为是 GeForce 把 CUDA 带到了所有人面前。

Researchers, scientists, they discovered CUDA on GeForce because they were all, you know… Many of 'em were gamers.

研究者、科学家是在 GeForce 上发现 CUDA 的,因为他们很多人本身就是玩家。

Many of them built their own PCs anyways. In a university lab, many of them built clusters themselves, you know, using PC components.

很多人反正也自己攒机。在大学实验室里,不少人就用 PC 配件自己搭集群。

And, and so that, you know, that's kind of how we got going.

我们大致就是这么起来的。

And then that became the platform and the foundation for the deep learning revolution.

然后它就成了深度学习革命的平台和地基。

That was also another great, great observation. Yeah.

这也是一个非常了不起的观察。

That existential moment, do you remember… Like, what were those meetings like? What were those discussions like, deciding as a company, risking everything?

那个生死时刻,你还记得吗——那些会议是什么样的?一家公司决定押上一切,那些讨论是什么样的?

Well I had to make it clear to the board what we're trying to do, and the management team knew our gross margins were gonna get crushed.

我必须跟董事会讲清楚我们想干什么,而管理层知道毛利要被压垮了。

So you could imagine a world where GeForce would carry the burden of CUDA and none of the gamers would appreciate it and none of the gamers would pay for it.

你可以想象那个局面:GeForce 背着 CUDA 的成本,而游戏玩家既不感激,也不会为它付钱。

You know, they only pay certain price and it doesn't matter what your cost is.

他们只肯付某个价,你的成本是多少跟他们没关系。

And so the… You know, we increased our cost by 50% and that consumed… And we were a 35% gross margin company, and so it was a… It was quite a difficult decision to make.

我们把成本抬高了 50%,而我们是一家 35% 毛利的公司,所以这个决定相当难做。

But you could imagine that someday this would go into workstations and it would go into supercomputers and in those segments, maybe we can capture more margin.

但你可以想象,有一天它会进工作站、会进超级计算机,而在那些细分市场里,或许我们能拿到更高的毛利。

So you could reason your way into being able to afford this, but it still took… It took a decade.

所以你可以一路推演出"这笔钱花得起",但它还是花了……整整十年。

But that, but that's more of, like, conversation with the board convincing them, but you psychologically- … as NVIDIA's continued to make bold bets that predict the future, and in part, especially now, define the future.

但那更像是跟董事会的对话、说服他们,而你自己心理上——NVIDIA 一直在做那种预判未来、而且现在某种程度上是在定义未来的大胆下注。

So I'm almost looking for wisdom about how you're able to make those decisions, to make leaps- … like that as a company.

所以我其实是想请教这里面的智慧:你是怎么做出这些决定的?一家公司怎么能这样跳跃?

Well, first of all, I'm informed by a lot of curiosity.

首先,支撑我的是大量的好奇心。

At some point, there's a reasoning system that convinces me so clearly this outcome will happen. That this will happen.

到某个时刻,会有一套推演把我说服到非常清楚:这个结果一定会发生。这件事一定会发生。

And so I believe it in my mind, and when I believe it in my mind, you know how it is.

于是我在心里相信了它。而当你心里真的相信,你知道那是什么感觉。

You manifest a future and that future is so convincing, there's no way it won't happen.

你把一个未来具象出来,那个未来清晰到不可能不发生。

There's a lot of suffering in between, but you've gotta believe what you believe.

中间会有很多痛苦,但你必须相信你所相信的。

So you, you, you envision the future- … and you essentially, from a sort of engineering perspective, manifest it?

所以你先看见那个未来——然后基本上是从工程的视角把它做出来?

Yeah. And you reason about how to get there. You reason about why it must exist.

对。然后你推演怎么走到那儿,推演它为什么必然存在。

And you know, I reason… We all reason it here. The management team would reason about it. All the people that I… We spend a lot of time reasoning about it.

我们这儿所有人都在推演。管理层会推演。我们花很多时间在推演上。

The thing that… The next part of it is probably a skill thing, which is, you know, oftentimes in leadership the leadership stays quiet or they learn about something, and then they do some manifesto, and it's a brand-new year, and somehow at the end of the year, next year, we're gonna have a brand-new plan.

接下来那部分大概算一种技巧:在领导层里常见的做法是,领导先闷着,或者他学到了点什么,然后搞个宣言,新年一到,不知怎么到了年底、到了明年,我们就要有一份全新的计划。

Big huge layoff this way, big huge organization change this way, new mission statement… brand new logos, you know, that kind of stuff.

这边大裁员,那边大改组,新的使命宣言……全新 logo,诸如此类。

We've just never, I never do things that way.

我们从来不这么干,我从来不这么干。

When I learn about something and it's starting to influence how I think, I'll make it very clear to everybody near me that, you know, this is interesting.

当我学到一件事、它开始影响我的想法,我会非常明确地告诉身边所有人:这个东西有意思。

This is going to make a difference. This is going to impact that. And I reason about things step by step by step.

这件事会带来改变。这件事会影响到那件事。然后我一步一步往下推。

Oftentimes, I've already made up my mind, but I'll take every possible opportunity—external information, new insights, new discoveries, new engineering revelations, new milestones developed—I'll take those opportunities and I'll use it to shape everybody else's belief system.

很多时候我其实已经拿定主意了,但我会抓住每一个可能的机会——外部信息、新洞见、新发现、工程上的新突破、新达成的里程碑——用这些机会去塑造其他所有人的信念系统。

And I'm doing that literally every single day. I'm doing that with my board, I'm doing that with my management team, I'm doing that with my employees.

而我几乎每一天都在做这件事:对董事会做,对管理层做,对员工做。

I'm trying to shape their belief systems such that when I come the day I say, "Hey, let's buy Mellanox," it's completely obvious to everybody that we absolutely should.

我在塑造他们的信念系统,好让我哪天说"我们把 Mellanox 买了"的时候,所有人都觉得这完全是理所当然、必须买。

On the day that I said, "Hey guys, let's go all in on deep learning," and let me tell you why. I've already been laying down the bricks to different organizations inside the company.

在我说"各位,我们全力押注深度学习"并解释为什么的那一天,我其实已经在公司里各个组织那儿铺好了砖。

Every organization and everybody, many of the people might have heard everything. Most of the company hears, of course, pieces of it.

每个组织、每个人,很多人可能已经听过全部。公司大部分人当然只听到了其中的片段。

And on the day that I announce it, everybody's kind of bought in to many pieces of it.

而到我宣布的那天,所有人对其中很多片段已经是买账的了。

And in a lot of ways, I like to announce these things, and I imagine that the employees are kind of saying, "You know, Jensen, what took you so long?"

很多时候我喜欢宣布这些事,然后想象员工在心里说:"黄仁勋,你怎么这么慢?"

And in fact, I've been shaping their belief system for some time, and therefore leadership.

而事实上我已经塑造他们的信念系统有一段时间了,这也正是领导力。

Sometimes it looks like you're leading from behind, but you've been shaping their, you know, to the point where on the day that I declared it, 100% buy-in.

有时候看起来像是你在从后面领导,但你其实一直在塑造他们——到我正式宣布那天,百分之百买账。

But that's what you want. You want to bring everybody along.

而这正是你要的。你要把所有人一起带上。

Otherwise, we announce something about deep learning and everybody goes, "What are you talking about?"

否则我们宣布一个深度学习的事,所有人都会问:"你在说什么?"

You know, you announce something about let's go all in on this thing, and your management team, your board, your employees, your customers, they're kind of like, "Where's this coming from?"

你宣布"我们要全力押注这个东西",然后你的管理层、董事会、员工、客户全都是:"这是从哪儿冒出来的?"

You know, this is insane."

"这太疯了。"

And so, so GTC effect, if you go back in time, you look at, look at the keynotes, I'm also shaping the belief system of my partners in the industry and, and I'm using that to shape, you know, the belief system of my own employees.

所以 GTC 的效果——你回头去看那些 keynote,我同时也在塑造行业里合作伙伴的信念系统,再用它来塑造我自己员工的信念系统。

And, and, and so by the time that I announce something, like for example, we just announ- we just announced Grok. We've been late… I've been talking about the stepping stones for two and a half years.

所以等我真正宣布某件事的时候——比如我们刚刚发布了 Grok。关于通往它的那些台阶,我已经讲了两年半。

You just go back and go, "Oh my gosh, they've been talking about it for two and a half years."

你回头一看就会说:"我天,他们已经讲了两年半了。"

And so I've been laying the foundation step by step by step, so when the time comes you announce it, everybody's saying, "You know, what took you so long?"

所以我是一步一步在打地基,等时候到了宣布出来,所有人都在说:"你怎么这么慢?"

But it's not just inside the company. You're shaping the landscape, the broader global landscape of innovation.

但这不只发生在公司内部。你是在塑造整个格局,塑造全球创新的大格局。

Like, putting those ideas out there, you really are manifesting reality.

把那些想法放出去,你真的是在把现实造出来。

We don't build computers. We actually don't build clouds. We don't… As it turns out, we're a computing platform company.

我们不造计算机。我们其实也不造云。说到底,我们是一家计算平台公司。

And so nobody can buy anything from us. That's the weird thing.

所以没人能从我们这儿买到任何东西——这是个很怪的事。

You know, we vertically design, vertically integrate to design and optimize, but then we open up the entire platform at every single layer to be integrated into other companies' products and services and clouds and supercomputers and OEM computers, and so the amazing thing is, I can't do what I do without having convinced them first.

我们做垂直设计、垂直整合,以此做优化,但接着我们把整个平台在每一层都开放出去,让它被整合进其他公司的产品、服务、云、超级计算机和 OEM 计算机里。所以最神奇的地方在于:我不先把他们说服,就没法做我要做的事。

And so most of GTC is about manifesting a future that by the time that we… My product is ready, they're going, "What took you so long?"

所以 GTC 大部分内容是在把一个未来具象出来——等到我的产品准备好那天,他们的反应是:"你怎么这么慢?"

Yeah. So one of the things you've been a believer for a long time is scaling laws, broadly defined. So are you still a believer in the scaling laws?

你长期以来一直相信的一件事,是广义上的 scaling law。你现在还信 scaling law 吗?

Yeah, yeah. Yeah, we have more scaling laws now.

信,信。而且我们现在有更多条 scaling law 了。

So I think you've outlined four of them with pre-training, post-training, test time, and agentic scaling.

我记得你梳理出了四条:预训练、后训练、test-time,以及 agentic scaling。

What do you think, when you think about the future, deep future and the near-term future, what are the blockers that you're most concerned about that keep you up at night that you have to overcome in order to keep scaling?

当你想未来——很远的未来和很近的未来——你最担心、最让你睡不着、必须跨过去才能继续扩展的阻碍是什么?

Well, we can go back and reflect on what people thought were blockers. So in the beginning, we were… The pre-training scaling law.

我们可以回头看看,过去大家以为的阻碍都是什么。最开始是预训练的 scaling law。

You know, people thought, rightfully so, that the amount of data that we have, high-quality data that we have, will limit the intelligence that we achieve.

当时大家认为——也有道理——我们手上的数据量、高质量数据的量,会限制我们能达到的智能水平。

And that scaling law was an important, very important scaling law. The larger the model, the correspondingly more data results in a smarter AI. And so that was pre-training.

那条 scaling law 很重要,非常重要:模型越大、相应地数据越多,AI 就越聪明。这就是预训练。

And Ilya Sutskever said, "We're out of data," or something like that. "Pre-training is over," or something like that.

然后 Ilya Sutskever 说了类似"数据用完了"的话,还有"预训练结束了"之类的。

The industry panicked, you know, that this is the end of AI. And of course, that's obviously not true.

整个行业慌了,以为这是 AI 的终点。当然,这显然不是事实。

We're gonna keep on scaling the amount of data that we have to train with.

我们会继续把用来训练的数据量扩上去。

A lot of that data is probably gonna be synthetic, and that also confused people, you know?

其中很大一部分数据大概会是合成的,而这一点也让人困惑。

And what people don't realize is they've kind of forgotten that most of the data that we are training, that we teach each other with, inform each other with, is synthetic.

大家没意识到、其实是忘了:我们用来训练、用来互相教、互相告知的大部分数据,本来就是合成的。

You know, it's synthetic because it didn't come out of nature. You created it. I'm consuming it. I modify it, augment it, I regenerate it, somebody else consumes it.

它是合成的,因为它不是从大自然里长出来的。是你创造了它,我消费它;我修改它、增补它、再生成一遍,然后别人来消费。

And so we've now reached a level where AI is able to take ground truth, augment it… Enhance it, synthetically generate an enormous amount of data.

所以我们现在到了这样一个水平:AI 能拿到 ground truth,把它增补、增强,再合成出海量的数据。

And that part of post-training continues to scale, and so the amount of data that we could use that is human generated will be smaller, and smaller, and smaller.

后训练的这部分会继续扩展,所以我们能用的人类生成数据的占比会越来越小、越来越小。

The amount of data that we use to train models is going to continue to scale to the point where we're no longer limited… Training is no longer limited by… Data is now limited by compute.

而我们用来训练模型的数据总量会继续扩大,一直扩到我们不再受限于——训练不再受限于数据,数据现在受限于算力。

And the reason for that is most of the data is synthetic.

原因就是大部分数据都是合成的。

Then the next phase is test time, and I still remember people telling me that, "Inference? Oh, yeah, that's easy. Pre-training, that's hard."

下一个阶段是 test-time。我还记得有人跟我说:"推理?哦,那个简单。难的是预训练。"

These are giant systems that people are talking about. Inference must be easy. And so inference chips are gonna be little tiny chips, and-

大家谈的是那些庞然大物的系统。推理肯定很简单。所以推理芯片会是很小很小的芯片,然后——

… you know, they're not, they're not like NVIDIA's chips. Oh, those are gonna be complicated and expensive, and, you know, we could make…

……它们不会像 NVIDIA 的芯片。哦,那些又复杂又贵。

And this is- … in, in the future, inference is gonna be the biggest market, and it's gonna be easy, and we're gonna commoditize it. You know, everybody can build their own chips.

而那个说法是:未来推理会是最大的市场,而且它很简单,我们会把它做成白菜生意——人人都能造自己的芯片。

And, and that was always illogical to me because inference is thinking, and I think thinking is hard. Thinking is way harder than reading.

这在我看来一直不合逻辑,因为推理就是思考,而我认为思考很难。思考比阅读难得多。

You know, pre-training is just memorization and generalization, you know, and looking for patterns in relationships.

预训练只是记忆和泛化,是在关系里找模式。

You're reading and reading, versus thinking, reasoning, solving problems, taking unexplored experiences, new experiences, and breaking it down into… Decomposing it into, you know, solvable pieces that we then go off, either through first principle reasoning, or, you know, through previous examples, prior experiences.

你是在不停地读、不停地读——而另一边是思考、推理、解决问题,是把没探索过的经历、新的经历拆开,分解成可解的小块,然后我们再去攻:要么靠第一性原理推理,要么靠以前的例子、过往的经验。

You know, or just exploration and search and, you know, trying different things.

或者干脆就是探索、搜索、试各种不同的做法。

And that whole process of test time scaling inference, is really about thinking. And it's about reasoning, it's about planning, it's about search, it's about…

而 test-time scaling 这整个推理过程,本质就是思考。是推理,是规划,是搜索,是……

And so how could that possibly be compute light? And we were absolutely right about that. You know, so test time scaling is intensely compute intensive.

所以这怎么可能是"算力轻"的呢?这一点我们完全押对了——test-time scaling 是极度吃算力的。

Then the question is, okay, now we're at inference and we're at test time scaling, what's beyond that?

接下来的问题是:好,我们现在到了推理、到了 test-time scaling,再往后是什么?

Well, obviously we have now created, you know, one agentic person, and that one agentic person has a large language model that we've now developed.

显然,我们现在造出了一个"agentic 的人",而这个 agentic 的人身上装着我们做出来的大语言模型。

But during test time, that agentic system goes off and does research and bangs on databases, and it goes out and, you know, uses tools, and one of the most important things it does is spins off and spawns off a whole bunch of sub-agents.

但在 test time,这个 agentic 系统会跑出去做研究、反复砸数据库,会出去调用工具——而它做的最重要的事情之一,是分裂出、派生出一大堆子 agent。

Which means we're now creating large teams. It's so much easier to scale NVIDIA by hiring more employees than it is to scale myself.

这意味着我们现在是在造大团队。靠多招员工来扩展 NVIDIA,比扩展我自己容易得多。

And so the next scaling law is the agentic scaling law. It's kind of like multiplying AI.

所以下一条 scaling law 就是 agentic scaling law。它有点像是在"乘"AI。

Multiplying AI, we could spin off agents as fast as you want to spin off agents.

乘 AI——你想多快派生 agent,就能多快派生。

And so, you know, I… You know, I have four scaling laws.

所以我有四条 scaling law。

And as we use the agentic systems, they're gonna create a lot more data, they're gonna create a lot of experiences.

而当我们开始用这些 agentic 系统,它们会产生大量新数据、大量新经验。

Some of it we're gonna say, "Wow, this is really good. We ought to memorize this."

其中有些我们会说:"哇,这个真好,我们应该把它记下来。"

That data set then comes all the way back to pre-training. We memorize and generalize it.

那份数据集接着一路回流到预训练。我们把它记住、泛化掉。

We then refine it and fine-tune it back into post-training.

然后我们把它精炼、微调,回灌进后训练。

Then we enhance it even more with test time, you know, and the agentic systems, you know, put it out to the industry.

再用 test time 和 agentic 系统进一步增强它,然后把它推给整个行业。

And so this loop, this cycle, is gonna go on and on and on.

所以这个环、这个循环会一轮一轮转下去。

It kinda comes down to basically intelligence is gonna scale by one thing, and that's compute.

归根结底就是一句话:智能只会随一件事扩展,那就是算力。

But there's a tricky thing there that you have to anticipate and predict, which is some of these components, it requires different kind of hardware to really do it optimally.

但这里有个棘手的地方需要你提前预判:其中有些环节,要做到最优就得用不同类型的硬件。

So you have to anticipate where the AI innovation's going to lead. For example, a mixture of-

所以你必须预判 AI 的创新会走向哪里。比如说,专家混合——

Perfect.

太对了。

… experts with sparsity.

……带稀疏性的专家混合。

Perfect.

太对了。

With hardware, you can't just pivot on a week's notice. You have to anticipate what that's going to look like. It has some-

硬件不可能提前一周通知就转向。你必须提前预判它会长成什么样。这就有点——

So good.

问得真好。

… that's so scary and difficult to do, right?

……这件事既可怕又难做,对吧?

For example, these AI model architectures are being invented about once every six months. Right?

比如说,AI 模型架构大概每六个月就被发明出一代新的,对吧?

And system architectures and hardware architectures kind of every three years.

而系统架构和硬件架构大概三年一代。

And so you need to anticipate what likely is going to happen, you know, two, three years from now.

所以你得预判两三年之后大概会发生什么。

And there's a couple ways that you could do that. First of all, we could do research internally ourselves, and that's one of the reasons why we have basic research, we have applied research.

有几种办法可以做到。第一,我们自己在内部做研究——这也是我们为什么要有基础研究、应用研究。

We create our own models. And so we have hands-on life experience right here. This is part of the co-design that I'm talking about.

我们自己做模型,所以我们在这儿有第一手的亲身体验。这就是我说的协同设计的一部分。

We're also the only AI company in the world that works with literally every AI company in the world.

我们还是全世界唯一一家跟世界上每一家 AI 公司都合作的 AI 公司。

And to the extent that we can, we try to get a sense of what are the challenges that people are experiencing.

在能做到的范围内,我们尽量去摸清:大家现在遇到的困难是什么。

So you're listening to the whispers across the industry, the AI labs.

所以你是在听整个行业、那些 AI 实验室的风声。

That's right. You got to listen and learn from everybody.

对。你得听所有人说话,从所有人身上学。

And have a… And then the last part is to have an architecture that's flexible, that can adapt and move with the wind.

最后一部分,是要有一个灵活的架构,能顺着风向调整和移动。

And one of the benefits of CUDA is that it's, you know, on the one hand, an incredible accelerator. On the other hand, it's really flexible.

CUDA 的好处之一是:一方面它是个极强的加速器,另一方面它非常灵活。

And so that balance, incredible balance between specialization, otherwise we can't accelerate the CPU, versus generalization, so that we can adapt with changing algorithms, that's really, really important.

所以那种平衡——专精(不然我们没法加速 CPU)和通用(好让我们能跟着算法变化调整)之间的绝妙平衡——真的非常重要。

That's the reason why CUDA has been so resilient on the one hand, and yet we continue to enhance it.

这就是 CUDA 一方面这么有韧性,同时我们又能不断增强它的原因。

We're at CUDA 13.2, and so we're evolving the architecture so fast that we can stay with the modern algorithms.

我们现在是 CUDA 13.2,架构演进得非常快,快到能跟上现代算法。

For example… When mixtures of experts came out, that's the reason why we had NVLink 72 instead of NVLink 8.

举个例子:专家混合出来的时候,这就是我们为什么做的是 NVLink 72,而不是 NVLink 8。

We could now take an entire 4 trillion, 10 trillion parameter model and put it in one computing domain as if it's running on one GPU.

我们现在可以把一个 4 万亿、10 万亿参数的完整模型放进一个计算域里,就像它跑在一块 GPU 上一样。

People probably didn't notice, I said it, but if you look at the architecture of the Grace Blackwell racks, it was completely focused on doing one thing, processing the LLM.

大家可能没注意到——我说过这事——但你去看 Grace Blackwell 机架的架构,它完全是围绕一件事做的:处理 LLM。

All of a sudden, one year later, you're looking at a Vera Rubin rack. It has storage accelerators. It has this incredible new CPU called Vera.

结果一年之后,你看到的是 Vera Rubin 机架。它有存储加速器,还有一颗叫 Vera 的、非常厉害的新 CPU。

It has Vera Rubin and NVLink 72 to run the LLMs.

它有 Vera Rubin 和 NVLink 72 来跑 LLM。

It also has this new additional rack called Rock.

它还多了一个新机架,叫 Rock。

And so this entire rack system is completely different than the previous one, and it's got all these new components in it.

所以整套机架系统跟上一代完全不同,里面全是这些新部件。

And the reason for that is because the last one was designed to run MoE large language models, inference.

原因是上一代是为跑 MoE 大语言模型的推理设计的。

And this one is to run agents and agents bang on tools, and-

而这一代是为跑 agent 设计的——而 agent 会反复砸工具,然后——

Obviously, the design of the system had to have been done before Claude Code, Codex, OpenClaw. So you were anticipating the future, essentially.

显然,这套系统的设计必须在 Claude Code、Codex、OpenClaw 之前就完成。所以你本质上是在预判未来。

And that, and that comes from what? From the whispers, from the understanding what all the state-

而这来自什么?来自那些风声,来自对最前沿——

No

不是。

… of the art is about?

……对最前沿状态的理解?

No, it's easier than that. You just reason about it. First of all, you just reason.

不是,比那简单。你只要推演一遍就行。首先,你就是推演。

No matter, no matter what happens, at some point in order for that large language model to be a digital worker… Let's just use that metaphor.

不管发生什么,到某个时候,要让那个大语言模型成为一个数字员工——我们就用这个比方。

Let's say that we want the LLM to be a digital worker. What does that have to do?

假设我们想让 LLM 当一个数字员工。它得干什么?

It has to access ground truth. That's our file system. It has to be able to do research. It doesn't know everything.

它得能访问 ground truth,那就是我们的文件系统。它得能做研究,因为它不是什么都知道。

We don't have… And I don't wanna wait until this AI becomes, you know, universally smart about everything, past, present, and future before I make it useful.

而我不想等到这个 AI 对过去、现在、未来的一切都无所不知了,才让它变得有用。

And so therefore, I might as well let it go do research. It's obvious; if it wants to help me, it's gotta use my tools.

所以那我不如让它自己去做研究。这很显然:如果它想帮我,它就得用我的工具。

You know, a lot of people would say, "You know AI is gonna completely destroy software. We don't need software anymore. We don't even need tools anymore." That's ridiculous.

很多人会说:"AI 会把软件彻底摧毁。我们不再需要软件了,连工具都不需要了。"这太荒谬了。

Let's use the… Let's use a thought experiment. And you could just sit there, enjoy a glass of whiskey, and think about all these things, and it would become completely obvious.

我们来做个思想实验。你就坐在那儿,喝一杯威士忌,把这些事想一遍,答案会变得完全显然。

Like, if I were to create the most amazing agent that we can imagine in the next 10 years. Let's say it'd be a humanoid robot.

比如,假设我要造出未来十年里我们能想象到的最惊人的 agent。就说它是个人形机器人吧。

If that humanoid robot were to be created, is it more likely that the humanoid robot comes into my house and uses the tools that I have to do the work that it needs to do?

如果这个人形机器人真被造出来了,更可能的情况是:它走进我家,用我家现有的工具去干它要干的活?

Or does this hand turn into a 10-pound hammer in one instance, turn into a scalpel in another instance, and in order to boil water, it beams, you know, microwaves out of its fingers?

还是说,它这只手一会儿变成一把十磅重的锤子,一会儿变成一把手术刀,而为了把水烧开,它从手指头里射出微波?

You know, or is it more likely just to use a microwave, you know? And the first time it goes up to the microwave, it probably doesn't know how to use it.

还是说,它更可能就是直接用微波炉?而它第一次走到微波炉前面,大概不知道怎么用。

But that's okay. It's connected to the internet. It reads the manual of this microwave, reads it, instantly becomes an expert. And so it uses it.

但没关系,它连着互联网。它把这台微波炉的说明书读一遍,读完,瞬间成为专家。于是它就会用了。

And so I think the… I just described, in fact, almost all of the properties of OpenClaw.

而我刚才描述的,其实几乎就是 OpenClaw 的全部特性。

You know, that it's gonna use tools, that it's gonna access files, it's gonna be able to do research. It has an IO subsystem.

它会用工具、会访问文件、能做研究,它有一个 IO 子系统。

And when you're done reasoning through it, reasoning about it in that way, then you say, "Oh, my gosh, the impact to the future of computing is deeply profound."

当你这样一路推演完,你就会说:"我天,这对计算的未来影响之深远,难以想象。"

And the reason for that is, I think we've just reinvented the computer.

原因是,我认为我们刚刚把计算机重新发明了一遍。

And then now you say, "Okay, when did we reason about that? When did we reason about OpenClaw?"

然后你会问:"好,那我们是什么时候推演出这件事的?什么时候推演出 OpenClaw 的?"

If you take the OpenClaw schematic that I used at GTC, you'll find it two years ago.

你把我在 GTC 上用的那张 OpenClaw 示意图翻出来,你会在两年前找到它。

Literally, two years ago at GTC, I was talking about agentic systems that exactly reflect OpenClaw today. And, of course, the confluence of many things had to happen.

真的,两年前的 GTC 上,我讲的 agentic 系统就和今天的 OpenClaw 一模一样。当然,这需要很多事情汇聚到一起才行。

First of all, we needed Claude and GPT and, you know, all of these models to reach a level of capability.

首先,我们需要 Claude、GPT 以及所有这些模型达到某个能力水平。

So their innovation and their breakthroughs and their continued advances was really important.

所以他们的创新、他们的突破、他们持续的进展都非常重要。

And then, of course, somebody had to create an open source project that was sufficiently robust and sufficiently complete and that we can all put to work.

然后当然,还得有人做出一个足够健壮、足够完整、我们所有人都能拿来干活的开源项目。

And I think OpenClaw did for agentic systems what ChatGPT did for generative systems. And I just think it's a very big deal.

我觉得 OpenClaw 对 agentic 系统做的事,就是 ChatGPT 对生成式系统做的事。我认为这是件非常大的事。

Yeah, it's a really special moment. I'm not exactly sure why it captured so much of the world's attention, but it did, more than Claude Code and Codex and so on.

对,这是个很特别的时刻。我也说不清它为什么抓住了全世界这么多注意力,但它确实做到了,比 Claude Code、Codex 这些都更甚。

Because consumers could reach it.

因为普通消费者能碰到它。

Sure, yeah. But there's also so much of this is vibes. And Peter, I had a podcast with him, he's a wonderful human being.

是,当然。但这里面很大一部分也是"氛围"。Peter 我跟他录过一期播客,他是个很好的人。

So part of it is also the humans that represent the thing.

所以其中一部分也在于:代表这个东西的是什么人。

Yeah, no doubt.

对,毫无疑问。

Part of it is memes and the— 'Cause we're all trying to figure it out.

一部分是 meme——因为我们所有人都还在摸索。

There's really serious and complicated security concerns about when you have such powerful technology, how do you hand over your data so they can do useful stuff?

当你手里有这么强的技术,安全上有非常严肃又复杂的顾虑:你怎么把自己的数据交出去,让它们去做有用的事?

But then there's scary things associated with that. And we, as a civilization, as individual people and as a civilization, are figuring out how to find that right balance.

但随之而来的又有可怕的东西。而我们作为一个文明、作为个体也作为文明,正在摸索那个正确的平衡点在哪。

Yeah, we jumped on it right away and we sent a bunch of security experts this way. And we did this thing called OpenShell. It's already been integrated into OpenClaw.

对,我们马上就扑上去了,派了一批安全专家过去。我们做了个叫 OpenShell 的东西,它已经集成进 OpenClaw 了。

And NVIDIA put forward NemoClaw.

而且 NVIDIA 还推出了 NemoClaw。

Yep, exactly.

对,正是。

They install super easy. It makes sure that it's secure.

它们安装超级简单,而且能保证安全。

We give you two out of three rights.

我们给你三项权限里的两项。

Agentic systems can access sensitive information, it can execute code, and it can communicate externally.

agentic 系统可以访问敏感信息、可以执行代码、可以对外通信。

We could keep things safe if we gave you two out of those three capabilities at any time, but not all three.

只要我们任何时刻只给你这三项中的两项、而不是全部三项,就能保证安全。

And out of those two out of three capabilities, we also give you access control based on whatever rights that you're given by enterprise.

而在那两项能力之内,我们还会按企业给你的权限做访问控制。

And then we connect it to a policy engine that all these enterprises already have.

然后我们把它接到这些企业本来就有的策略引擎上。

And so we're going to try to do our best to help OpenClaw become a better claw.

所以我们会尽全力,帮 OpenClaw 变成一只更好的爪子。

So you eloquently explained how we have a long history of blockers that we thought were going to be blockers, and we overcame them.

你刚才很清楚地讲了:我们有一长串"以为会是阻碍"的阻碍,后来都跨过去了。

But now looking into the future, what do you think might be the blockers now that it's clear that agents will be everywhere?

但往前看,既然现在已经很清楚 agent 会无处不在,你觉得接下来的阻碍可能是什么?

So it's obviously we're going to need compute. So what is going to be the blocker for that scaling?

算力显然是需要的。那这轮扩展的阻碍会是什么?

Power is a concern, but it's not the only concern.

电力是个隐忧,但不是唯一的隐忧。

But that's the reason why we're pushing so hard on extreme co-design, so that we can improve the tokens per second per watt orders of magnitude every single year.

而这正是我们这么用力推极限协同设计的原因——好让每瓦每秒 token 数每年都提升好几个数量级。

And so in the last 10 years, Moore's Law would have progressed computing about 100 times in the last 10 years. We progressed and scaled up computing by a million times in the last 10 years.

过去十年,摩尔定律大概能让计算进步 100 倍。而我们在过去十年把计算推进、扩展了 100 万倍。

And so we're gonna keep on doing that through extreme co-design.

我们会靠极限协同设计继续这么干。

So energy efficiency, perf per watt, completely affects the revenues of a company. It affects the revenues of a factory.

所以能效、每瓦性能,是完完全全影响一家公司收入的事,也影响一座工厂的收入。

And we're just going to push that to the limit so that we can keep on driving token costs down as fast as we can.

我们就是要把它推到极限,好让 token 成本以最快的速度往下走。

You know, our computer price is going up, but our token generation effectiveness is going up so much faster that token cost is coming down.

我们计算机的价格在涨,但我们 token 生成的效能涨得快得多,所以 token 成本是在往下走的。

It's just coming down an order of magnitude every year.

它就是每年往下掉一个数量级。

So power, that's an interesting one. So the way to try to get around the power blocker is to try to, with the tokens per second per watt, try to make it more and more efficient.

电力这个点很有意思。绕开电力这个阻碍的办法,是靠每瓦每秒 token 数,让它越来越高效。

Of course, there's the question of how do we get more power.

当然还有另一个问题:我们怎么拿到更多电。

We should also get more power.

我们也应该去拿更多电。

That's a really complicated one. You've talked about small modular nuclear power plants. There's all kinds of ideas for energy.

这个特别复杂。你提过小型模块化核电站,能源方面各种想法都有。

How much does it keep you up at night? The bottlenecks in the supply chain of AI, like ASML with EUV lithography machines, TSMC with advanced packaging like CoWoS, and SK Hynix with the high bandwidth memory?

AI 供应链上的那些瓶颈,让你有多睡不着?比如 ASML 的 EUV 光刻机、TSMC 的 CoWoS 这类先进封装、SK Hynix 的高带宽内存?

All the time, and we're working on it all the time.

一直睡不着,而且我们一直在处理它。

No company in history has ever grown at a scale that we're growing while accelerating that growth. It's incredible. And it's hard for people to even understand this.

历史上没有任何一家公司在我们这个体量上增长,同时还在加速这个增长。这太不可思议了,连外人都很难理解。

In the overall world of AI computing, we're increasing share. And so supply chain, upstream and downstream, are really important to us.

在整个 AI 计算的版图里,我们的份额还在上升。所以供应链——上游和下游——对我们非常重要。

I spend a lot of time informing all the CEOs that I work with: what are the dynamics that's going to cause the growth to continue or even accelerate?

我花大量时间去告知所有合作的 CEO:是什么动力会让这个增长持续、甚至加速?

It's part of the reasons why to the entire right-hand side of me were CEOs of practically the entire IT industry upstream and practically the entire infrastructure industry downstream.

这也是为什么在我右手边整排坐着的,几乎是整个上游 IT 产业和整个下游基础设施产业的 CEO。

And they were all… There were several hundred CEOs. And I don't think there's ever been keynotes where several hundred CEOs show up.

他们全都……那是好几百位 CEO。我不觉得历史上有哪场 keynote 来了好几百位 CEO。

And part of it is, I'm telling them about our business condition now. I'm telling them about the growth drivers in the very near future and what's happening.

其中一部分原因是,我在告诉他们我们现在的业务状况,告诉他们最近的未来里有哪些增长驱动、正在发生什么。

And I'm also describing where are we going to go next so that they could use all of this information and all of the dynamics that are here to inform how they want to invest.

我也描述我们下一步要往哪走,好让他们能用这些信息和这里的所有动力,来决定自己要怎么投。

And so I inform them that way like I inform my own employees.

我用这种方式告知他们,跟我告知自己员工的方式一样。

And then of course, then I make trips out to them and make sure that, "Hey, listen, I want you to know this quarter, this coming year, this next year, these things are going to happen."

然后我当然还会亲自跑去他们那儿,确保他们知道:"听着,我要你知道这个季度、今年、明年,这些事情会发生。"

And if you look at the CEOs of the DRAM industry—the number one DRAM in the world was DDR memory for CPUs in data centers.

你去看 DRAM 行业那些 CEO——当时世界第一大 DRAM 是给数据中心 CPU 用的 DDR 内存。

About three years ago, I was able to convince several of the CEOs that even though at the time HBM memory was used quite scarcely, and barely by supercomputers, that this was going to be a mainstream memory for data centers in the future.

大约三年前,我说服了其中几位 CEO:尽管当时 HBM 内存用得非常少、几乎只有超级计算机在用,但它未来会成为数据中心的主流内存。

At first it sounded ridiculous, but several of the CEOs believed me and decided to invest in building HBM memories.

一开始这听起来很荒唐,但有几位 CEO 信了我,决定投钱去建 HBM 内存的产能。

Another memory was rather odd to put into a data center: the low power memories that we use for cell phones.

还有一种内存放进数据中心也挺奇怪:我们手机上用的那种低功耗内存。

And we wanted them to adapt them for supercomputers in the data center. And they go, "Cell phone memory for supercomputers?" And I explained to them why.

我们希望他们把它改造成能给数据中心超级计算机用的。他们说:"手机内存给超级计算机用?"我给他们解释了为什么。

Well, look at these two memories, LPDDR5, HBM4. The volumes are so incredible.

现在你看这两种内存,LPDDR5、HBM4,出货量惊人。

All three of them had record years in history, and these are 45-year companies.

他们三家都创下了历史最高的年份——而这些都是四十五年的老公司。

And so, you know, I… That's part of my job, is to inform and shape, inspire, you know.

这就是我工作的一部分:告知、塑造、激发。

So you're not just manifesting the future and maybe inspiring NVIDIA, the different engineers of the company, you're manifesting the supply chain of the future.

所以你不只是把未来具象出来、激励 NVIDIA 公司内部各路工程师,你是在把未来的供应链也具象出来。

So you're having conversations with TSMC, with ASML.

所以你在跟 TSMC、跟 ASML 谈。

Upstream, downstream.

上游,下游。

Upstream, downstream. So that's the thing.

上游、下游。原来是这样。

GEV, Caterpillar. Yeah, that's downstream from us. Yeah, yeah, there you go.

GEV、Caterpillar。对,那些是我们的下游。对,就是这样。

Yeah, the whole thing. I mean, but that's so… There's so much incredibly difficult engineering that happens in the entire semiconductor industry, and it just feels scary how intricate the supply chain is, how many components there are, but it works somehow.

对,整条链。整个半导体产业里有太多极其困难的工程。这条供应链有多精密、有多少零部件,想想都让人发怵——但它居然就运转起来了。

Exactly, the deep science. The deep engineering, the incredible manufacturing, and so much of the manufacturing is already robotics, but we have a couple of hundred suppliers that contribute the technology that goes into our 1.3 million component rack.

正是——深科学、深工程、惊人的制造能力,而其中很大一部分制造已经是机器人在做了。但我们有两百多家供应商,把技术贡献进我们那个 130 万零件的机架里。

Each rack is 1.3, one and a half million components. There are 200 suppliers across the Vera Rubin rack.

每个机架是 130 万到 150 万个零件。整个 Vera Rubin 机架背后有 200 家供应商。

So it's interesting that you don't list that as the thing that keeps you up at night in the list of blockers.

所以有意思的是,你并没有把它列进"让你睡不着的阻碍"清单里。

But I'm doing, I'm doing all the things necessary to-

但我在做,我在做所有必要的事情来——

Okay.

好。

… yeah, see? I can go to sleep because I checked it off.

……看到了吧?我能睡着,因为我把它勾掉了。

I said, "Okay," you know, I go, I can go to sleep and I go, "Well, let's see, let's reason about this. What's important for us?" Because let's reason about this.

我会说:"好,我们来推一下。对我们来说什么最重要?"

Because we changed the system architecture from the original DGX-1 that you remembered to NVLink-72 rack scale computing- … what's gonna… What does that, what does that mean?

因为我们把系统架构从你记得的那个最早的 DGX-1,换成了 NVLink-72 的机架级计算——那意味着什么?

What does that mean to software? What does that mean to engineering? What does that mean to how we design and test? And what does that mean to the supply chain?

对软件意味着什么?对工程意味着什么?对我们怎么设计和测试意味着什么?对供应链又意味着什么?

Well, one of the things that it meant was we moved supercomputer integration at the data center into supercomputer manufacturing in the supply chain.

其中一个结论是:我们把原来在数据中心做的超级计算机集成,搬进了供应链里的超级计算机制造环节。

If you're doing that, you also have to recognize you're gonna move one…

你这么做的话,还得意识到你会挪动一件事——

And if your total footprint of whatever data center you're gonna build, let's say you would like to have, you know, 50 gigawatts of supercomputers that are running simultaneously, and it takes one week to manufacture that 50 gigawatts of supercomputers, then each week in the supply chain, the supercomputers are gonna need a gigawatt of power.

假设你要建的数据中心总规模是,比如说 50 吉瓦的超级计算机同时在跑,而制造这 50 吉瓦的超级计算机需要一周时间,那么供应链里每一周,这些超级计算机就要吃掉一吉瓦的电。

And so we're gonna need the supply chain to increase the amount of power it has to build and test the supercomputers in the supply chain before I ship it.

所以我们需要供应链把它的用电量提上来,好在我出货之前,在供应链里把这些超级计算机造出来并测完。

Oh.

哦。

Well, NVLink-72 literally builds supercomputers in the supply chain and ships 'em two, three tons at a time per rack.

NVLink-72 确实是在供应链里就把超级计算机造好,然后一个机架两三吨地发货。

It used to be they used to come in parts and we used to assemble 'em inside the data center. But that's impossible now because NVLink-72 is so dense.

以前是零件运过来,我们在数据中心里组装。但现在不可能了,因为 NVLink-72 太密了。

And so that's an example. And I would have to go and fly into the supply chain, go meet my partners saying, "Hey," I said, "guess what? So here's what I'm going to do with… This is the way we used to build our DGXs. We're gonna build them this way. This is gonna be so much better because we're going to need 'em for inference."

这就是一个例子。我就得飞到供应链里去,见我的合作伙伴,说:"你猜怎么着?我打算这么干——我们以前是这样造 DGX 的,现在要改成这样造。这样会好得多,因为我们需要它们来做推理。"

The market for inference is, you know, coming. The inflection point for inference is coming. It's gonna be a big market.

推理的市场要来了。推理的拐点要来了。这会是个大市场。

And so I first explain to them what's going on, why it's gonna happen, and then I ask 'em to make several billion dollars of capital investments each.

所以我先跟他们解释发生了什么、为什么会发生,然后请他们每家投出几十亿美元的资本开支。

And because they trust me and I'm very respectful of 'em, and I give 'em every opportunity to question me and I spend time to explain things to people and I reason about it.

因为他们信我,而我也非常尊重他们。我给他们每一个质疑我的机会,我花时间给人解释,把逻辑推一遍。

I draw pictures and I reason about it in first principles. And by the time I'm done with them, they know what to do.

我画图,按第一性原理一步步推。等我跟他们讲完,他们就知道该怎么做了。

So a lot of it is about relationships and building a shared view of the future. But do you worry about certain bottlenecks? I mean, what are the biggest bottlenecks in the supply chain?

所以这里面很大一部分是关系,以及建立一个共同的未来图景。但你会担心某些瓶颈吗?供应链里最大的瓶颈是什么?

Are you worried about ASML's EUV tooling? Are you worried about the packaging, CoWoS packaging of TSMC, about how fast it could scale?

你担心 ASML 的 EUV 设备吗?担心 TSMC 的 CoWoS 封装、担心它能扩多快吗?

Like you said, you're not only growing incredibly fast, you're accelerating your growth. So it feels like everybody in the supply chain, and those are certainly bottlenecks, would have to scale up.

像你说的,你不只是长得极快,你还在加速。所以感觉供应链里每一个环节——那些确实是瓶颈——都得跟着扩。

Are you having conversations with them, like, how can you scale up faster?

你会跟他们谈"你们怎么能扩得更快"吗?

All the time.

一直在谈。

Do you worry about it?

你担心吗?

No.

不担心。

Okay.

好。

Because I told 'em what I needed. They understood what I need. They told me what they're gonna go do, and I believe them what they're going to do.

因为我告诉了他们我需要什么。他们听懂了我需要什么。他们告诉我他们会去做什么,而我信他们会去做。

Interesting. That's great to hear. So maybe if we can just linger on the power for a little bit. What are your hopes for how to solve the energy problem?

有意思,听你这么说很好。那我们在电力上多停一会儿。关于解决能源问题,你有什么期待?

One of the areas, Lex, that I would love us to talk about and just get the message out, you know, our power grid is designed for the worst case condition with some margin.

Lex,有一件事我特别希望我们聊聊、把话传出去:我们的电网是按最坏情况再加一点余量设计的。

Well, 99% of the time we're nowhere near the worst case condition because the worst case condition is a few days in the winter, a few days in the summer, and extreme weather.

而 99% 的时间里,我们离最坏情况差得很远——因为最坏情况就是冬天那几天、夏天那几天,加上极端天气。

Most of the time we're nowhere near the worst case condition and we're probably running around, call it 60% of peak.

大部分时候我们远远够不上最坏情况,大概只跑在峰值的 60% 左右。

And so 99% of the time, our power grid has excess power, and they're just sitting idle, but they have to be there sitting idle because just in case, when the time comes, hospitals have to be powered and, you know, infrastructure has to be powered and airports have to run and so on and so forth.

所以 99% 的时间里,我们的电网都有余电,就那么空着——但它必须空在那儿,以防万一:真到那时候,医院得有电、基础设施得有电、机场得能运转,等等。

And so the question that I have is whether we could go and help them understand and create contractual agreements and design computer architecture systems, data centers, such that when they need the maximum power for infrastructure in society, that the data centers would get less.

所以我的问题是:我们能不能去帮他们理解这件事、签出相应的合同,并把计算机架构系统、数据中心设计成——当社会基础设施需要最大功率的时候,数据中心就少拿一些。

But that's in a very rare instance anyways.

反正那也是极少数的情况。

And during that time, we either have a backup generator for that little part of it, or we just have our computers shift the workload somewhere else, or we have the computers just run slower.

而在那段时间里,我们要么给那一小部分配备用发电机,要么让计算机把负载挪到别处去,要么就让计算机跑慢一点。

You know, we could degrade our performance, reduce our power consumption and provide for a, you know, slightly longer latency response, you know, when somebody asks for, you know, asks for an answer.

我们可以让性能降级、把功耗降下来,在有人来要答案的时候,响应延迟稍微长一点。

And so I think that that, that way of using computers, of building data centers, instead of expecting 100% uptime—and these contracts that are really, really quite rigorous, it's putting a lot of pressure on the grid to be able to… Now, they're gonna have to increase from their maximum.

所以我认为,这种使用计算机、建数据中心的方式——而不是一味期待 100% 可用率。那些合同真的非常苛刻,给电网施加了巨大压力,逼它去……他们就得在自己的最大值之上再往上加。

I just wanna use their excess. It's just sitting there.

而我只想用他们的余量。那些电就在那儿闲着。

Yeah, that's not talked about enough. So what's stopping there? Is it regulation? Is it bureaucracy?

对,这件事讨论得太少了。那卡在哪儿?是监管吗?是官僚流程吗?

I think it's a three-way problem. It starts with the end customer.

我认为这是个三方问题。它从终端客户开始。

The end customer puts requirements on the data centers that they can never not be available, okay? So that the end customer expects perfection.

终端客户对数据中心提的要求是:绝对不能不可用。所以终端客户期待的是完美。

Now, in order to deliver that perfection, you need a combination of backup generators and your grid power supplier to deliver on perfection. And so everybody's gotta have six nines.

而要交付这份完美,你需要备用发电机加上你的电网供电方一起来兑现完美。于是所有人都必须做到六个 9。

Well, I think first of all, right now, we ought to have everybody understand that when the customer asks for these things, you have somebody in your data center operations team disconnected from the CEO.

我觉得首先,现在我们应该让所有人明白:当客户提这些要求的时候,你数据中心运维团队里签字的那个人,跟 CEO 是脱节的。

I bet the CEO doesn't know this. I'm gonna talk to all the CEOs.

我打赌 CEO 并不知道这件事。我会去跟所有 CEO 谈。

The CEOs are probably not paying any attention to the contracts that are being signed, and so everybody wants to sign the best contract, of course.

CEO 们大概根本没在关注那些正在签的合同,而所有人当然都想签到最好的合同。

And they go down to cloud service providers, and the two contract negotiators that are… You, I could just see them now. You know, negotiating these multi-year contracts. Both sides want, you know, the best contract.

这些要求一路压到云服务商那儿,然后那两个谈合同的人——我现在都能想象出那个画面。他们在谈这些多年期合同,双方都想要最好的条款。

As a result, the CSPs then have to go down to the utilities, and they expect the nine, the six nines.

结果云服务商就得再往下压到电力公司,要求做到那个六个 9。

And so I think, I think the first thing is just make sure that, that all of the customers, the CEOs and the customers realize what they're asking for.

所以我觉得第一件事,就是确保所有客户、所有 CEO 和客户,都清楚自己在要求什么。

Now, the second thing is we have to build data centers that gracefully degrade.

第二件事,是我们必须把数据中心建成能优雅降级的。

And so if the power, if the utility, if the grid tells us, "Listen, we're gonna have to back you down to about 80%," we're gonna say, "That's no problem at all."

所以如果电力公司、电网跟我们说:"听着,我们得把你压到 80% 左右",我们会说:"完全没问题。"

We're just gonna move our workload around. We're gonna make sure that data's never lost, but we can reduce the computing rate and use less energy. The quality of service degrades a little bit.

我们就把负载挪一挪。我们会确保数据永不丢失,但可以把计算速率降下来、少用点电。服务质量会稍微降一点。

For the critical workloads, I shift that somewhere else right away so I don't have that problem, and so, you know, whichever data center still has 100% uptime, and so…

至于关键负载,我立刻把它挪到别处去,这样就没这个问题了——挪到那些还保持 100% 可用率的数据中心去。

How difficult of an engineering problem is that, that smart, dynamic allocation of power in a data center?

在数据中心里做这种聪明的、动态的电力分配,是个多难的工程问题?

As soon as you could specify, you could engineer it. Beautifully put. So long as it obeys the laws of physics on first principles, I think we're good.

只要你能把它写成规格,你就能把它做出来。说得真好。只要它在第一性原理上不违反物理定律,我觉得就没问题。

What was the third thing you were mentioning?

你刚才说的第三件事是什么?

So the second thing is the, the data centers. And the third thing is we need the utilities to also recognize that this is an opportunity-

第二件事是数据中心。第三件事,是我们需要电力公司也认识到这是个机会——

… and instead of saying, "Look, it's gonna take me five years to increase my grid capability," if you have, if you're willing to take power of this level of guarantee, I can make them available for you next month and at this price.

别再说"听着,我要扩电网容量得花五年",而是说:如果你愿意接受这个保障等级的电力,我下个月就能按这个价给你。

And so if utilities also offered more segments of power delivery promises, then I think everybody will figure out what to do with it.

所以如果电力公司也能提供更多档位的供电承诺,我觉得大家都能想出该怎么用它。

Yeah, but there's just way too much waste in the grid right now. We should go after it.

现在电网里的浪费实在太多了。我们应该去把它拿回来。

You've highly lauded Elon and xAI's accomplishment in Memphis, in building Colossus supercomputer, probably in record time in just four months. It's now at 200,000 GPUs and growing very quickly.

你高度赞扬过 Elon 和 xAI 在孟菲斯的成就——他们建成 Colossus 超级计算机,大概只用了四个月,创了纪录。它现在有 20 万块 GPU,而且增长很快。

Is there something that you could speak to the understanding about his approach that's instructive to, broadly to all the data center creators that enabled that kind of accomplishment?

关于他的方法,有什么你能讲讲、对所有建数据中心的人都有启发的东西吗?是什么让这种成就成为可能?

His approach to engineering, his approach to the whole management of construction, everything?

他做工程的方式、他管理整个施工的方式,所有这些?

First of all, Elon is deep in so many different topics. Yet he's also a really good systems thinker.

首先,Elon 在很多不同领域都钻得很深。同时他还是一个非常好的系统思考者。

And so he's able to think through multiple disciplines, and he obviously pushes things, questions everything, where they're, number one, is it necessary? Number two, does it have to be done this way? And then number three, you know, does it have to take this long?

所以他能跨多个学科把问题想通。而他显然会往前推、会质疑一切:第一,这是必要的吗?第二,非得这么做吗?第三,非得花这么长时间吗?

And so he has the ability to question everything to the point where everything is down to its minimal amount that's necessary, you can't take anything else out.

所以他有能力把一切都质疑到只剩下最必要的那点,再也拿不掉任何东西为止。

And yet the necessary capabilities of the product remains, you know? And so he is as minimalist as you could possibly imagine, and he does it at a system scale.

但产品必要的能力还都在。所以他是你能想象到的最极简的那种人,而且他是在系统层面上做这件事。

I think… I also love the fact that he is represented. He is present at the point of action.

我也很喜欢他"在场"这一点——他会出现在事情发生的那个点上。

You know, he'll just go there. If there's a problem, he'll just go there and then, "Show me the problem."

他就是直接去现场。有问题,他就过去,然后说:"把问题给我看看。"

You know, when you do all of this in combination, you overcome a lot of previous, "This is just the way we do it." "You know, I'm waiting for them." You know, I mean, it's just, everybody has a lot of excuses.

当你把这些全都组合起来,你就能跨过很多以前的说法:"我们一直是这么干的"、"我在等他们"。每个人都有一堆借口。

And so, and then the last thing is when you act personally with so much urgency, it causes everybody else to act with urgency, you know?

最后一点是:当你本人带着这么强的紧迫感行动,就会让所有其他人也带着紧迫感行动。

And every supplier has a lot of customers going on. Every supplier has a lot of projects going on, and he makes it his business that he's the top priority of everybody else's projects. And so he does that by demonstrating it.

每个供应商都有很多客户在排队,每个供应商都有很多项目在跑,而他会想办法让自己成为所有人项目里的第一优先。他是靠亲自示范做到这一点的。

Yeah, I've been in a bunch of those meetings. It's just, it's fun to watch, 'cause really, not enough people ask the question like, "Okay, so can this be done a lot faster, and how? Why does it have to take this long?"

我参加过好几次那种会。看着挺有意思的,因为真的很少有人会问:"好,这件事能不能快得多?怎么快?为什么非得花这么长时间?"

Yeah, right.

对,是的。

And then in the… That becomes an engineering question often. And yes, I think when you get the ground truth of actually… I remember… One of the times I was hanging out with him, he literally is going through the entire process of how to plug in cables into a rack.

而这往往就变成了一个工程问题。是的,我觉得当你拿到真正的 ground truth……我记得有一次跟他待在一起,他真的在把"怎么把线插进机架"这整个流程走一遍。

He's working with an engineer on the ground that's doing that task, and he's just trying to understand what does that process look like so it can be less error-prone.

他跟现场那个干这活的工程师一起工作,就是想搞清楚这个流程长什么样,好让它更不容易出错。

And just building up that intuition from every single task involved in putting together a data center-

然后从组装数据中心涉及的每一个任务里,一点点建起那种直觉——

… you start to immediately get a sense at the detailed scale and at the broad systems scale of where the inefficiencies are, and so you can make it more and more and more efficient.

……你会立刻在细节层面和宏观系统层面同时感觉到低效在哪儿,于是你能让它越来越高效。

Plus you have the big hammer of being able to say, "Let's do it totally different-"

再加上你手里还有一把大锤,可以直接说:"我们换一种完全不同的做法——"

Yeah. That's right.

对,没错。

"… and remove all possible blockers."

"……把所有可能的阻碍都清掉。"

That's right.

没错。

Is there parallels in the NVIDIA Extreme Systems co-design approach that you see in the way Elon approaches systems engineering?

在 NVIDIA 的极限系统协同设计方法里,你有看到跟 Elon 做系统工程的方式相似的地方吗?

Well, first of all, co-design is an ultimate systems engineering problem. And so we approach the work that we do from that first principle.

首先,协同设计就是终极的系统工程问题。所以我们做事情就是从这个第一性原理出发。

The other thing that we do and this is a philosophy that, a thought, a state of mind, I guess, a method that I started 30 years ago, and it's called the speed of light.

另外我们还做一件事——这是一种哲学、一种想法、一种心态,或者说一种方法,我三十年前就开始用了,它叫"光速"。

The speed of light is not just about the speed. The speed of light is my shorthand for what's the limit of what physics can do.

"光速"不只是关于速度。"光速"是我对"物理能做到的极限是什么"的简写。

And so every single thing that we do is compared against the speed of light. Memory speed, math speed, power, cost, time, effort, number of people, manufacturing cycle time.

所以我们做的每一件事,都要拿去跟光速比一比:内存速度、运算速度、功耗、成本、时间、投入、人数、制造周期。

And when you think about latency versus throughput when you think about cost versus throughput, cost versus capacity, all of these things you test against the speed of light to achieve all of these different constraints separately.

当你考虑延迟对吞吐、成本对吞吐、成本对容量,所有这些你都要分别拿去跟光速比,看每一个约束单独能做到什么程度。

And then when you consider it together, you know you have to make compromises because a system that achieves extremely low latency versus a cheap, a system that achieves very high throughput are architected fundamentally differently.

然后当你把它们合起来考虑,你就知道必须做妥协——因为一个追求极低延迟的系统,和一个追求极高吞吐的便宜系统,架构从根上就不一样。

But you want to know what's the speed of light of a system that achieves high throughput, what's the speed of light of a system that achieves low latency?

但你想知道:高吞吐系统的光速是多少?低延迟系统的光速是多少?

And then when you think about the total system, you can make trade-offs. And so I force everybody to think about what's the first principles, the limits-

然后当你考虑整个系统时,你就能做权衡了。所以我强迫所有人先想清楚第一性原理、那些极限——

… the physical limits for everything before we do anything. And we test everything against that. And so that's a good frame of mind.

……在我们动手做任何事之前,先想清楚每一样东西的物理极限,然后拿一切去跟它比。这是一个很好的思维框架。

I don't love the other methods, which is continuous improvement.

我不喜欢另一种方法,就是持续改进。

The problem with continuous improvement, it… First of all, you should engineer something from first principles at the speed, you know, with speed of light thinking. Limit it only by physical limits, and physics limits.

持续改进的问题是……首先,你应该用光速思维、从第一性原理去做工程设计,只被物理极限约束。

And after that, of course you would improve it over time.

在那之后,你当然会随时间不断改进它。

But I don't like going into a problem and somebody says, "Hey, you know, it takes 74 days to do this today-" "… Right now. And we can do it for you in 72 days." You know, I'd rather strip it all back to zero-

但我不喜欢进到一个问题里,有人说:"这件事现在要花 74 天,我们能给你做到 72 天。"我宁愿把它全部剥回到零——

… and say, "First of all, explain to me why 74 days in the first place. And l- let's note, let's think about what's possible today. And if I were to- to build it completely from scratch, you know, how long would it take?"

——然后说:"首先,你先跟我解释,为什么一开始是 74 天?我们来想想今天到底什么是可能的。如果我完全从零开始造它,要多久?"

Oftentimes, you'd be surprised. It might come to six days.

很多时候你会大吃一惊,答案可能是 6 天。

Now, the rest of the six days, the 74, could be very well-reasoned and compromises, and, you know, cost reductions, and all kinds of different things. But at least you know what they are.

那么剩下那些天——74 天里多出来的部分——可能都是很有道理的妥协、降成本,以及各种各样的取舍。但至少你知道它们是什么了。

And then now that you know that six days is possible, then the conversation from 74 to six, surprisingly much more effective.

而一旦你知道 6 天是可能的,从 74 谈到 6 这场对话,会出乎意料地有效得多。

In such incredibly complex systems that you're working with, is simplicity sometimes a good heuristic to reach for?

在你们做的这种极其复杂的系统里,"简单"有时候是个好的启发式吗?

I mean, if I can just… I mean, the pod, the Vera Rubin pod that you announced is just incredible.

你们发布的那个 pod、Vera Rubin pod,简直不可思议。

We're talking about seven chips, seven chip types, five purpose-built rack types, 40 racks, 1.2 quadrillion transistors, nearly 20,000 NVIDIA dies, over 1,100 Rubin GPUs, 60 exaflops, 10 petabytes per second of scale bandwidth. That's all just one…

我们说的是七种芯片、七种芯片类型、五种专门定制的机架类型、40 个机架、1.2 千万亿个晶体管、将近 2 万颗 NVIDIA 裸片、1100 多块 Rubin GPU、60 exaflops、每秒 10 PB 的扩展带宽。而这全都只是一个——

That's just one pod.

这只是一个 pod。

That's just one pod .

这只是一个 pod。

Yeah, that's just one pod.

对,这只是一个 pod。

I mean, in- … so you have the… And then even the NVL72 rack alone is 1.3 million components, 1300 chips, 4,000 pods crammed into a single 19-inch wide rack.

而且光是 NVL72 这一个机架,就有 130 万个零件、1300 颗芯片、4000 个 pod,全塞进一个 19 英寸宽的机架里。

And Lex, we're probably gonna have to crank out about 200 of these pods a week, just to put it in perspective.

Lex,给你一个参照:我们大概每周要造出 200 个这样的 pod。

The amount of different components, I suppose simplicity is impossible, but is that a metric that you kind of reach for in trying to design things?

零部件种类这么多,我猜"简单"是不可能的了——但在设计东西的时候,这算是你会去追求的一个指标吗?

You know, the phrase that I use most often is, we need things to be as complex as necessary, but as simple as possible.

我最常用的一句话是:东西要该复杂就复杂,能简单就简单。

And so the question is, is all that complexity there necessary? And we ought to test for that. And we got to challenge that.

所以问题是:那些复杂性都是必要的吗?我们应该去检验它,应该去挑战它。

And then after that, everything else above it, you know, is gratuitous.

在那之后,任何超出的部分都是多余的。

But it's still almost incredible. Semiconductor industry broadly, but what NVIDIA is doing is some of the greatest engineering in history. So these systems are just truly, truly marvels of engineering.

但这仍然近乎不可思议。整个半导体产业都是,而 NVIDIA 在做的事是历史上最伟大的工程之一。这些系统真的是工程上的奇迹。

It is the most complex computer the world has ever made.

它是世界上造出过的最复杂的计算机。

Yeah, the engineering teams, I mean- … I don't, it's not a competition, but I don't know. If it was like an Olympics of engineering teams, I mean, TSMC does incredible engineering.

对,这些工程团队——我不是说要比,但我不知道,如果真有个工程团队的奥运会……TSMC 的工程非常了不起。

Like I said, ASML at every scale, but NVIDIA is gonna give them a run for their money. Just incredible, incredible teams.

我说过 ASML 在每个尺度上也是,但 NVIDIA 会跟他们掰手腕、不落下风。真的是了不起、了不起的团队。

Well, it's gold medalists in every single, in every single sport, all assembled right here.

这里聚齐了每一个项目、每一个单项的金牌选手。

And have to work together. And report directly to you. This is wonderful. You recently traveled to China.

而且还得一起工作,还得直接向你汇报。这太棒了。你最近去了中国。

So it's interesting to ask you, China's been incredibly successful in building up its technology sector.

所以有个问题很有意思:中国在建立自己的科技产业上极其成功。

What do you understand about how China's able to, over the past 10 years, build so many incredible world-class companies, world-class engineering teams, and just this technology ecosystem- … that produces so many incredible products?

关于过去十年中国怎么能建起这么多世界级公司、世界级工程团队,以及这样一个能产出这么多惊人产品的科技生态,你的理解是什么?

A whole bunch of reasons for… Well, first of all, let's start, let's start with some facts.

原因有一大堆……先从一些事实说起。

50% of the world's AI researchers are Chinese, plus or minus, and they're mostly in China still. We have many of them here, but there's amazing researchers still in China.

全世界 AI 研究者里大约有 50% 是华人,上下浮动,而且他们大多数还在中国。我们这儿有很多,但中国国内仍然有非常厉害的研究者。

They—their tech industry showed up at precisely the right time.

他们的科技产业出现的时间点恰好正确。

At the time of the mobile cloud era, their way of contributing was software, and so this is a country's incredible science and math really well-educated kids.

在移动云时代,他们贡献的方式是软件,而这个国家的孩子在科学和数学上受过极好的教育。

Their tech industry was created during the era of software. They're very comfortable with modern software.

他们的科技产业是在软件时代建立起来的,所以他们对现代软件非常熟悉。

China is not one giant economic country. It's got many provinces and cities with mayors all competing with each other.

中国不是一个铁板一块的巨型经济体。它有很多省和城市,市长们互相竞争。

That's the reason why there's so many EV companies. That's the reason why there's so many AI companies.

这就是为什么会有那么多电动车公司,为什么会有那么多 AI 公司。

That's the reason why there's so many—every company you could imagine, they all create some of them.

这就是为什么——你能想到的每一类公司,他们都会冒出好几家。

And, and as a result, they have insane competition internally. And, you know, what remains is an incredible company.

结果就是,他们内部的竞争激烈到疯狂。而最后活下来的,是一家极强的公司。

They also have a social culture where, where it's family first, friends second, and company third.

他们还有一种社会文化:家庭第一,朋友第二,公司第三。

And so the amount of conversation that goes back and forth between… They're essentially open source all the time.

所以彼此之间来回交流的量……他们本质上是一直在开源。

So the fact that they contribute more to open source is so sensible because they're probably, "What are we protecting?"

所以他们对开源贡献更多,这件事非常合乎逻辑,因为他们大概会想:"我们在保护什么?"

You know, my engineers, their brothers are in that company, their friends are in that company, and they're all schoolmates.

我的工程师,他们的兄弟在那家公司,他们的朋友在那家公司,而且他们都是同学。

You know, the schoolmate concept. There's a, you know, one schoolmate, you're brother for life.

同学这个概念——做过一次同学,就是一辈子的兄弟。

And and so they, they, they share knowledge very, very quickly. And so there's no sense keeping technology hidden. You might as well put it on open source.

所以他们分享知识的速度非常非常快。所以把技术藏起来没有意义,不如干脆放到开源上。

And so the open source community then amplifies, accelerates the, the innovation process.

然后开源社区又会放大、加速这个创新过程。

So you get this rapid, incredible great talent, rapid innovation because of open source and just, you know, the nature of friends, and, and insane competition.

所以你得到的是:极好的人才、因为开源而飞快的创新、朋友之间那种天然的关系,以及疯狂的竞争。

Among the company, what emerges is incredible stuff. And so this is the fastest innovating country in the world today.

在这些公司里长出来的东西非常惊人。所以这是当今世界创新最快的国家。

And this is something that has everything that, everything that I've just said is fundamental to just how the kids were grown, the fact that they have excellent education, the fact that they, parents want them to do well in school, the fact that they, their culture is that way.

而我刚才说的这一切,根子都在孩子是怎么长起来的:他们受过极好的教育,父母希望他们在学校里表现好,他们的文化就是这样。

These are, you know, these are just the thing about their country, and they showed up at precisely the time when technology is going through that exponential.

这些就是这个国家本身的特质——而他们出现的时间点,恰好是技术走上那条指数曲线的时候。

Plus culturally, it's pretty cool to be an engineer. It connects to all the components that you're mentioning…

而且在文化上,当工程师是件挺酷的事。这跟你提到的所有要素都连在一起……

It's a builder nation.

这是一个建造者之国。

It's a builder nation.

一个建造者之国。

Yeah, it's a builder nation. Our country's leaders, incredible, but they're mostly lawyers.

对,一个建造者之国。我们国家的领导人很了不起,但他们大多是律师。

Their country's leaders—and because we're, they're trying to keep us safe, rule of law governing—their country was built out of poverty.

他们是在努力保护我们的安全、以法治来治理。而他们那个国家是从贫穷里建起来的。

And so most of their leaders are incredible engineers. Some of the brightest minds.

所以他们的领导人大多是非常出色的工程师,是一些最聪明的头脑。

To take a small tangent, because you mentioned open source, I have to go to Perplexity here, who you have been a fan of a long time.

稍微岔一下,因为你提到开源,我得说说 Perplexity——你很久以来就是它的拥护者。

Love it, yeah.

很喜欢,是的。

And thank you for releasing open source Nemotron 3 Super, which you can also use inside Perplexity to look stuff up. Now, which is 120 billion parameter open weight MoE model.

也谢谢你们开源了 Nemotron 3 Super——你在 Perplexity 里面也能用它来查东西。它是一个 1200 亿参数的开放权重 MoE 模型。

What's your vision with open source? So you mentioned China with, with DeepSeek and MiniMax, with all these companies really pushing forward the open source AI movement, and NVIDIA is really leading the way in close to state-of-the-art open source LLMs. What's your vision there?

你在开源上的构想是什么?你刚提到中国,DeepSeek、MiniMax,这些公司都在真正推动开源 AI 运动,而 NVIDIA 在接近最前沿的开源 LLM 上确实走在前面。你的构想是什么?

First, if we're gonna be a great AI computing company, we have to understand how AI models are evolving.

第一,如果我们要成为一家优秀的 AI 计算公司,我们就必须理解 AI 模型是怎么演进的。

One of the things that I love about Nemotron 3 is it's not just a pure transformer model, it's transformer and SSMs.

我很喜欢 Nemotron 3 的一点是:它不只是一个纯 transformer 模型,而是 transformer 加 SSM。

And we were early in developing the, the conditional GANs, which, that progressive GANs, which led step-by-step to diffusion.

我们很早就在做条件 GAN、渐进式 GAN,那条线一步步通向了扩散模型。

And so the fact that we're doing basic research in model architecture and in different domains gives us visibility into, you know, what kind of computing systems would do a good job for future models. And so it is part of our extreme co-design strategy.

所以我们在模型架构和不同领域做基础研究这件事,让我们能看清:什么样的计算系统能把未来的模型跑好。所以它是我们极限协同设计战略的一部分。

Second, I think we rightfully recognize that on the one hand, we want world-class models as products, and they should be proprietary.

第二,我认为我们很正确地认识到:一方面,我们想要世界级的模型作为产品,它们应该是专有的。

On the other hand, we also want AI to diffuse into every industry and every country, every researcher, every student.

另一方面,我们也希望 AI 扩散到每个行业、每个国家、每个研究者、每个学生。

And if everything is proprietary, it's hard to do research and it's hard to innovate on top of, around, with.

如果一切都是专有的,做研究就很难,在它之上、围绕它、用它来创新也很难。

And so… Open source is fundamentally necessary for many industries to join the AI revolution.

所以开源对很多行业加入 AI 革命来说是根本必要的。

NVIDIA has the scale and we have the motives—not only skills, scale, and motivation—to build and continue to build these AI models for as long as we shall live. And so therefore, we ought to do that.

NVIDIA 有这个体量,也有这个动机——不只是技能,还有规模和动力——去建、并且在我们活着的每一天继续建这些 AI 模型。所以我们理应这么做。

We can open up, we can activate every industry, every researcher, you know, every country to be able to join the AI revolution.

我们可以把它开放出来,激活每个行业、每个研究者、每个国家,让他们都能加入 AI 革命。

There's the third reason, which is from that, to recognizing that AI is not just language. These AIs will likely use tools and models and sub-agents that were trained on other modalities of information.

还有第三个理由:由此我们认识到,AI 不只是语言。这些 AI 很可能会用到在其他信息模态上训练出来的工具、模型和子 agent。

Maybe it's biology or chemistry or you know, laws of physics, or you know, fluids and thermodynamics, and not all of it is in language structure.

可能是生物、化学,或者物理定律,或者流体和热力学——这些并不都是语言结构。

And so somebody has to go make sure that weather prediction, biology, AI, AI for biology, physical AI, all of that stuff stays, can be pushed to the limits and pushed to the frontier.

所以得有人去确保:天气预报、生物学、生物学 AI、physical AI,所有这些东西都能被推到极限、推到前沿。

We don't build cars, but we wanna make sure every car company has access to great models.

我们不造车,但我们想确保每一家汽车公司都能用上很好的模型。

We don't discover drugs, but I wanna make sure that Lilly has the world's best biology AI systems, so that they can go use it for discovering drugs.

我们不做新药发现,但我想确保 Lilly 拥有世界上最好的生物学 AI 系统,好让他们拿去发现药物。

And so these three fundamental reasons, both in recognizing that AI is not just language, that AI is really broad, that we wanna engage everybody into the world of AI, and then also co-design of AI.

所以就是这三个根本理由:认识到 AI 不只是语言、AI 的范围其实非常广;我们想把所有人都带进 AI 的世界;以及 AI 的协同设计。

Well, I have to say, once again, thank you for open sourcing, really truly open sourcing Nemotron 3 and …

我得再说一次,谢谢你们把 Nemotron 3 真正地、彻底地开源出来,还有……

Yeah, I appreciate you were saying that. We open sourced the models, we open sourced the weights, we open sourced the data, we open sourced how we created it. Yeah, it's pretty amazing.

谢谢你这么说。我们开源了模型,开源了权重,开源了数据,也开源了我们是怎么做出来的。确实挺了不起。

It's really incredible. You're originally from Taiwan and have a close relationship with TSMC.

真的了不起。你出生在台湾,和 TSMC 关系很近。

So I have to ask TSMC I think also is a legendary company in terms of the engineering teams, in terms of the incredible engineering work that they do.

所以我得问问 TSMC——我觉得它也是一家传奇公司,无论是工程团队,还是他们做出的那些惊人的工程。

What do you understand about TSMC culture and their approach that explains how they're able to achieve this singular unmatched success in everything they're doing with semiconductors?

关于 TSMC 的文化和他们的做法,你的理解是什么?是什么解释了他们在半导体上能取得这种独一无二、无人能及的成功?

You know, first of all, the deepest misunderstanding about TSMC is that their technology is all they have.

首先,对 TSMC 最深的误解是:以为技术就是他们的全部。

That somehow they have a really great transistor, and if somebody shows up another transistor, game over.

以为他们不过是有一个特别好的晶体管,而只要别人也拿出一个晶体管,游戏就结束了。

It's the technology and, of course, you know, I don't mean just the transistor, the metallization systems, the packaging, the 3D packaging, the silicon photonics, the, you know, all of the technology that they have.

技术当然重要——我说的不只是晶体管,还有金属化系统、封装、3D 封装、硅光,他们手里所有的技术。

That technology is really what makes the company special. Their technology makes the company special.

那些技术确实让这家公司变得特殊。他们的技术让这家公司特殊。

But their ability to orchestrate the demands, the dynamic demands of hundreds of companies in the world as they're moving up, shifting out, you know, increasing, decreasing, pushing out, pulling in, changing from customer to customer, wafer starting, wafer stopping, emergency wafer starts, you know, all of this dynamics of the world's complexity as the world is shape-shifting all the time, and somehow they're running a factory with high throughput, high yields, really great costs, excellent customer service.

但他们真正厉害的是调度需求的能力——全世界几百家公司动态变化的需求:往上调、往后推、加量、减量、延后、提前、客户之间来回切换、开片、停片、紧急开片。世界一直在变形,所有这些复杂动态都压过来,而他们居然还能把工厂跑出高产出、高良率、很好的成本、极佳的客户服务。

They take their promises seriously.

他们把自己的承诺当真。

They, when your wafer—because they know that they're helping you run your company—when the wafers were promised to show up, the wafers show up, you know, so that you could run your company appropriately.

因为他们知道自己是在帮你经营你的公司——所以承诺哪天到货的晶圆,那天就一定到,好让你能正常经营公司。

And so their system, their manufacturing system is completely miraculous, I would say.

所以我会说,他们的系统、他们的制造系统简直是个奇迹。

Then the second thing is their culture. This culture is simultaneously technology focused on one hand, advancing technology; simultaneously customer service oriented on the other hand.

第二件事是他们的文化。这种文化一方面聚焦技术、推进技术,另一方面同时以客户服务为导向。

A lot of companies are very customer service oriented, but they're not very technology excellent. They're not at the bleeding edge of technology.

很多公司非常以客户服务为导向,但技术上并不卓越,不在技术的最前沿。

There are a lot of companies who are tech, at the bleeding edge of technology, but they're not the best customer service oriented company.

也有很多公司站在技术最前沿,但不是客户服务最好的公司。

And so it just depends on somehow they've, they've balanced these two and they're world-class at both.

而他们不知怎么就把这两件事平衡住了,而且两边都是世界级。

And then probably the third thing is the technology that I most value in them that they created this, you know, this, this intangible called trust.

第三点,可能是我最看重他们的那项"技术":他们创造出了一个叫"信任"的无形之物。

I trust them to put my company on top of them. That's a very big deal.

我信任他们,愿意把我整个公司架在他们身上。这是件非常大的事。

When they trust, I mean, there's a really close relationship there that you've established, and that trust is established based on many years of performance, but there's human relationships involved there as well.

说到信任——你们之间建立了非常紧密的关系,而这份信任是建立在很多年的履约表现之上的,但里面同时也有人与人的关系。

Three decades, I don't know how many tens, hundreds of billions of dollars of business we've done through them, and we don't have a contract. That's pretty great.

三十年,我不知道我们通过他们做了几百亿、几千亿美元的生意,而我们之间没有合同。这挺了不起的。

Amazing. Okay, there's this story … … That in 2013, the founders of TSMC, Morris Chang offered you the chance to become TSMC's chief executive and you said you already had a job. Is this story true?

太惊人了。好,还有一个故事——说 2013 年,TSMC 的创始人张忠谋(Morris Chang)曾邀请你去做 TSMC 的 CEO,而你说你已经有工作了。这个故事是真的吗?

Story is true. I didn't, I didn't dismiss it.

故事是真的。我没有轻率地打发掉它。

But I was deeply honored and, and of course, I knew then as I know now, TSMC is one of the most consequential companies in history.

我深感荣幸。当然,我当时就跟现在一样清楚:TSMC 是历史上最有影响力的公司之一。

And Morris is one of the highest regarded executives and business and personal friend that I've had in my life. And, for him to ask, I was humbled and really honored.

而 Morris 是我这一生里最受敬重的经营者之一,也是生意上和私人的朋友。他来问我,我既谦卑又非常荣幸。

But the work that I'm doing here is really important, and I've seen, you know, in my mind's eye, what NVIDIA was going to be and what the impact that we could have.

但我在这儿做的工作真的很重要,而且我在脑子里已经看见了 NVIDIA 会变成什么样、我们能带来什么影响。

And it was really important work. And it's my responsibility, you know, my sole responsibility to make this happen.

那是非常重要的工作,而让它发生是我的责任,是我唯一的责任。

And so I declined it, not because it wasn't an incredible offer. It's an unbelievable offer, but I simply couldn't take it.

所以我拒绝了——不是因为那不是个了不起的机会。那是个难以置信的机会,但我就是不能接。

I think NVIDIA, both NVIDIA and TSMC are two of the greatest companies in the history of human civilization.

我认为 NVIDIA 和 TSMC 都是人类文明史上最伟大的公司之一。

And running either one, I'm sure, is an incredibly complicated effort and takes… You have to truly be all in.

而经营任何一家,肯定都是极其复杂的事,需要……你必须真正全力投入。

Everybody at every scale, not just at the CEO level. Everybody is really truly all in-

每一个层级上的每一个人,不只是 CEO 那一层。每个人都真真正正全力投入——

Yeah. Yeah, no doubt.

对,毫无疑问。

… To, to accomplish this kind of complexity.

……才能完成这种复杂度的事。

So now I can help both companies.

所以现在我可以帮到这两家公司。

Exactly. So NVIDIA is now the most valuable company in the world. I have to ask, what is the NVIDIA's biggest moat, as the folks in the tech sector say? The edge you have that protects you from the competition.

正是。NVIDIA 现在是全世界市值最高的公司。我得问一句:用科技圈的说法,NVIDIA 最大的护城河是什么?那个保护你不受竞争侵害的优势。

Our single most important property as a company is the install base of our computing platform. Our single most important thing today is the install base of CUDA.

作为一家公司,我们最重要的那一项资产,是我们计算平台的 install base。今天我们最重要的一件事,就是 CUDA 的 install base。

Now, the reason why 20 years ago, of course, there was no install base. But what makes… And if somebody came up with a GUDA or TUDA, it wouldn't make any difference at all.

当然,二十年前根本没有 install base。而如果今天有人搞出一个 GUDA 或者 TUDA,那完全不会有任何影响。

And the reason for that is because it's never been just about the technology. The technology, of course, was incredible, visionary.

原因是这件事从来就不只是技术。技术当然很了不起、很有远见。

But it's the fact that the company was dedicated to it, stuck with it, expanded its reach.

但关键在于这家公司死磕它、坚持它、把它的覆盖面一点点扩大。

It wasn't three people that made CUDA successful. It was 43,000 people that made CUDA successful.

让 CUDA 成功的不是三个人,是四万三千人。

And the several million developers that believed in us that trusted that we were going to continue to make CUDA 1, 2, 3, 13, that they decided to port and dedicate their software on top of it, their mountain of software on top of it.

还有那几百万开发者,他们信我们、信我们会一路把 CUDA 做到 1、2、3、13 版,所以他们决定把自己的软件、把自己那座软件大山移植过来、架在它上面。

And so the install base is the number one most important advantage.

所以 install base 是第一重要的优势。

That install base, when you amplify it with the velocity of our execution at the scale that we're talking about, no company in history had ever built systems of this complexity, period.

而当你用我们在这个体量上的执行速度去放大这个 install base——历史上没有任何一家公司造过这种复杂度的系统,没有,一个都没有。

And then to build it once a year is impossible.

而要做到一年造一代,更是不可能的。

And that velocity combined with the install base, in the developer's mind, you just go now, take the developer's mind. From the developer's perspective, if I support CUDA, tomorrow it'll be 10 times better. I just have to wait six months on average.

那个速度叠加 install base,在开发者心里……你站到开发者的位置想。从开发者的角度:如果我支持 CUDA,明天它就会好十倍,我平均只要等六个月。

Not only that, if I develop it on CUDA, I reach a few hundred million people, computers.

不只如此,如果我在 CUDA 上开发,我能触达好几亿人、好几亿台计算机。

I'm in every cloud, I'm in every computer company, I'm in every single industry, I'm in every single country.

我在每一个云上、每一家计算机公司里、每一个行业里、每一个国家里。

So if I create an open source package and I put it on CUDA first, I get these both attributes simultaneously.

所以如果我做一个开源包,先放到 CUDA 上,我就同时拿到了这两个属性。

And not only that, I trust 100% that NVIDIA is going to keep CUDA around and maintain it and improve it and keep optimizing the libraries for as long as they shall live.

还不只如此:我百分之百相信,NVIDIA 会一直把 CUDA 留着、维护它、改进它,并且在他们活着的每一天继续优化那些库。

You could take that to the bank, and that last part, trust.

这话你可以拿去当保票——最后那一项,就是信任。

You put all that stuff together, if I were a developer today, I would target CUDA first. I would target CUDA most.

把这些全部加起来:如果我今天是个开发者,我会优先瞄准 CUDA,会最多地瞄准 CUDA。

And that's the reason that I think in the final analysis is our first, that's even our first-

所以我认为归根结底,这是我们的第一——这甚至是我们的第一——

… core advantage. Our second one is our ecosystem.

……核心优势。第二个是我们的生态。

The fact that we vertically integrated this incredibly complex system, but we integrate it horizontally into every single company's computers.

我们把这套极其复杂的系统做了垂直整合,但又把它横向整合进每一家公司的计算机里。

We're into Google Cloud, we're into Amazon, we're in Azure. You know, we're ramping up AWS like crazy right now.

我们进了 Google Cloud、进了 Amazon、进了 Azure。我们现在在 AWS 上疯狂扩张。

We're in new companies like CoreWeave and Nscale. We're in supercomputers at Lilly. We're in enterprise computers. We're at the edge in radio base stations.

我们进了 CoreWeave、Nscale 这样的新公司。我们在 Lilly 的超级计算机里,在企业计算机里,在边缘的无线基站里。

You know, I mean, it's just crazy. One architecture is in all these different systems.

这真的太疯狂了。同一个架构存在于所有这些不同的系统里。

We're in cars, we're in robots, we're in satellites, we're out in space.

我们在汽车里、在机器人里、在卫星里,我们还在太空里。

And so the fact that you have this one architecture and the ecosystem is so broad, it basically covers every single industry in the world.

所以你有这么一个统一架构,加上如此宽广的生态,它基本上覆盖了世界上每一个行业。

Well, how does the CUDA install base evolve into the future with AI factories as a moat? What do you… Do you think it's possible that NVIDIA of the future is all about the AI factory?

那么当 AI 工厂成为护城河,CUDA 的 install base 往未来会怎么演变?你觉得未来的 NVIDIA 有可能整个都围绕 AI 工厂展开吗?

Well, the unit of computing used to be GPU to us. Then it became a computer, then it became a cluster. Now it's an entire AI factory.

对我们来说,计算的单位以前是 GPU。后来变成了一台计算机,再后来变成了一个集群。现在是一整座 AI 工厂。

When I see a computer, when I see what NVIDIA builds, in the old days, I would, you know, I visualize the chip.

当我看一台计算机、看 NVIDIA 造的东西,过去我脑子里浮现的是那颗芯片。

And then when I announced the new product, new generation, like, "Ladies and gentlemen, we're announcing Ampere today," I'd pick up the chip. That was my mental model- … of what I was building.

然后我发布新产品、新一代的时候,会说"各位,我们今天发布 Ampere",然后把芯片举起来。那就是我对"我在造什么"的心智模型。

Today, I wouldn't… Picking up the chip is kind of still adorable.

今天我不会……不过把芯片举起来还是挺可爱的。

But it's adorable. It's not my mental model of what I'm doing.

它确实可爱,但那已经不是我对自己在做什么的心智模型了。

My mental model is this giant gigawatt thing that has power generations connected to the grid. It's got cooling systems and networking of incredible monstrosity, you know.

我的心智模型是那个巨大的、吉瓦级的东西,带着接入电网的发电设施。它有冷却系统,还有庞大到骇人的网络。

10,000 people are in there trying to install it, hundreds of networking engineers in there, thousands of engineers behind it trying to power it up.

里面有一万人在试着把它装起来,有几百个网络工程师在里面,背后有几千个工程师在设法把它通上电。

You know, powering up one of those factories, as you know, it's not somebody going, "It's on now." It takes thousands of people to bring it up.

给那样一座工厂上电,不是某个人喊一句"通了"就完事。那需要几千人才能把它启动起来。

So mentally, you're actually… When you're thinking about a single unit of compute, you're like literally, when you go to bed at night, you're thinking now about a collection of racks, so pods, not individual chips.

所以在你心里……当你想"一个计算单元"的时候,你晚上睡觉前想的其实是一堆机架、是 pod,而不是单颗芯片。

Entire infrastructure. And I'm hoping my next click is when I'm thinking about building computers, it's planetary scale. That'll be the next click.

是整套基础设施。而我希望我的下一档是:当我想造计算机的时候,想的是行星级。那会是下一档。

Well, what do you think about the space angle that Elon has talked about, doing compute in space for solving some of the… It makes some of the energy issues in terms of scaling energy easier.

那 Elon 讲的太空这个角度你怎么看?在太空里做计算,来解决一部分……它能让能源扩展的一些问题变得更容易。

Cooling issues is not easy. Yeah.

散热问题就不容易了。

Cooling. Well, there's a large number of engineering complexities involved with that. So what… You know, NVIDIA has also announced that you're already thinking about that.

散热。这里面涉及大量的工程复杂性。而 NVIDIA 也宣布过你们已经在想这件事了。

Yeah, we're already there. NVIDIA GPUs are the first GPUs in space.

对,我们已经在那儿了。NVIDIA 的 GPU 是第一批上太空的 GPU。

And I didn't realize it, it was so interesting to… I would have declared it maybe. We're in space. You know, little, little astronaut suit on one of our GPUs.

我当时都没意识到,这事挺有意思的——我本来应该宣布一下的。我们在太空里。给我们其中一块 GPU 套个小小的宇航服。

But we've been in space. It's the right place to do a lot of imaging.

但我们确实已经在太空里了。那儿是做大量成像的合适地方。

You know, because those satellites have really high resolution imaging systems, and they're sweeping the Earth, you know, continuously now.

因为那些卫星带着分辨率非常高的成像系统,而且现在是在持续扫过整个地球。

And you want, you know, centimeter scale imaging that is done continuously for the world, so that, you know, you'll basically have real time telemetry of everything.

你要的是对全世界持续做厘米级成像,这样你基本上就有了一切事物的实时遥测。

You don't wanna beam that back down to Earth. It's just, you know, petabytes and petabytes of data.

你不会想把这些数据传回地球——那是一个 PB 又一个 PB 的量。

You gotta just do AI right there at the edge, throw away everything you don't need, you've seen before, didn't change, and then just keep the stuff that you need.

你只能就在边缘那里做 AI:把不需要的、见过的、没变化的全部扔掉,只留下你需要的部分。

And so AI had to be done at the edge. Obviously we have 24/7 solar, if we put it at the polars. And but, you know, there's no conduction, no convection.

所以 AI 必须在边缘做。显然,如果我们把它放在极地轨道,就有全天候的太阳能。但那里没有传导,也没有对流。

And so, you know, you're pretty much just radiation. And but, you know, space is big. I guess, you know, we're just gonna put big, giant radiators out there.

所以你基本上只剩辐射这一条路。不过太空很大,我猜我们就往那儿放几个巨大的散热器。

How crazy of an idea do you think it is? Like is this five years out, 10 years out, 20 years out? So we're talking about blockers for AI scaling.

你觉得这个想法有多疯狂?这是五年后、十年后,还是二十年后的事?我们现在说的是 AI 扩展的阻碍。

You know, I'm just so much more practical. I look for where my next, next bucket of opportunities are first. Meanwhile, I'm cultivating space.

我要务实得多。我先去找我下一桶、再下一桶机会在哪儿。同时我在培育太空这件事。

And so I send, I send engineers to go work on the problem. We're starting to… We're learning a lot about it.

所以我派工程师去攻这个问题。我们正在学到很多东西。

How do we deal with radiation? How do we deal with degrading performance? How do we deal with a continuous testing and attestation of defects? And you know, how do we deal with redundancy? And how do we degrade gracefully and things like that?

辐射怎么应对?性能衰减怎么应对?怎么做持续测试和缺陷认证?冗余怎么做?怎么优雅降级?

And so we could do a… What about software? How do you think about software and redundancy and performance out in space?

还有软件呢?在太空里,软件、冗余和性能该怎么想?

Make it so that the computer never breaks, it just gets slower, you know. And I… So we could start doing a lot of engineering exploration upfront.

把它做成计算机永远不会坏,只会变慢。所以我们可以先做很多前期的工程探索。

But in the meantime, my favorite answer is eliminate waste. You know, we've got all that idle power, I want to evacuate it as fast as possible.

但在这期间,我最喜欢的答案是:消除浪费。我们手上有那么多闲置的电力,我想尽快把它抽干。

Yeah. There, there… Yeah, there's a lot of low-hanging fruit here on Earth- … That we can utilize for the AI scaling.

对,地球上还有很多唾手可得的东西——我们可以拿来支撑 AI 扩展。

Quick pause. Quick 30-second thank you to our sponsors. Check them out in the description. It really is the best way to support this podcast. Go to lexfridman.com/sponsors.

先暂停一下,用 30 秒感谢我们的赞助商。在简介里看看他们,这真的是支持这档播客最好的方式。去 lexfridman.com/sponsors。

We got Perplexity for curiosity-driven knowledge exploration, Shopify for selling stuff online, LMNT for electrolytes, Fin for customer service AI agents, and Quo for a phone system, like calls, texts, contacts, for your business. Choose wisely, my friends.

我们有 Perplexity 做好奇心驱动的知识探索,Shopify 做在线卖货,LMNT 做电解质饮品,Fin 做客服 AI agent,以及 Quo 提供电话系统——通话、短信、联系人,给你的生意用。各位,明智地选择吧。

And now, back to my conversation with Jensen Huang. Do you think NVIDIA may be worth 10 trillion at some point? Let's, let's ask it this way. What does the future of the world look like where that's true?

现在回到我和黄仁勋的对话。你觉得 NVIDIA 有可能在某个时点值 10 万亿吗?换个问法吧:如果那是真的,那时候世界的未来会是什么样?

I think that NVIDIA's growth is extremely likely, and in my mind, inevitable. And let me explain why.

我认为 NVIDIA 的增长极有可能发生,在我看来是必然的。我来解释为什么。

We're the largest computer company in history. That alone should beg the question, why? And the reason of course… Two reasons. First, two foundational technical reasons.

我们是历史上最大的计算机公司。光这一点就该让人问一句:为什么?原因当然有两个,两个基础性的技术原因。

The first reason is that computing went from being a retrieval-based, file retrieval system. Almost everything is a file… We pre-write something, we pre-record something.

第一个原因是:计算从原来那种基于检索的、文件检索的系统变了。过去几乎一切都是文件——我们预先写好、预先录好。

You know, we draw something, we put it on the web, we put it in a file. And we use a recommender system, some smart filter, to figure out what to retrieve for you.

画点什么,放到网上,存成文件。然后我们用一个推荐系统、某种聪明的过滤器,来决定给你取回哪一份。

And so we were a pre-recording, human pre-recording, and file retrieving system. That's what a computer is, largely.

所以我们过去是一个"人类预先录制 + 文件检索"的系统。计算机大体上就是这个东西。

To now, AI computers are contextually aware, which means that it has to process and generate tokens in real time.

而现在,AI 计算机是有上下文意识的,这意味着它必须实时处理并生成 token。

So we went from a retrieval-based computing system to a generative-based computing system.

所以我们从检索式的计算系统,变成了生成式的计算系统。

We're gonna need a lot more processing in this new world than in the old world. We need a lot of storage in the old world. We need a lot of computation in this new world.

在这个新世界里,我们需要的处理量比旧世界多得多。旧世界我们需要大量存储;新世界我们需要大量计算。

And so that's the first part of it. We fundamentally changed computing and the way how computing is done. The only thing that would cause it to go back……

这是第一部分。我们从根本上改变了计算,以及计算的完成方式。唯一能让它退回去的东西……

is if this way of computation, this way of computing generating information that's contextually relevant, situationally aware, that is grounded on new insight before it generates information, this computation-intensive way of doing computing would only go back if it's not effective.

……就是如果这种计算方式无效:这种生成上下文相关、情境感知、在生成信息之前先落在新洞见上的计算方式,这种极度吃算力的做法——只有当它无效时才会退回去。

So if… For the last 10, 15 years while working on deep learning, if at any single moment I would have come to the conclusion that, "You know what? This is not gonna work out. I think this is a dead end."

所以在过去十年、十五年做深度学习的过程里,如果任何一个时刻我得出结论:"你知道吗?这条路走不通,我觉得是死胡同。"

Or, "It's not gonna scale, it's not gonna solve this modality, not gonna be used in this application." Then, of course, I would feel very differently about it, but I think the last five years has given me more confidence than the previous ten years.

或者"它扩展不上去,它解决不了这个模态,它在这个应用里用不上"——那我当然会有完全不同的感受。但我觉得过去五年给我的信心,比前面十年更多。

The second idea is computers, because it was a storage system, it was largely a warehouse. We're now building factories.

第二个想法是:因为过去的计算机是个存储系统,它基本上是个仓库。而我们现在在建工厂。

Warehouses don't make much money. Factories directly correlates with the company's revenues.

仓库不太赚钱。工厂则直接跟公司的收入挂钩。

And so, the computer did two things. Not only did it change the way it did it, its purpose in the world changed.

所以计算机发生了两件事:它不只改变了做事的方式,它在这个世界上的目的也变了。

It's no longer a computer, it's a factory. It's a factory, it's used for generation of revenues.

它不再是一台计算机,它是一座工厂。它是一座工厂,用来产生收入。

We're now seeing not only is this factory generating products, commodities that people want to consume, we're seeing that the commodities are so interesting, so valuable to so many different audiences that the tokens are starting to segment, like iPhones.

而我们现在看到的不只是这座工厂在生产产品、生产人们想消费的商品——我们看到这些商品有意思到、对这么多不同受众有价值到,token 开始像 iPhone 一样分层了。

You have free tokens, you have premium tokens, and you have several tokens in the middle.

有免费 token,有高端 token,中间还有好几档。

And so intelligence, as it turns out, you know, it's a scalable product.

所以事实证明,智能是一个可以分层定价的产品。

There's extremely high intelligence products, tokens that you could… that are used for specialized things, people be willing to pay.

有极高智能的产品、用于专门用途的 token,人们愿意为它付钱。

You know, the idea that somebody's willing to pay $1000 per million tokens is just around the corner. It's not if, it's only when.

有人愿意为一百万 token 付 1000 美元这件事,就在眼前了。这不是"会不会",只是"什么时候"。

And so, so now we're seeing that the commodity that this factory makes is actually valuable, and is revenue generating and profit generating.

所以我们现在看到:这座工厂生产的商品其实是有价值的,是能产生收入和利润的。

Now the question is how many of these factories does the world need? How many tokens does the world need? And how much is society willing to pay for these tokens?

那么问题变成:这个世界需要多少座这样的工厂?世界需要多少 token?社会又愿意为这些 token 付多少钱?

And what would happen to the world's economy if the productivity were to improve so substantially? What would happen…

而如果生产力提升得如此之大,世界经济会发生什么?会发生什么……

Are we, are we gonna discover new drugs, new products, new services?

我们会不会发现新药、新产品、新服务?

And so when you take these things in combination, I am absolutely certain that the world's GDP is going to accelerate in growth.

把这些合起来看,我百分之百确定,世界 GDP 的增长会加速。

I'm absolutely certain the percentage of that GDP that will be used for computation will be 100 times more than the past—mm-hmm—because it's no longer a storage unit. It's a product generation unit.

我百分之百确定,GDP 中用于计算的比例会是过去的 100 倍——因为它不再是一个存储单元,而是一个产品生产单元。

And so when you look at it in that context and then you back into what is NVIDIA's, what does NVIDIA sh—what does NVIDIA do and how much of that new economics, new industry would we have to benefit t—to address, I think we're gonna be a lot, lot bigger.

所以当你在这个语境下看,再倒推回 NVIDIA 是做什么的、这个新经济、新产业里有多少是我们要去覆盖、去受益的,我觉得我们会大得多、大得多。

And then the rest of it, to me, is: is it possible for NVIDIA to be a, you know, $3 trillion revenue company in the near future? The answer is, of course, yes.

剩下的问题对我来说是:NVIDIA 在不远的将来有没有可能成为一家 3 万亿美元收入的公司?答案当然是:有可能。

And the reason for that is because it's not limited by any physical limits. There's nothing that I see that says, you know, gosh $3 trillion is not possible.

原因是它并不受任何物理极限的限制。我看不到任何东西告诉我"天哪,3 万亿不可能"。

And as it turns out, NVIDIA's supply chain is—the burden is shared by 200 companies.

而事实上,NVIDIA 的供应链——这个负担是由 200 家公司分担的。

And the fact that we scale out on the backs of, with the partnership of this ecosystem, the question is: do we have the energy to do so? And surely we will have the energy to do so.

我们是靠这个生态的背、靠这个生态的合作往外扩的,所以问题变成:我们有足够的能源做到吗?而我们肯定会有。

And so all of these things combined, that number is just a number, you know?

所以这些加在一起,那个数字只是个数字。

And I still remember, NVIDIA was a… the first time we crossed a billion dollars, I was reminded of a CEO who told me, "You know, Jensen, it's theoretically impossible for a fabless semiconductor company to exceed a billion dollars."

我还记得,NVIDIA 第一次跨过十亿美元的时候,我想起有位 CEO 跟我说过:"黄仁勋,一家无晶圆厂的半导体公司在理论上不可能超过十亿美元。"

And I won't bore you with why, but of course it's illogical and there's a lot of evidence we're not.

我就不拿他的理由烦你了,但那当然不合逻辑,而且有大量证据表明我们并没有被卡住。

And then somebody told me, "You know, Jensen, you'll never be more than $25 billion because of some other company." Somebody told me that, "You'll never be, you know, because…"

然后又有人跟我说:"黄仁勋,因为某家别的公司,你永远超不过 250 亿美元。"有人这么跟我说过:"你永远不会,因为……"

And so those aren't principled, first principled reason thinking. And the simple way to think about that is what is it that we make and how large is the opportunity that we can create?

这些都不是有原则的、第一性原理式的思考。简单的想法是:我们做的是什么?我们能创造的机会有多大?

Now, NVIDIA is not in the market share business. Almost everything that I just talked about don't exist. That's the part that's hard.

NVIDIA 做的不是抢市场份额的生意。我刚才讲的几乎所有东西,现在都还不存在。难的地方就在这里。

You know, if NVIDIA was a $10 billion company trying to take NVIDIA's share, then it's easy to see for shareholders that, oh, yeah, if they could just take 10% share, they could be this much larger.

如果 NVIDIA 是一家 100 亿美元的公司、想去抢 NVIDIA 的份额,那股东很容易看懂:哦对,只要抢到 10% 份额,就能大这么多。

But it's hard for people to imagine how large we could be because there's nobody I could take share from. You know?

但大家很难想象我们能变得多大,因为我没有份额可抢。

And so I think that that's one of the challenges for the world is the imagination of the future.

所以我觉得这是这个世界的一个难题:对未来的想象力。

But I got plenty of time, and I'll keep reasoning about it, and I'll keep talking about it, and every single GTC will become more and more real.

但我有大把时间,我会继续推演,继续讲,而每一届 GTC 都会让它变得更真实一点。

You know, and then more and more people will talk about it, and one of these days, you know, we'll get there. But I'm 100% we'll get there.

然后越来越多的人会谈它,总有一天我们会走到那儿。但我百分之百确定我们会走到那儿。

Yeah, this view of you know, token factories essentially, this token per second per watt, and every token having value.

对,这个"token 工厂"的视角、每瓦每秒 token 数,以及每个 token 都有价值。

Like it's an actual thing that brings value, and it brings different kinds of value, different amounts of value to different people with value.

它是一个真实带来价值的东西,而且它对不同的人带来不同种类、不同数量的价值。

That's the actual product—it really could be loosely thought of as the token. And so you have a bunch of token factories.

那才是真正的产品——可以粗略地把它想成 token。所以你有一堆 token 工厂。

And then it's very easy, first principles, to imagine a future, given all the potential things that AI can solve, that you're going to need an exponential number more of token factories.

然后从第一性原理出发,很容易想象出一个未来:考虑到 AI 能解决的所有潜在问题,你会需要指数级更多的 token 工厂。

Yeah. And what's really interesting, the reason why I was so excited about it, the iPhone of tokens arrived.

对。而真正有意思的、我之所以这么兴奋的原因是:token 的 iPhone 时刻到了。

What do you call it? Wait, are you saying OpenClaw's iPhone?

你叫它什么?等等,你是说 OpenClaw 就是那个 iPhone?

Yeah.

对。

That's interesting.

这有意思。

Agents.

是 agent。

Yeah, agents. True.

对,agent。确实。

Agents in general. The iPhone of tokens arrived. It is the fastest-growing application in history. It went straight up. Went straight up.

泛指 agent。token 的 iPhone 时刻到了。它是历史上增长最快的应用。它是直线上去的,直线上去。

That says something.

这本身就说明了点什么。

Yep, there's no question OpenClaw is the iPhone of tokens.

对,毫无疑问,OpenClaw 就是 token 的 iPhone。

Is there something truly, as you know, something truly special happening from about December, where people have really woke up to the power of Claude Code of Codex, of OpenClaw?

大概从十二月开始,是不是真有什么特别的事情在发生——人们真正意识到了 Claude Code、Codex、OpenClaw 的力量?

I mean, I'm embarrassed to admit that on the way here in the airport, I've… It's the first time I've done this in public. I was programming, quote unquote, by talking to my laptop.

我有点不好意思承认,来这儿的路上在机场,我……那是我第一次在公共场合这么干:我在对着笔记本说话,所谓"编程"。

Yeah, exactly.

对,正是。

And I was embarrassed because I was pretending like I'm talking to a human colleague.

我不好意思是因为我像在跟一个人类同事说话。

I'm not sure how I feel about the future where everybody- … is walking around talking to their AI, but it's such an efficient way to get stuff done.

我不太确定自己对那个"所有人都边走边跟自己的 AI 说话"的未来是什么感觉,但这确实是把事情办完的极高效方式。

And it's more likely that your AI is bothering you all the time. And the reason for that is because it's getting stuff done so fast.

而更可能的情况是:你的 AI 一直在烦你。原因是它把事情办得太快了。

It's reporting back to you, "I got that done." "You know, what do you want me to do next?"

它不停回报你:"那件事我搞定了。""你接下来想我干什么?"

You know, it… That's the part that I think most people don't realize is the person who's gonna be chatting with them, texting them most, is their, is their claws or lobster.

我觉得大多数人还没意识到的一点是:以后跟你聊天、给你发消息最多的那个"人",会是你的 claw、你的龙虾。

What an incredible future. I read that you attribute a lot of your success to your ability to work harder than anyone and withstand more suffering than anyone.

多么了不起的未来。我读到你把自己很大一部分成功归因于两件事:比任何人都能更拼命工作,以及比任何人都能承受更多痛苦。

So we can list many of the things that entails. I mean, dealing with failure, the cost and engineering problems we've talked about.

这里面包含的东西可以列一长串:应对失败、我们聊过的成本和工程难题。

The human problems, uncertainty, responsibility, exhaustion, embarrassment, the near-death company moments that you've mentioned but also the pressure.

还有人的问题、不确定性、责任、疲惫、尴尬,你提过的公司濒死时刻,以及压力。

Now, as the CEO of this company that economies and nations strategize around, plan their financial allocations around, plan their AI infrastructure around, how do you deal with this much pressure?

现在,作为这样一家公司的 CEO——各国经济体围着它制定战略、安排财政拨款、规划自己的 AI 基础设施——你怎么应对这么大的压力?

What gives you strength, given how many nations and peoples depend on you?

在这么多国家和人群都依赖你的情况下,什么给你力量?

I'm conscious about the fact that NVIDIA's success is very important to the United States. We generate enormous amounts of tax revenues. We established technology leadership for our nation.

我很清楚,NVIDIA 的成功对美国非常重要。我们创造了巨额税收,我们为国家建立了技术领先地位。

Technology leadership is important for national security. National security not just in one aspect of national security, all aspects of national security.

技术领先对国家安全很重要——不只是国家安全的某一个方面,而是所有方面。

When our country's more prosperous, we could do a better job with domestic policies and helping social benefits.

当国家更繁荣,我们就能在国内政策和社会福利上做得更好。

Because we're generating so much re-industrialization in the United States, we're creating mountains of jobs.

因为我们在美国带动了这么大规模的再工业化,我们创造了成山的就业。

We're helping shift how we build things back to the United States in so many different plants, chips, computers, and of course, these AI factories. I'm completely aware that, that…

我们在帮着把"东西怎么造"这件事搬回美国,分布在那么多不同的工厂里:芯片、计算机,当然还有这些 AI 工厂。我完完全全清楚这些。

And I have the benefit, and this is a real gift with mainstream investors, teachers, policemen who have somehow, for whatever reason, invested in NVIDIA or because they watched Jim Cramer, bought some stock and now are millionaires.

而我还有一份好运,这真是一份礼物:那些普通投资者、老师、警察,不知出于什么原因投了 NVIDIA,或者因为看了 Jim Cramer 的节目买了些股票,现在成了百万富翁。

And I am completely aware of that circumstance. I'm aware of the circumstance that NVIDIA is central to a very large network of ecosystem partners behind us and downstream from us.

我完全清楚这个处境。我清楚 NVIDIA 处在一个庞大网络的中心:身后和下游都是生态伙伴。

And so the way I deal with that is exactly what I just did. I reason about what is… what is it that we're doing? What is it causing?

我应对的方式,就是我刚才做的那件事:我推演——我们在做什么?它造成了什么?

What's the impact that has on other people benefit, you know, positively or even through great burden, for example, to supply chain?

它对其他人有什么影响——是正面的受益,还是像对供应链那样带来巨大负担?

And the question is therefore, what are you gonna do about it?

所以问题接着就是:那你要拿它怎么办?

In almost everything that I feel, I break it down, I reason about, "Okay, what's the circumstance? What has changed? What's hard? And what am I gonna do about it?" And I'm…

几乎我感受到的每一件事,我都会把它拆开、推演一遍:"好,处境是什么?什么变了?哪里难?我要拿它怎么办?"

I break it down, decompose the problem, and the decomposition of these circumstances turns it into manageable things that I can do.

我把它拆开、分解掉,而对这些处境的分解会把它变成一件件我做得到的、可管理的事。

And the only thing that after that I could do is, "Did you do it? Did you either do it or did you get somebody else to do it?

在那之后我唯一能做的事就是问:"你做了吗?你是自己做了,还是找了别人去做?"

And if you didn't do it, you reasoned that you need to do it, and you didn't do it, and you didn't get anybody else to do it, then stop crying about it."… you know? And so, and so-

"如果你没做——你明明推演出这件事必须做,你却没做,也没找任何人去做——那就别再抱怨了。"

so I'm fairly tough on myself. And, but I also break things down so that I don't panic.

所以我对自己相当严厉。但我也会把事情拆开,这样我就不会慌。

I can go to sleep because I've made the list of things that needed to be done, and I've made sure that everything that could put our company in harm's way, could put my partners in harm's way, put our industry in harm's way, I've told somebody.

我能睡着,因为我已经把该做的事列成清单了。而且我确保:任何可能让公司受损、让合作伙伴受损、让整个行业受损的事情,我都已经告诉了某个人。

Everything that I feel could put anybody in harm's way, I've told someone. And I've told that someone who could do something about it.

任何我觉得可能让任何人受损的事,我都告诉了某个人——而且是告诉那个能对此做点什么的人。

And so I've gotten it off my chest or I'm doing something about it. And so after that, Lex, what else can you do?

所以我要么把它从心里卸下来了,要么正在处理它。在那之后,Lex,你还能做什么呢?

So given all the insane, intense amount of suffering on the journey of building up NVIDIA, have you hit low points psychologically?

那么,在建起 NVIDIA 这一路上那些疯狂而剧烈的痛苦里,你在心理上有跌到过低谷吗?

Oh, yeah. Oh, yeah. Sure. All the time. All the time.

有,有。当然。一直都有,一直都有。

And there-

那——

All the time

一直都有。

… you just break down the problem into pieces? See what you could do about it?

……你就是把问题拆成一块块?看看自己能做点什么?

And part of it, Lex, part of it is forgetting. One of the most important attributes of AI learning, as you know, is, right? Systematic forgetting.

还有一部分,Lex,一部分是遗忘。AI 学习最重要的属性之一,你知道的,对吧?系统性遗忘。

You need to know when to forget some things. You can't memorize everything. You can't keep everything and, you know, you don't want to carry everything.

你得知道什么时候该忘掉一些东西。你不可能记住一切,不可能留住一切,你也不想背着一切走。

One of the things that I do very quickly is decompose the problem, I reason about the problem, and I share the load with it. When I say I tell everybody, I'm essentially sharing that burden.

我很快会做的一件事是:把问题分解、推演一遍,然后把负担分摊出去。我说"我告诉所有人",本质上就是在分摊那份负担。

As quickly as possible. Whatever worries me, tell somebody else. Don't just keep it. You know, don't freak them out.

越快越好。任何让我担心的事,就告诉别人,别自己憋着。但也别把人吓着。

Decompose the problem into smaller parts and get people to, and inspire them to be able to go do something about it.

把问题分解成更小的部分,让大家、激励大家去对它做点什么。

But part of it is just forgetting. You know, like, a lot of it is you gotta be tough on yourself. You know, just come on, stop crying about it. Let's get going. You know? And then you get out of bed.

但另一部分就是遗忘。很大一部分是你得对自己狠:得了,别抱怨了,我们上。然后你就从床上起来了。

And then the other part is you're attracted to the next shiny light, the next future, the next opportunity, the next, "Okay, that's behind us. What's next?"

还有一部分是:你被下一道亮光吸引住了——下一个未来、下一个机会、下一个"好,那个过去了,下一个是什么?"

It's a lot, I think, you know, you watch this with great athletes. They just worry about the next point. The last point is behind them. The embarrassment, the, you know- … the setback.

我觉得你在顶级运动员身上能看到很多这种东西。他们只操心下一分,上一分已经在身后了。那些尴尬,那些挫败。

You know, and because I do so much of my job publicly, you know? Lex, you do a fair amount of your job publicly too. And so I do a lot of my job publicly.

而且因为我很大一部分工作是公开做的——Lex,你也有相当一部分工作是公开做的。

And so you know, I say a lot of things that seem sensible at the time or funny at the time, mostly it's just because it's funny to me at the time.

所以我说了很多当时听起来合理、或者当时觉得好笑的话,大多数只是因为那一刻我自己觉得好笑。

And then, you know, you reflect on it, it's less funny, but…

然后你回头再想,就没那么好笑了,不过……

Yeah. No, trust me, I know. But you basically allow yourself to be pulled by the light of the future. Forget the past and just keep-

对。相信我,我懂。但你基本上是让自己被未来那道光拉着走。忘掉过去,就一直——

That's right.

没错。

… keep working towards that. I mean, you did say, there's this kind of famous thing you said that if you knew how hard it would be to build NVIDIA it turned out to be—what is it? A million times more hard than you anticipated—that you wouldn't do it.

……一直朝那个方向做。你说过一句挺有名的话:如果你早知道建 NVIDIA 有多难——是多少?比你预想的难一百万倍——你就不会去做了。

Yeah, right.

对,是的。

But isn't… You know, when I hear that, that's probably true about everything worth doing, right?

但难道不是……我听到这话时想,所有值得做的事大概都是这样,对吧?

Exactly. That is, by the way, what I was trying to explain, is that there's an incredible superpower of having the mind of a child.

正是。顺便说,这就是我想解释的:拥有一颗孩子的心,是一种不可思议的超能力。

You know? And I say to myself oftentimes when I look at something, and almost everything my first thought is, "How hard can it be?"

我常常在看一件事的时候对自己说——几乎每一次,我的第一个念头都是:"这能有多难?"

You know? And so you get yourself into that mode, how hard could it be? And nobody's ever done it. It looks gigantic. It's gonna cost hundreds of billions of dollars. It's gonna take, you know, all this…

于是你就把自己带进那个状态:这能有多难?而且从来没人做过。它看起来庞大无比。它要花几千亿美元。它要耗掉……

And you just go, "Yeah, but how hard could it be?" You know? How hard could it be?

而你就说:"是啊,但这能有多难?"这能有多难?

And so, you gotta get yourself into that state of mind. You don't wanna actually over-simulate everything and all the setbacks and all the trials and tribulations and all the disappointments.

所以你得把自己带到那个心态里。你并不想真的把一切、所有挫折、所有磨难和折腾、所有失望都提前模拟一遍。

You don't wanna simulate all that in advance. You don't wanna know that.

你不想提前模拟那些。你不想知道那些。

You wanna go into a new experience thinking it's gonna be perfect, it's gonna be great, it's gonna be incredibly fun.

你想带着"这会很完美、会很棒、会特别好玩"的心态进入一段新经历。

And then while you're there, you know, you need to have endurance, you need to have grit, so that when the setbacks actually happen, and those setbacks are gonna surprise you, the disappointments are gonna surprise you, the embarrassments are gonna surprise you, the humiliations are gonna surprise you.

然后到了里面,你需要耐力,需要 grit——这样当挫折真的发生时。而那些挫折会让你意外,那些失望会让你意外,那些尴尬会让你意外,那些屈辱会让你意外。

You just can't let… Now you just gotta turn on the other bit, which is just forget about it. Move on, keep moving.

你不能让……这时候你要把另一个开关打开,就是:忘掉它。往前走,继续走。

And to the extent that my assumptions about the future and why the future is gonna manifest, so long as those assumptions and that input doesn't change or didn't change materially, then I should expect that the output won't change.

而只要我关于未来、关于未来为什么会实现的那些假设和输入没有变、没有实质性变化,那我就该预期输出不会变。

And so my simulated output of the future is still gonna happen. And if it's still gonna happen, I'm still gonna go after it.

所以我模拟出来的那个未来仍然会发生。而如果它仍然会发生,我就仍然会去追它。

I believe it's gonna, you know, and so there's a combination of two or three human characteristics: the ability to go into an experience fresh-minded, the ability to forget the setbacks, the ability to believe in yourself, you know, to believe what you believe and stay true to that belief.

我相信它会。所以这里是两三种人格特质的组合:能以一颗新鲜的心进入一段经历的能力;能忘掉挫折的能力;能相信自己的能力——相信你所相信的,并对那个信念保持忠实。

But you're constantly reevaluating.

但你同时在不断重新评估。

This combination of three, four, five things I think is really important for resilience.

这三、四、五样东西的组合,我觉得对韧性非常重要。

And, you know, I'm fortunate that whatever life experiences led to this, I've got kind of those four, five things.

我很幸运,不管是什么样的人生经历造成的,我大概是有这四五样东西的。

You know, I'm always curious, always learning. I'm always learning from everybody, you know? I'm always asking my…

我一直很好奇,一直在学。我一直在从每个人身上学。

And because I'm humble about everything, I'm always thinking, "Gosh, they did that so nicely. They did that so wonderfully." You know, I wonder what they're thinking through. How do they… So I'm simulating everybody.

而因为我对一切都保持谦卑,我总在想:"天哪,他们那件事做得真漂亮,做得真好。"我会好奇他们是怎么想通的。所以我在模拟每一个人。

In a lot of ways, you know, I'm emulating almost everybody I watch, right? You're empathetic towards everything that they do that you're observing and respect. And so you're constantly learning and, you know.

很多时候,我几乎在模仿我看到的每一个人。你对他们所做的、你观察到并尊重的一切,都抱有共情。所以你就在不断地学。

You're now one of the wealthiest people on Earth. One of the most successful humans on Earth. Is it harder to be humble and to be able to…

你现在是地球上最富有的人之一,也是最成功的人之一。保持谦卑变难了吗?

Do you feel the effect of money and power and fame in making it harder for you to sort of be wrong in your own head? Enough to hear out an opinion of somebody else when they disagree with you and learn from them? Those kinds of things.

你有感觉到金钱、权力和名声让你更难在自己脑子里承认"我错了"吗?难到听不进别人跟你不同的意见、难到没法从他们身上学东西?这类事情。

Surprisingly, no. And I would actually go the other way. Because I do so much of my work publicly, when I'm wrong, pretty much everybody sees it.

出人意料地,没有。我甚至会说是反方向的。因为我很大一部分工作是公开做的,我错的时候,基本上所有人都看得见。

You get humbled. Fair enough.

你被打回原形了。有道理。

And when I'm wrong—when I'm wrong or it didn't turn out that way or, you know, I mean, most of the things that I say outside I'm fairly certain about.

而当我错了——我错了,或者事情没往那个方向走……我在外面说的大部分事情,我是相当有把握的。

And the reason for that is because it's gonna impact somebody else and I want to be quite concerned about that and quite circumspect about that.

原因是它会影响到别人,所以我想对此格外在意、格外谨慎。

For stuff that I'm reasoning about inside a meeting, you know, a lot of things could turn out differently. And so, but it doesn't ever stop me from reasoning.

至于我在会议里推演的那些东西,很多事的结果可能完全不同。但这从来不会让我停止推演。

The way that I manage and lead, I'm constantly reasoning in front of people. And even when I'm talking to you, you can kind of see me reasoning through things.

我管理和领导的方式,就是不断在人前推演。哪怕我现在跟你说话,你也能看到我在一步步推。

And I want to make sure that you understand what I'm saying not because I told you-

而我想确保你理解我说的话,不是因为"是我告诉你的"——

… because I'm so humble about what I'm about to tell you. I kind of show you the steps that I got there. And then you can decide whether you believe what I said in the end.

——而是因为我对自己要告诉你的东西非常谦卑。我会把我走到那儿的步骤展示给你,然后你自己决定要不要相信我最后说的结论。

And so I'm doing that all day long in meetings. With all of my employees, I'm constantly reasoning through, "Let me tell you how I see it." And then I reason through it.

所以我整天在会议里都在做这件事。跟我所有员工,我都在不断推演:"我先说说我怎么看这件事。"然后我把推理走一遍。

It gives everybody the opportunity to intercept and say, "I disagree with that part."

这就给了所有人机会插进来说:"我不同意那一段。"

The nice thing about reasoning through things and letting people interact with it is that they don't have to disagree with your outcome. They can disagree with your reasoning steps.

把推理过程摊开、让人参与进来,好处是他们不必反对你的结论——他们可以反对你的推理步骤。

And they could pull me in different directions, and then we can reason forward. And so we're kind of, you know, a collective path searching method. And it's really fantastic.

他们可以把我往不同的方向拉,然后我们再一起往前推。所以我们有点像是一种集体的路径搜索方法。这真的很棒。

Yeah, you have this way about you of … When you're explaining stuff, I can feel you actually reasoning on the spot about it with a constant open-mindedness where you could … I could feel like I could steer your thinking.

对,你身上有那种气质——当你解释东西的时候,我能感觉到你真的在当场推演,而且始终保持开放。我甚至觉得自己能引导你的思路。

And that's a—that's really beautiful that you've been able to maintain that after so many years of success, and pain. I think sometimes pain closes you down a bit. And I think to maintain-

在这么多年的成功和痛苦之后你还能保持这一点,真的很美。我觉得痛苦有时候会让人稍微关闭起来。而要保持——

Yeah. Tolerance for embarrassment, I think is…

对。我觉得是对尴尬的耐受度……

Yes, that's… The tolerance… I mean, that's a real thing. Is many years of embarrassing yourself.

是的,那个耐受度……这是真实存在的东西。是很多年不断让自己出丑。

Even those meetings knowing that there's people around you where you declared one idea and it was shown that that idea was wrong- … and be able to admit that and to grow from that. That's not—that's very difficult on a human level.

哪怕在那些会议上,你知道身边有一群人,你宣布了一个想法,然后它被证明是错的——而你还能承认它、并从中成长。这在人性层面上是非常难的。

Yeah. Well, you know. They knew I was—they knew that recently my first job was cleaning toilets, so.

对。不过他们知道我——他们知道,不久前我的第一份工作是洗厕所,所以。

I'm glad you maintained that same spirit of Denny's, the work. I mean, that was beautiful. Your whole journey starting from Denny's is a beautiful one.

我很高兴你一直保持着 Denny's 时期那种对工作的心态。那真的很美。你从 Denny's 起步的整段旅程,本身就很美。

Let me ask you about video games. So I'm a big gaming fan. So I have to say thank you to NVIDIA for many years of incredible graphics.

我想问问电子游戏。我是个重度游戏迷,所以我得感谢 NVIDIA 这么多年带来的惊人画面。

By the way, GeForce is our still, to this day- … our number one marketing strategy.

顺便说,GeForce 到今天依然是我们的第一营销策略。

Right. People learn about NVIDIA while they're in their teenage years. And then they go to college and they know who NVIDIA is and in the beginning it's just, you know, playing Call of Duty, Fortnite.

人们在十几岁的时候就认识了 NVIDIA。然后他们上大学,已经知道 NVIDIA 是谁了——一开始只是玩《使命召唤》、《堡垒之夜》。

And then later they're using CUDA, and then later they're using NVIDIA and, you know, Blender and Dassault and Autodesk.

后来他们开始用 CUDA,再后来在 Blender、Dassault、Autodesk 里用 NVIDIA。

Yeah. I mean, I should say I mentioned to a friend that I'm talking with you. He said, "Oh, they make great gaming GPUs."

对。我得说,我跟一个朋友提到我要跟你聊,他说:"哦,他们做很好的游戏 GPU。"

Yeah, exactly.

对,正是。

It's like-

就像是——

Exactly.

正是。

You know, there's more to it, but, yeah, people really love it. It really brought a lot of joy to a lot of people. The hardware really brings these worlds to life.

当然不只是这样,但大家确实很爱它。它真的给很多人带来了快乐。硬件真的把这些世界变活了。

There was some controversy around this with DLSS 5. Can you explain to me the drama around this? I guess people, the gamers online were concerned that it makes games look like AI slop. What do you think of this drama?

关于 DLSS 5 有一些争议。你能给我讲讲这场风波吗?我理解是网上的玩家担心它会让游戏看起来像 AI slop。你怎么看这场风波?

Yeah. I think their perspective makes sense and I could see where they're coming from, because I don't love AI slop myself.

我觉得他们的视角有道理,我能理解他们为什么这么想,因为我自己也不喜欢 AI slop。

You know, all of the AI-generated content increasingly looks similar and they're all beautiful, and so I'm empathetic towards what they're thinking.

所有 AI 生成的内容越来越像,而且个个都很漂亮,所以我对他们的想法是有共情的。

That's just not what DLSS 5 is trying to do. I showed several examples of it. But DLSS 5 is 3D-conditioned, 3D-guided. It's ground truth structure data guided.

但 DLSS 5 想做的根本不是那件事。我展示过好几个例子。DLSS 5 是以 3D 为条件、以 3D 为引导的,是被 ground truth 的结构数据引导的。

And so the artist determined the geometry. We are completely truthful to the geometry maintained in every single frame. It's conditioned by the textures, the artistry of the artist.

所以几何形状是由艺术家决定的。我们在每一帧里都完全忠实于被维持的那个几何。它以贴图、以艺术家的创作为条件。

And so every single frame, it enhances but it doesn't change anything.

所以每一帧,它是在增强,而不是改变任何东西。

Now, the question is about enhancing. DLSS 5 also lets, because the system is open, you could train your own models to determine, and you could even in the future prompt it.

那么"增强"这件事怎么说。因为这个系统是开放的,DLSS 5 还允许你训练自己的模型去决定效果,未来你甚至可以对它下 prompt。

You know, I want it to be a toon shader. I want it to look like this kind of, so you can give it even an example.

比如我想要一个卡通渲染器,我想让它看起来像某种风格——你甚至可以给它一个样例。

And it would generate in the style of that, all consistent with the artistry, the style, the intent of the artist.

然后它就会按那个风格生成,并且全部与艺术家的创作、风格、意图保持一致。

And so all of that is done for the artist, so that they can create something that is more beautiful but still in the style that they want.

所有这些都是为艺术家做的,好让他们能创造出更美、但仍然是他们想要的那种风格的东西。

I think that they got the impression that the games are gonna come out the way the games are, shipped the way they do, and then we're gonna post-process it. That's not what DLSS is intended to do.

我觉得他们的印象是:游戏照原样做出来、照原样发布,然后我们去做后处理。那不是 DLSS 想做的事。

DLSS is integrated with the artist, and so it's about giving the artist the tool of AI, the tool of generative AI. They could decide not to use it, you know?

DLSS 是和艺术家整合在一起的,它是把 AI 这个工具、生成式 AI 这个工具交给艺术家。他们也可以决定不用它。

I think people are very sensitive to human faces. And we're now living in this moment, which I think is a, is a beautiful one, which is people are sensitive to AI slop.

我觉得人对人脸非常敏感。而我们现在正处在一个我认为很美的时刻:人们对 AI slop 变得敏感。

It puts a mirror to ourselves to help us realize that what we seek is imperfections. What we seek is sometimes not perfect graphics.

它像一面镜子照向我们自己,帮我们意识到:我们追求的其实是不完美。我们追求的有时候并不是完美的画面。

It helps us understand what we find compelling in the worlds we create. And that's beautiful. And as long as it's tools that help us create those worlds-

它帮我们理解:在我们创造的那些世界里,究竟什么才是打动我们的东西。这很美。而只要它是帮我们创造那些世界的工具——

Yeah, that's right.

对,没错。

… it's wonderful.

……那就太好了。

That's right. Yet, yet another tool, and they want the generative models to generate the opposite of photo real. Yeah, it'll do that too. And so it's just yet another tool.

没错。又是一个工具而已。而如果他们想让生成模型生成"照片真实"的反面——对,它也能做到。所以它就是又一个工具。

I think the gamers might also appreciate that in the last couple of years, we introduced skin shaders to the game developers.

我觉得玩家可能也会欣赏这一点:过去这几年我们给游戏开发者提供了皮肤着色器。

And many of those games have skin shaders that include subsurface scattering that make skin look more skin-like.

很多游戏里的皮肤着色器包含次表面散射,让皮肤看起来更像皮肤。

And so the industries, you know, game developers are looking for more and more tools to express their art. And so this is just yet one more tool, and they get to decide what to use.

所以这个行业、游戏开发者们,一直在寻找越来越多的工具来表达他们的艺术。所以这只是又多了一个工具,用不用由他们决定。

Ridiculous question. What do you think is the greatest or most influential game ever made? Maybe from NVIDIA's perspective?

一个无理的问题:你觉得史上最伟大、或者最有影响力的游戏是哪一款?也许从 NVIDIA 的角度看?

Doom.

《Doom》。

Doom, unquestionably. That was the start of the 3D.

《Doom》,毫无疑问。那是 3D 的起点。

I would say Doom, from an art, the intersection of the cultural implication as well as the industry, turning a PC into a gaming device. That was a very important moment.

我会说是《Doom》——从艺术、文化影响与产业的交汇点来看,它把 PC 变成了一台游戏设备。那是个非常重要的时刻。

Now, of course, flight simulation companies were before it. And but they just didn't have the popularity that Doom did to have made the industry turn the PC from an office automation tool into a personal computer for families and gamers and things like that.

当然,飞行模拟类的公司在它之前就有了。但它们没有《Doom》那种普及度,没能让整个行业把 PC 从办公自动化工具变成给家庭、给玩家用的个人计算机。

And so Doom was really impactful there. From an actual game technology perspective, I would say Virtua Fighter. And so we're great friends with both of them, you know?

所以《Doom》在这一点上影响非常大。从真正的游戏技术角度看,我会说是《VR 战士》(Virtua Fighter)。而我们跟这两边都是好朋友。

And then there's games more recently—I mean, Cyberpunk 2077, really nice GPU-accelerated graphics. Like-

还有比较近的游戏——比如《赛博朋克 2077》,GPU 加速的画面非常棒。像——

Fully ray traced.

全光线追踪。

Fully ray traced. Also, I like, I personally, I'm a huge fan of Skyrim, Elder Scrolls, and the, you know, it's, it's been released a long, long time ago, but people release mods and-

全光线追踪。另外我个人非常喜欢《天际》、《上古卷轴》。它发布已经很久很久了,但人们会发布 mod,然后——

We love mods.

我们爱 mod。

… they create these inc- I mean, it's like a different game and it just allows me to replay the game over and over.

……他们做出那些不可——简直就像换了一个游戏,让我可以一遍又一遍重玩。

It makes you realize that you can re-experience in a totally new way the world you already love. So-

它让你意识到:你可以用一种全新的方式,重新体验你本来就已经爱的那个世界。

That's right.

没错。

… I do that all the time. One of my favorite things is just walk across Skyrim.

……我一直这么干。我最喜欢的事情之一,就是在《天际》里一路走过去。

We created this thing called RTX Mod. Yeah, it's a modding tool.

我们做了个叫 RTX Mod 的东西,是个 mod 工具。

Awesome.

太棒了。

It allows the community to inject the latest technology into an old game.

它让社区可以把最新的技术注入到一个老游戏里。

Of course, like what makes a great video game is not just graphics, it's also story and character development, but-

当然,一款好游戏靠的不只是画面,还有故事和角色塑造,但——

That's right

没错。

… beautiful graphics can add to the immersion. The feeling like it's another place you're transported to.

……但漂亮的画面能加强沉浸感,那种"你被传送到另一个地方"的感觉。

Ah, what you said, I think accurately, that the AGI timeline question rests on your definition of AGI. So let's, let me ask you about possible timelines here.

你说过一句我觉得很准确的话:AGI 时间表这个问题,取决于你对 AGI 的定义。那我们来聊聊可能的时间表。

Let's, this ridiculous definition perhaps of what AGI is, but an AI system that's able to essentially do your job. So, run, no, start, grow, and run a successful technology company that's worth-

用一个也许有点无理的 AGI 定义:一个基本上能做你这份工作的 AI 系统。也就是能运营——不,是从零创办、发展并运营一家成功的科技公司,估值达到——

A good one or a one?

一家好公司,还是随便一家?

No. It has to be worth more than a billion, more than a billion dollars. So, you know, you know how hard it is to do all those components. So, how far are we away from that?

不。它得值超过十亿、超过十亿美元。你知道要做到这些环节有多难。那我们离那一步还有多远?

So, we're talking about Open-Claude that does all the incredibly complex stuff that are required to, first of all, innovate, to find customers, to sell to them, to manage, to build a team of some agents, some humans, all that kind of stuff. Is this five, 10, 15, 20 years away?

我们说的是一个 Open-Claude,它要做完所有那些极其复杂的事:首先要创新,要找到客户、卖给他们、做管理、组建一个由一些 agent 和一些人类组成的团队,所有这类事情。这是五年、十年、十五年,还是二十年之后的事?

I think it's now. I think we've achieved AGI.

我觉得就是现在。我认为我们已经实现 AGI 了。

Do you think you could have a company run by an AI system like this?

你觉得真能有一家公司由这样的 AI 系统来运营?

Possible, and the reason for that is this. You said a billion, and you didn't say forever.

有可能,原因是这样:你说了十亿,但你没说"永远"。

And so for example… It is not out of the question that a Claude was able to create a web service, some interesting little app that all of a sudden, you know, a few billion people used for 50 cents, and then it went out of business again shortly after.

举个例子:完全不能排除这种情况——某个 Claude 做出一个网络服务、一个有意思的小应用,突然有几十亿人花五十美分用它,然后不久之后它又倒闭了。

Now, we saw a whole bunch of those type of companies during the internet era, and most of those websites were not anything more sophisticated than what Open-Claude could generate today.

互联网时代我们见过一大堆这类公司,而那些网站里大部分的复杂度,并不比今天 Open-Claude 能生成的东西更高。

Interesting. Achieve virality and monetize that virality.

有意思。做出病毒式传播,再把这种传播变现。

Yeah. It's just that I don't know what it is, but I couldn't have predicted any of those companies at the time either, you know? And –

对。只是我不知道那会是什么——但当年那些公司我也一个都预测不出来。而且——

You're gonna get a lot of people excited with that statement.

这句话会让很多人兴奋起来。

Yeah, no. Yeah.

是啊。

It's like, what do you mean? I can just launch an agent and make a lot of money.

大家会说:什么意思?我只要放一个 agent 出去,就能赚一大笔钱?

Well, by the way, it's happening right now, right?

顺便说,这件事现在就在发生了,对吧?

You know that when you go to China you're gonna see, you're gonna see a whole bunch of people teaching their, getting their Claudes to try to go out and look for jobs and, you know, do work, make money.

你去中国就会看到,一大堆人在教自己的 Claude、让它们出去找活干、干活、赚钱。

And I'm not, I'm not actually… I wouldn't be surprised if some social thing happened or somebody created a digital influencer, super, super cute, or some social application that, you know, feeds your little Tamagotchi or something like that, and it become out of the blue an instant success.

而我一点都不会意外:哪天出现某个社交类的东西,或者有人做了一个超级超级可爱的数字网红,或者某个社交应用——比如喂你的电子宠物之类的——然后它莫名其妙就一夜爆红。

A lot of people use it for a couple of months and it kind of dies away. Now, the odds of 100,000 of those agents building NVIDIA is zero percent.

很多人用了几个月,然后它就慢慢消失了。但要说十万个这样的 agent 能建出一个 NVIDIA,概率是零。

And then, and then the one part that I will, I won't do and I wanna make sure we all do, is to recognize that people are really worried about their jobs.

还有一件我不会回避、而且希望我们所有人都做到的事,是要认识到:人们真的很担心自己的工作。

And I just want to remind them that the purpose of your job and the tasks and tools that you use to do your job are related, not the same.

我只想提醒他们:你工作的目的,和你为完成工作而使用的任务与工具,是相关的,但不是同一件事。

I've been doing my job for 33 years. I'm the longest running tech CEO in the world, 34 years.

我做这份工作已经 33 年了。我是全世界在任时间最长的科技 CEO,34 年。

And the tools that I've used to do my job has changed continuously in the last 34 years, and sometimes quite dramatically, you know, over the course of a couple, two, three years.

而过去 34 年里,我用来做这份工作的工具一直在变,有时候在两三年里变得相当剧烈。

And the one story that I really wanna make sure that everybody hears is the story that the first job that computer scientists said, AI researchers said was gonna go away was radiology.

我特别想让所有人都听到的一个故事是:计算机科学家、AI 研究者说第一个会消失的职业,是放射科。

Because computer vision was going to achieve superhuman levels, and it did. CV… Computer vision was superhuman in 2019, 20, maybe maybe a little bit later, 2020?

因为计算机视觉将会达到超越人类的水平——而它确实做到了。计算机视觉在 2019、2020 年就超越人类了,或许再晚一点,2020?

Okay? And so it's been a long time since computer vision has been superhuman. And so the prediction was radiologists would go away because studying radiology scans was a thing of the past. AI will do that.

所以计算机视觉超越人类已经很久了。当时的预测是:放射科医生会消失,因为看放射影像会成为过去式,AI 会来做这件事。

Well, they were absolutely right. Computer vision is completely superhuman. Every radiology platform and package today is driven by AI, and yet the number of radiologists grew.

他们说得完全没错——计算机视觉彻底超越了人类。今天每一个放射影像平台和软件包都是由 AI 驱动的。然而放射科医生的数量增长了。

And so the question is why? And we now have a shortage of radiologists in the world.

所以问题是:为什么?而现在全世界还闹放射科医生短缺。

And so, one, the alarmist warning went too far and it scared people from doing this profession that is so important to society. And so it did harm.

第一,那个警报式的预警走得太远了,它把人们从这个对社会如此重要的职业里吓跑了。所以它造成了伤害。

Now, why was it wrong? The reason why is because the purpose of a radiologist, the purpose is to diagnose disease and help patients and doctors diagnose disease.

那它为什么错了?原因是:放射科医生的目的,是诊断疾病、帮助病人和医生诊断疾病。

And because we're able to study scans so much faster now, you could study more scans, you could diagnose better, you could in-patient faster, you can see people more.

而因为我们现在读影像快得多,你就能读更多影像、诊断得更好、更快收治病人、看更多的人。

The hospitals are making more money. You have more patients in the hospital. You need more radiologists.

医院赚更多钱,医院里病人更多,你就需要更多放射科医生。

I mean, the amazing thing is, it's so obvious this was gonna happen. The number of software engineers at NVIDIA is gonna grow, not decline.

最神奇的是,这件事本来就非常显然。NVIDIA 的软件工程师数量会增长,不会下降。

And the reason for that is because the purpose of a software engineer and the task of a software engineer coding are related, not the same.

原因是:软件工程师的目的,和软件工程师写代码这个任务,是相关的,但不是同一件事。

I wanted my software engineers to solve problems. I didn't care how many lines of code they wrote, you know? But their job, their purpose of their job didn't change.

我要的是我的软件工程师解决问题。我不在乎他们写了多少行代码。但他们工作的目的没有变。

Solving problems, working as a team, diagnosing problems, evaluating the result, looking for new problems to solve, innovation, connecting dots. You know, none of that stuff is gonna go away.

解决问题、作为团队协作、诊断问题、评估结果、寻找新的问题去解决、创新、把点连起来。这些东西一样都不会消失。

Do you think it's possible that… Let's even take coding. Do you think the number of programmers in the world might increase, not decrease?

你觉得有可能……我们就拿写代码来说。你觉得世界上程序员的数量可能是增加,而不是减少?

Yes. And the reason for that is this. What is the definition of coding?

是的。原因是这样:什么叫写代码?

I believe it is… The definition of coding, as of today, is simply specifying, specification, and maybe if you want to be rather directive, you could even give it an architecture of the software that you wanted to write.

我认为到今天为止,写代码的定义就是"写规格"——说明清楚你要什么;如果你想更有指导性,你甚至可以给它一个你想写的软件的架构。

So the question is, how many people could do that? Describe a specification for a computer to go… telling the computer what to go build. How many people?

那么问题是:有多少人能做这件事?为一台计算机描述一份规格,告诉计算机去造什么。有多少人能做?

I think we just went from 30 million to probably 1 billion.

我认为我们刚刚从 3000 万人变成了大概 10 亿人。

And so every carpenter in the future will be a coder, except a carpenter with AI is also an architect. They've just increased the value that they could deliver to the customer. Their artistry just elevated tremendously.

所以未来每个木匠都会是一个写代码的人——只不过一个带着 AI 的木匠同时还是一个建筑师。他们能交付给客户的价值刚刚提高了,手艺水准大幅跃升。

I believe that every accountant is, you know, also your financial analyst, also your financial advisor. So, all of these professions have just been elevated…

我相信每个会计同时也是你的财务分析师,也是你的财务顾问。所以所有这些职业刚刚都被抬高了一档。

and if I were a carpenter, I see AI, I would just completely go berserk. You know, the services I can bring to my clients if I were a plumber, completely go berserk.

如果我是个木匠,我看到 AI,我会彻底玩疯。如果我是个水管工,想想我能带给客户的服务——彻底玩疯。

And the, the people that are currently programmers and software engineers, I think they're at the cutting edge of understanding intuitively how to communicate with the agents using natural language in order to design the best kind of software.

而那些现在是程序员和软件工程师的人,我觉得他们站在最前沿——他们在直觉层面上最懂怎么用自然语言跟 agent 沟通,来设计出最好的软件。

That's right, exactly.

对,正是。

So over time they'll converge, but I think there's still value in getting, I think learning how to program, like learning what programming languages are.

所以随着时间它们会收敛,但我觉得学会编程仍然有价值——比如了解什么是编程语言。

The old kind of programming, what are good practices for programming languages, what are design principles for programming-

老式的那种编程:编程语言有哪些好实践,有哪些设计原则——

That's right

没错。

… Languages for large software systems?

……用于大型软件系统的语言?

And the reason for that, Lex, and you know, as you're saying for the audience, I think the goal of, the goal of specification, the artistry of specification, the goal and the artistry of it is going to depend on what problem you're trying to solve.

而原因是,Lex——就像你在跟观众说的——我认为写规格这件事的目标和它的手艺,取决于你想解决什么问题。

When I'm thinking, when I'm thinking about giving the company strategies and formulating corporate directions and things that we should do, I describe it at a level that is sufficiently specific that people generally understand the direction and it's actionable.

当我在想公司战略、在制定公司方向、在决定我们该做什么的时候,我会把它描述到一个足够具体的程度:大家大致理解方向,而且这个方向是可执行的。

It's specific enough that they can take action on it, but I under-specify it on purpose, so that enables 43,000 amazing people to make it even better than I imagined.

具体到他们能据此行动,但我又故意写得不够详尽——好让四万三千个了不起的人把它做得比我想象的更好。

And so when I'm working with engineers and when I'm working with people, I think about who, what problem am I trying to solve? Who am I working with? And the level of specification, the level of architecture definition relates to that.

所以当我跟工程师、跟人一起工作时,我会想:我要解决的是什么问题?我在跟谁一起工作?而规格的详尽程度、架构定义的详尽程度,跟这两点相关。

And so everybody's going to have to learn how, where in the spectrum of coding they want to be. Writing a specification is coding.

所以每个人都得学会:在"写代码"这条谱系上,自己想站在哪个位置。写规格就是写代码。

And so you might decide to be quite prescriptive because there's a very specific outcome you're looking for.

你可能决定写得非常规定性,因为你要的是一个非常具体的结果。

You might decide that, you know, this is an area you want to be much more exploratory, and so you might under-specify and enable you to go back and forth with the AI to even push your own boundaries of creativity.

你也可能觉得这是个你想更多探索的领域,于是你故意写得不那么详尽,好让你能和 AI 来回拉扯,甚至把自己创造力的边界往外推。

And so this artistry of where you are in the spectrum, this is the future of coding.

所以"你在这条谱系上站在哪里"这门手艺,就是写代码的未来。

But just to linger on it outside of coding, I think a lot of people, rightfully so, are worried about their jobs, have a lot of anxiety about their jobs, especially in the white-collar sector.

但我想在写代码之外多停一下:很多人——也有道理——在担心自己的工作,对工作有很多焦虑,尤其是白领这一块。

I don't think any of us know what to do with tumultuous times that always come when automations and new technology arrives.

我觉得每次自动化和新技术到来时那种动荡期,我们谁都不知道该怎么应对。

And I just… First of all, I think we all need to have compassion and the responsibility to feel sort of the burden of what the actual suffering feels like for individual people and families that lose their job.

首先,我认为我们都需要有慈悲心,也有责任去感受那份重量:对那些失去工作的个人和家庭来说,真实的痛苦是什么感觉。

I think whenever you have transformative technology like that's coming with artificial intelligence, there's going to be a lot of pain, and I don't know what to do about that pain.

我觉得每当有人工智能这种变革性技术到来,一定会有大量的痛苦,而我不知道该怎么处理那份痛苦。

Hopefully, it creates much more opportunities for those same people for the same kind of job as the tooling evolves and makes them more productive and makes them more fun, hopefully, as it does in the programming.

希望随着工具演进,它能为这些人、在同类工作上创造出更多机会,让他们更有生产力、也更快乐——就像在编程里那样。

I have been having so much fun programming, I have to say. Like, I've never had this much fun.

我必须说,我最近写代码写得特别开心,从来没这么开心过。

So hopefully it makes their job, automates the boring parts and makes the creative parts the ones that the human beings are responsible for. But still there's going to be a lot of pain and suffering.

所以希望它能把工作里枯燥的部分自动化掉,把创造性的部分留给人类负责。但仍然会有大量的痛苦和折磨。

So my first recommendation before… And this is now how I deal with anxiety. In fact, we just talked about it earlier.

我的第一个建议……这也是我现在应对焦虑的方式,其实我们刚才聊过。

Enormous anxiety about the future, enormous anxiety about the pressure, enormous anxiety about uncertainty, I first break it down, and then I'm gonna tell myself, "Okay, there are some things you can do something about, there's some things you can't do anything about. But for the stuff that you can do something about, let's reason, reason about it and let's go do it."

对未来的巨大焦虑、对压力的巨大焦虑、对不确定性的巨大焦虑——我先把它拆开,然后告诉自己:"好,有些事你能做点什么,有些事你什么都做不了。但对于你能做点什么的那些,我们来推演一遍,然后去做。"

If we were to hire a new college graduate today, and I have a choice between two, one that has no clue what AI is and one that is expert in using AI, I would hire the one who's expert in using AI.

如果我们今天要招一个应届毕业生,我在两个人之间选:一个完全不知道 AI 是什么,一个是用 AI 的专家——我会招那个用 AI 的专家。

If I had an accountant, a marketing person, the one that is expert in using AI, supply chain, customer service, a salesperson, business development, a lawyer, I would hire the one who is expert in using AI.

如果是会计、市场、供应链、客服、销售、商务拓展、律师——我都会招那个用 AI 的专家。

And so I would advise that every college student, every teacher should encourage their student to go use AI. Every college student should graduate and be an expert in AI.

所以我会建议每一个大学生、每一位老师都鼓励自己的学生去用 AI。每个大学生毕业时都应该是用 AI 的专家。

And everybody, if you're a carpenter, if you're an electrician, go use AI. Go see what it can do to transform your current job, elevate yourself.

所有人——如果你是木匠、如果你是电工,去用 AI。去看看它能怎么改造你现在的工作,把你自己抬高一档。

If I were a farmer, I would absolutely use AI. If I were a pharmacist, I would use AI.

如果我是农民,我绝对会用 AI。如果我是药剂师,我会用 AI。

I wanna see how, what it could do to elevate my job so that I could be the innovator to revolutionize this industry myself. And so that would be the first thing that I would do.

我想看看它能怎么把我的工作抬高一档,好让我自己成为那个革新这个行业的创新者。这是我会做的第一件事。

And then I would also help them… It is the case that the technology will dislocate and will eliminate many tasks.

然后我也会帮他们……事实确实是:这项技术会造成错位,会消灭掉很多任务。

And because it will automate it, if your job is the task—then you're very highly going to be disrupted.

因为它会把那些任务自动化——所以如果你的工作就等于那个任务,那你被颠覆的概率极高。

If your job's purpose includes you, certain tasks- … then it's vital that you go learn how to use AI to automate those tasks. And then there's the world of spectrum in between.

如果你工作的目的里包含你本人、再加上某些任务,那你就非常有必要去学会用 AI 把那些任务自动化。而中间还有一整片光谱。

And by the way, the beautiful thing about AI, so the chatbot versions, is you can break down… You have anxiety and you can break down the problem by talking to it.

顺便说,AI——尤其是聊天机器人那种形态——一个很美的地方是:你有焦虑,而你可以通过跟它对话把问题拆开。

Like, I've recently… It's really just incredible how much you can think through your life's problems, and through… And I don't mean, like, therapy problems.

我最近……你能借它把自己人生的问题想清楚到什么程度,真的很惊人。我不是说心理治疗那类问题。

I mean, like, very practically, "Okay, I'm worried about my…" Literally, "I'm worried about my job. What are the skills? What are the steps I need to take?" How do I get better at AI?"

我是说非常实际的:"好,我担心我的……"就是字面意义上的"我担心我的工作。需要哪些技能?我该走哪些步骤?我怎么把 AI 用得更好?"

Everything you just said, you could literally ask and it's going to give you- … a point-by-point plan. I mean, it's just a great life coach, period. This-

你刚才说的每一件事,你都可以真的去问它,而它会给你一份一条一条的计划。它就是个很棒的人生教练,没别的。

I don't know how to use AI, and the AI goes, "Well, let me show you."

"我不知道怎么用 AI"——然后 AI 说:"来,我教你。"

Exactly. It's very meta, but it's- It's kind of incredible. So people definitely should-

正是。这非常"元",但也挺不可思议的。所以大家真的应该——

You can't walk up to Excel and say, "I don't know how to use Excel."

你没法走到 Excel 面前说:"我不知道怎么用 Excel。"

Exactly.

正是。

You're done.

那你就没戏了。

I mean, that's really what AI has done for me in all walks of life, is that initial friction of being a beginner of using a thing for the first time.

这正是 AI 在我生活各个方面为我做到的事:它抹掉了作为新手第一次用一样东西时的那种起始摩擦。

I can literally ask about any single thing, "What are the first steps I need to take?"

我可以对任何一件事直接问:"我需要走的第一步是什么?"

That's right.

没错。

And that handholding that it does, removing the friction of all the experiences that the world offers is… You know, like I mentioned to you offline, you mentioned, "I'm going to China and Taiwan."

它这种手把手带你的方式,把世界上所有体验的摩擦都抹掉了……就像我在录制外跟你提到的,你说"我要去中国和台湾"。

So awesome.

太棒了。

Just ask, "Where do I-"

直接问就行:"我该去哪——"

So excited for you.

真替你高兴。

"Where do I… What do…" "You know, where do I go? How do I…" All of those questions- … immediately answered, and it's beautiful.

"我该去哪……我该做什么……我该怎么……"所有这些问题都立刻有了答案,这很美。

Well, when you go to Taiwan, just ask AI… "What are Jensen's favorite restaurants in Taiwan?" And it'll actually-

等你去台湾的时候,直接问 AI:"黄仁勋在台湾最喜欢的餐厅是哪几家?"它真的会——

You don't know?

你不知道?

Oh, yeah.

哦,知道。

Is it accurate? Okay. All right.

它准吗?好,行。

It's all over Taiwan.

全台湾都传遍了。

Well, you're a rockstar over there. And like we also mentioned offline, maybe our paths will cross, which would be really wonderful in computing.

你在那儿是摇滚明星。而且像我们在录制外提到的,也许我们的路会在计算这件事上交汇,那会非常好。

COMPUTEX. NVIDIA GTC Taiwan.

COMPUTEX。NVIDIA GTC 台湾站。

Do you think there's some things about human nature, about human consciousness that is fundamentally non-computational? Maybe something a chip, no matter how powerful, can never replicate?

你觉得人性、人的意识里,有些东西是根本不可计算的吗?也许是某种再强的芯片都永远无法复制的东西?

I don't know if the chip will ever get nervous. And that's the, you know, of course, the conditions by which that causes anxiety or nervousness or whatever emotion.

我不知道芯片有没有一天会紧张。当然,还有导致焦虑、紧张或任何情绪的那些条件。

I believe that AI will be able to recognize those and understand those. I don't think my chips will feel those.

我相信 AI 能识别这些、理解这些。但我不认为我的芯片会"感受到"它们。

And therefore, the… How that anxiety, how that feeling, how that excitement, how that, how that, you know… All of those feelings manifest in human performance.

而那种焦虑、那种感觉、那种兴奋……所有这些情绪,会体现在人的表现上。

For example, extremely amazing human performance, athletic performance, you know, average or lesser than average.

比如极其惊人的人类表现、运动表现,或者平庸、低于平庸的表现。

That entire spectrum of human performance that comes out of exactly the same circumstances for different people, manifesting a different outcome, manifesting a different performance.

同样一模一样的处境,放在不同的人身上,会呈现出完全不同的结果、不同的表现——这一整条人类表现的光谱。

I don't think there's anything about anything that we're building that would suggest that two different computers being presented with all of exactly the same context would perfo-

我不觉得我们正在造的任何东西里,有任何迹象表明:两台不同的计算机,在完全相同的上下文下会表现得——

Of course, it would produce statistically different outcomes, but it's not because it felt different.

当然,它会产生统计意义上不同的结果,但那不是因为它"感受"不同。

Yeah, the subjective… Boy, there's something truly special about the subjective experience that we humans feel. Like I mentioned to you, I was pretty nervous talking to you.

对,那种主观的……天哪,我们人类感受到的那种主观体验里,确实有某种非常特别的东西。就像我跟你说的,我跟你聊之前挺紧张的。

Like I mentioned to you, that, the hope, the fear, the anxiety, and just life itself, the richness of life. How amazing everything is.

那种希望、恐惧、焦虑,以及生命本身、生命的丰盈。一切有多惊人。

How deeply we fall in love, how deeply our hearts get broken, how afraid we are of death and how much pain we feel when our loved ones pass away. All of that, the whole thing.

我们爱得有多深,心碎得有多深,我们对死亡有多恐惧,所爱之人离世时我们有多痛。所有这些,整整一整套。

I know it's very hard to- … think AI being able to… A computational device being able to do that. But there's so many mysteries about this whole thing that we're yet to uncover, that I am open to be surprised. I've been surprised a lot over the past-

我知道很难想象 AI 能做到、一个计算设备能做到这些。但这整件事里还有太多我们尚未揭开的谜团,所以我对被惊到是开放的。过去这段时间我已经被惊到很多次了——

… few months and few years. Scaling can create some incredible miracles in the space of intelligence. It has been truly marvelous to watch, so I'm open to surprise.

……过去几个月、几年。扩展能在智能这个空间里造出一些不可思议的奇迹。看着这一切真的令人惊叹,所以我对惊喜是开放的。

And it's just really important to break down what is intelligence. You know, the word, that word we use all the time, it's not a mysterious word. Intelligence has a meaning, you know?

而非常重要的一件事是把"智能"这个词拆开。这个我们天天用的词,它并不神秘。智能是有明确含义的。

And it's a system that… You know, it's something that we do that includes perception and understanding and reasoning and the ability to do plan.

它是一个系统……是我们在做的一件事,包含感知、理解、推理,以及做规划的能力。

And, you know, that loop, that loop, is the… Fundamentally what intelligence is.

而那个循环、那个 loop,从根本上就是智能。

Intelligence is not one word that is exactly equal to humanity. And that's, I think it's really important to separate the two. We have two words for that.

智能不是一个可以直接等同于"人性"的词。我觉得把这两者分开非常重要——我们有两个不同的词来说它们。

I'm not… I don't over-fantasize about, and I don't over-romanticize about intelligence. Intelligence is… And people have heard me say it before, I actually think intelligence is a commodity.

我不会过度幻想智能,也不会过度浪漫化智能。大家以前听我说过:我其实认为智能是一种大宗商品。

I'm surrounded by intelligent people. And I'm surrounded by intelligent people more intelligent than I am in each one of the spaces that they're in.

我身边都是聪明人。而且在他们各自的领域里,他们都比我更聪明。

And yet, I have a role in that circle. It's actually kind of interesting.

然而我在那个圈子里有我的角色。这其实挺有意思的。

They're more educated than I am. They went to better schools than I did. They're deeper in any of the fields that they're in. All of them.

他们受的教育比我好,他们上的学校比我好,他们在各自领域里都比我钻得深。全都是。

I have 60 of them. They're all superhuman to me. And somehow, I'm sitting in the middle orchestrating all 60 of them.

我手下有 60 个这样的人。在我眼里他们都是超人。而不知怎么,我坐在中间调度着这 60 个人。

And so you gotta ask yourself… What is it about a dishwasher that allows that dishwasher to sit in the middle of superhumans? Does that make sense?

所以你得问自己:一个洗碗工凭什么能坐在一群超人中间?你明白我的意思吗?

And so, but that's my point. My point is intelligence is a functional thing. Humanity is not specified functionally. It's a much, much bigger word.

这就是我的意思:智能是一个功能性的东西。而"人性"不是按功能来定义的,它是一个大得多、大得多的词。

And our life experience, our tolerance for pain, our determination, those are different words than intelligence.

我们的人生经历、我们对痛苦的耐受度、我们的决心——这些跟"智能"是不同的词。

And so the thing that I wanna help the audience understand, if I could give them one thing, is intelligence is a word that we've elevated to a very high form over time.

所以如果我只能给观众一样东西,我想帮他们理解的是:"智能"是一个被我们长期以来抬得过高的词。

The word we should really elevate is humanity.

我们真正该抬高的词,是"人性"。

Character, humanity.

品格,人性。

All those things.

所有这些。

All of those things. Compassion, generosity, all of the things that you say just now, I believe those are superhuman powers.

所有这些。慈悲、慷慨,你刚才说的所有那些——我认为那些才是超人的力量。

And that now intelligence is gonna be commoditized. Because we've spoken about it, the most important thing is your education.

而现在智能要被商品化了。我们谈过这件事:最重要的是你的教育。

Now, even when they said the most important thing is your education, when you went to school, there's more than just knowledge that you gained.

而即便当人们说"最重要的是你的教育"时——你去上学,你得到的也远不只是知识。

And so, but unfortunately, our society had put everything into one single word, and life is more than one word.

但不幸的是,我们的社会把一切都塞进了一个词里,而人生远不止一个词。

And I'm just telling you, my life would suggest that being lower on the intelligence curve than everybody around me doesn't change the fact I'm the most successful.

我就这么告诉你:我自己的人生说明,在智能曲线上比周围所有人都低,并不改变"我是最成功的那个"这个事实。

And so, and I think that kind of is—I'm trying to hopefully to inspire everybody else—that don't let this democratization of intelligence, this commoditization of intelligence, cause you anxiety. You should be inspired by that.

我希望这能激励其他所有人:不要让智能的民主化、智能的商品化引起你的焦虑。你应该被它激励。

Yeah. I think AI will help us celebrate humans more. And certainly humanity and human first, and I think what makes this world incredible is humans forever will be so, and just AI is this incredible tool that makes us-

对。我觉得 AI 会帮我们更多地去赞美人。当然是人性优先、人优先。我觉得让这个世界了不起的是人,而且永远会是人。AI 只是这个不可思议的工具,让我们——

That's exactly right.

完全正确。

… humans more powerful.

……让人类更强大。

That's exactly right.

完全正确。

So much of the success of NVIDIA and the lives of millions of people that I mentioned depend on you.

NVIDIA 的成功、以及我提到的数百万人的生活,有很大一部分都依赖你。

But you're just one human, like we mentioned, a mortal like all of us. Do you think about your mortality? Are you afraid of death?

但你只是一个人,像我们说的,和我们所有人一样是凡人。你会想自己的死亡吗?你怕死吗?

I really don't wanna die. I have a great life. I have a great family. I have really important work.

我真的不想死。我有很好的人生,有很好的家庭,有非常重要的工作。

This is not a once in a lifetime experience suggests that it has been experienced by many people, just not one person. This is a once in a humanity experience, what I'm going through.

"一生一次"这种说法意味着很多人都经历过、只是不是同一个人——而这不是。我正在经历的,是"人类一次"的经历。

NVIDIA is one of the most consequential technology companies in history. We're doing very important work. I take it very seriously.

NVIDIA 是历史上最有影响力的科技公司之一。我们在做非常重要的工作,我非常认真地对待它。

And so some of the things that of course are practical things, like how do we think about succession planning? And I'm famous in saying that I don't believe in succession planning.

所以当然有一些实际的事情,比如:接班人规划我们该怎么想?而我有句有名的话:我不信接班人规划。

Man.

天。

And the reason for that isn't because I'm immortal.

理由不是因为我不朽。

The reason for that is because if you're worried about succession planning, if you're worried all that anxiety of succession planning, then what should you do about it? Then you break it all the way back down.

理由是:如果你担心接班人规划、被接班人规划的那些焦虑困扰,那你该拿它怎么办?你就把它一路拆回去。

The most important thing you should do today, if you care about the future of your company, post you, is to pass on knowledge, information, insight, skills, experience as often and continuously as you can, which is the reason why I continuously reason about everything in front of my team.

如果你在乎公司在你之后的未来,那你今天最该做的事,就是尽可能频繁、尽可能持续地把知识、信息、洞见、技能、经验传出去——这就是我为什么持续地在团队面前推演每一件事。

Every single meeting is a reasoning meeting. Every moment I spend inside a company, outside a company is about passing on knowledge to people as fast as I can.

每一场会都是推演会。我在公司里、公司外的每一刻,都是在尽可能快地把知识传给别人。

Nothing I learn ever sits on my desk longer than, you know, a fraction of a second.

我学到的任何东西,在我桌上停留的时间不会超过一瞬间。

I'm passing that information, that knowledge—oh my gosh, this is cool. Before I even finish learning all of it myself, I'm already pointing it to somebody else. "Get on this. This is so cool. You're gonna wanna learn this."

我会立刻把那份信息、那份知识传出去——"天哪,这个太酷了。"在我自己还没学完之前,我已经在把它指给别人看了:"上手这个。这太酷了,你会想学的。"

And so I'm constantly passing knowledge, empowering people, elevating the capability of everybody around me, so that the outcome that I seek, that I hope for, is that I die on the job, you know?

所以我一直在传递知识、赋能他人、抬高身边每一个人的能力——好让我追求、我希望的那个结果发生:我死在岗位上。

And hopefully I die on the job instantaneously, you know? And there's no long periods of suffering, you know? It's, uh –

而且最好是瞬间死在岗位上,不要有漫长的受苦期。

Well, from a fan perspective, given your extremely enormous positive impact on civilization, of course, I hope you keep going.

从一个粉丝的角度,考虑到你对文明的巨大正面影响,我当然希望你继续下去。

But also it's just fun to watch what NVIDIA is doing, you know. It's just the rate of innovation. And I'm a huge fan of engineering.

而且看 NVIDIA 在做的事本身就很有意思——就那个创新速度。我是工程的重度爱好者。

There's so much incredible engineering continuously being done by NVIDIA. It's just fun to watch. It's a celebration of humanity, a celebration of great builders, a celebration of great engineering.

NVIDIA 在持续做出这么多不可思议的工程,看着就很过瘾。它是对人性的一场庆祝、对伟大建造者的庆祝、对伟大工程的庆祝。

So, it represents something special. So I hope you and NVIDIA keep going. What gives you hope about this whole thing we got going on, about humanity, about the future of humanity?

所以它代表了某种特别的东西。所以我希望你和 NVIDIA 继续下去。关于我们正在经历的这整件事、关于人类、关于人类的未来,什么给你希望?

When you look out, when you think about the future quite a bit, when you look out 10, 20, 50, 100 years from now, what gives you hope?

当你往外看、往后想很远,看向 10 年、20 年、50 年、100 年之后,什么给你希望?

I've always had a great confidence in the kindness, the generosity, the compassion, the human capacity. I've always been extremely confident of that.

我一直对善良、慷慨、慈悲,对人的能量抱有极大的信心。我一直极其相信这一点。

Sometimes more so than I should. And I get taken advantage of, but it doesn't ever cause me not to.

有时候信得比应该的还多。我会被人占便宜,但这从来不会让我停止这样相信。

I start with always that people want to do good. People want to help others. And vastly, I am proven right. Constantly proven right. And often it exceeds my expectations.

我总是从"人是想做好事的"出发。人想帮助别人。而绝大多数情况下,我被证明是对的,一次又一次被证明是对的,而且常常超出我的预期。

And so I have complete confidence in the human capacity. I think the things that give me incredible hope is what I see now as possible, and as I extrapolate based on the things that we're doing, what will very likely happen.

所以我对人的能量有完全的信心。而给我巨大希望的,是我现在看到什么已经成为可能——以及基于我们正在做的事往外推,什么极有可能发生。

And that there's so many things that we wanna solve. There's so many problems we wanna solve. There's so many things that we wanna build.

而我们想解决的事情太多了。我们想解决的问题太多了。我们想造的东西太多了。

There's so many good things that we wanna do that are now within our reach, and within the reach of my lifetime.

我们想做的好事太多了——而它们现在已经在我们伸手可及的范围里,而且在我这一生的范围里。

You just can't possibly not be romantic about that. You know what I'm saying? O-

你不可能对这件事不浪漫。你明白我的意思吧?

What an exciting time to be alive. Like, truly-

活在这个时代真是激动人心。真的——

How can-

怎么可能——

… truly so.

……真的如此。

How can you not be romantic about that? The fact that there is a—it's a reasonable thing to expect the end of disease. It's a reasonable thing to expect.

你怎么可能对这件事不浪漫?"期待疾病终结"现在是一件合理的期待。这是合理的。

It's a reasonable thing to expect that pollution will be drastically reduced.

"期待污染大幅减少"是合理的期待。

It's a reasonable thing to expect that traveling at the speed of light is actually in our future. And then, you know, not for long distances, but short distances.

"期待以光速旅行真的会出现在我们的未来"也是合理的期待。不是长距离,是短距离。

You know, and people ask me how. Well, first of all, very soon, I'm gonna put a humanoid on a spaceship, and it's gonna be, you know, my humanoid, and we're gonna send it out as soon as possible, and it's gonna keep improving and enhancing along the flight.

有人问我怎么做。首先,很快我要把一个人形机器人放到飞船上——是我的人形机器人。我们会尽快把它发出去,而它会在飞行途中不断改进、不断增强。

And then when it's time, all of my consciousness has already been—you know, so much of my life has been uploaded in the internet.

然后等时候到了,我的意识其实早就——我人生的很大一部分已经被上传到互联网上了。

Take all my inbox, take everything that I've done, everything I've said. You know, it's been collected and becoming my AI.

把我所有的收件箱、我做过的一切、我说过的一切都拿去。这些都已经被收集起来,正在变成我的 AI。

And I'm just, when the time comes, we'll just send that at the speed of light, catch up with my robot.

等时候到了,我们就把它以光速发出去,去追上我的机器人。

Oh, that's brilliant. I mean, but for me, that's sorta application-focused.

哦,这太妙了。不过对我来说那更偏应用向。

But also, for me, the curiosity-maxing perspective, I just, all of those mysteries. There's so much- … fascinating scientific questions there.

而对我来说,从"把好奇心最大化"的角度——就是那些谜团本身。那里面有太多迷人的科学问题。

Understanding the biological machine is right around the corner. It's, it's not 10 years. It's five years probably.

理解这台生物机器,已经近在眼前了。不是十年,大概是五年。

And then your biological machine, the, the human mind and cracking physics, theoretical physics open. It's so exciting.

然后是你那台生物机器、人的心智,以及把物理、理论物理彻底破开。太激动人心了。

Explaining consciousness, that one would be awesome.

解释意识——那一个会很了不起。

And it's all within our reach. Jensen, thank you so much for everything you've done over the years. Thank you for everything you're doing for the world. Thank you for being who you are.

而这一切都在我们伸手可及的范围里。Jensen,非常感谢你这些年做的一切。感谢你为这个世界做的一切。谢谢你成为你自己。

I can tell you're a great human being, and I wish you incredible success this year. I can't wait. As a fan, I can't wait to see what you do next, and hopefully I'll see you in Taiwan and thank you so much for talking today.

我能看出你是个很好的人,祝你今年取得惊人的成功。我很期待。作为粉丝,我很期待看到你接下来会做什么,希望能在台湾见到你。非常感谢你今天来聊。

Thank you, Lex. I had a great time. And also, if I could just say one more thing.

谢谢你,Lex。我很享受这段时间。另外,我想再多说一句。

And thank you for all the interviews that you do, the depth, the respect that you go through with and the research that you do to reveal, you know, for all of us the amazing people that you've interviewed over the years.

谢谢你做的所有这些访谈——你投入的深度、你带着的尊重,以及你为了把那些了不起的人呈现给我们所有人而做的功课。

I've enjoyed them immensely. And as an innovator, to have created this long form, unbelievable, and yet, you know, it's just captivating. So anyways, thank you for everything you do.

这些年你访谈过的那些人,我看得非常享受。而作为一个创新者,你创造出了这种长篇形式——难以置信,而且就是那么引人入胜。总之,谢谢你做的一切。

It means the world. Thank you, Jensen.

这对我意义重大。谢谢你,Jensen。

Thank you, Lex.

谢谢你,Lex。

Thank you for listening to this conversation with Jensen Huang. To support this podcast, please check out our sponsors in the description, where you can also find links to contact me, ask questions, give feedback, and so on.

感谢你收听这场与黄仁勋的对话。要支持这档播客,请看简介里的赞助商,你也可以在那儿找到联系我、提问、反馈的链接。

And now, let me leave you with some words from Alan Kay. "The best way to predict the future is to invent it." Thank you for listening, and hope to see you next time.

最后我留给你一句 Alan Kay 的话:"预测未来最好的方式,就是把它发明出来。"感谢收听,希望下次再见。