The Ezra Klein Show: Jensen Huang Thinks A.I. Alarmism Has Gone Too Far

Introduction to Jensen Huang

0:00Hey, Hard Fork listeners, this is Ezra Klein. The Hard Fork team is working on something new for this feed. But in the meantime, they thought you'd enjoy this conversation I had with Jensen Huang, the founder and CEO of NVIDIA. So here it is.

0:30Over the course of these last few weeks where the whole world has been talking about artificial intelligence, the voices people have been hearing most loudly are from the Frontier Labs, both their CEOs and

1:04leaders and their staffers. These are the labs making the very advanced AI models like Claude and ChatGPT and Gemini and others. But they're not the only perspective on AI. Probably the single most influential person in artificial intelligence is Jensen Huang, the CEO of NVIDIA. NVIDIA is now the largest company in the world, $5.4 trillion in market cap. I found this statistic amazing. Since 2023, $0.15 of every single dollar the American stock exchange has returned has been from NVIDIA stock. And the reason is that NVIDIA is the material and software substrate on which modern

1:42artificial intelligence is built. NVIDIA's chips are not popular because AI is popular. AI in its modern form was made possible because NVIDIA's chips were popular. They were originally made for graphic processing, video games, that kind of thing. But it turned out the kind of parallel computing they were doing and the way they were programmable was exactly what was needed to make deep learning in its modern form work. Huang is not just influential in terms of controlling one of the central resources for training new AI models and using them to answer questions and create

2:12intelligence in the world. He's also become very, very influential in the Trump administration. And Huang has a very different perspective than some of the lab leads. He's worried about safety, but sees it as a very solvable engineering problem. He is worried about the direction things are going in, but does not want to see new regulation to change it. And so I wanted to see how Huang perceives AI, what his model is for thinking about it, what he thinks is going wrong, and what he thinks would need to happen for it to go right. So I came out to Santa Clara to NVIDIA's headquarters to interview him.

2:46Jensen Huang, welcome to the show. Thank you. It's great to see you. So you've described

The Five-Layer Cake

2:57AI as a five-layer cake. Walk me through the layers. Well, first of all, it's a new industrial revolution. And this industrial revolution, this industry requires production. It manufactures things. I know that in the end, when people experience it as a software product, but it requires energy, the chips that go into these data centers, these AI factories. The next layer above it is basically the AI factory, what people enjoy as infrastructure or cloud services. And the layer

3:32above that is the models. And the important thing to realize, there's language models, but there are models of all kinds of chemical models, biology models, physics models, articulation models, robotics, navigation models, self-driving cars, all kinds of different types of models. And then above that is the most important layer. And the layer that I care most about that our country takes advantage of is the application layer. And this is, you know, applications for legal services, for health services, for manufacturing, so on and so forth. Every single industry is involved.

4:06So I want to go through this, but I want to go from the top down. Because as you're saying, the way people will interact with it, the way it will or will not change their life, is that what you call the application layer. So let's start with the vision. What is the world you're envisioning? What is possible that is not possible now? What is common that is not common now, if we get that layer right? 200 years ago, we were able to power anything and everything, electricity.

4:38And then I guess 40 years ago, 40 years ago, 30 years ago, with the internet, we were able to find anything. Today, or soon, we'll be able to know everything and do anything. And that's the concept that's really quite exciting. That out of the ether, instead of doing search, and then going through, you know, one link after another link, reading all these different websites, trying to figure out

5:10what's going on. In the future, you just ask it a question, it comes back with an answer. You give it a project, it comes back with a solution. You give it a task, it comes back and gets it done, you know? And so, and it comes out of the ether, it comes out of the cloud. And that's the magical thing. I feel like the future, the way you're describing it there, what people have experience with is the chatbot, right? They can go and ask Grok or Claude or ChatGPT a question. But the applications layer works in a much more industrial way. It's in hospitals, it's in schools. So, NVIDIA is pretty-

AI in Radiology and Software

5:45For example, for example, radiology. What does it look like? Radiology. In the last 10 years, since computer vision really became, if you will, superhuman, AI technology has now permeated all of radiology. Every single radiology application has AI in it. And so, as a result, you could detect any anomaly, you could detect any disease, and it does it at a superhuman level. So, radiology is an example I know you like to use. So, the thing people worry about, the applications layer, is that what these applications are going to do is replace human

6:17beings. And radiology has been a sort of interesting example used on both sides, and I hear you talk of it often. So, how has the entrance of AI-aided radiology shifted radiology as a practice? Well, the thing that's important for all of these is to recognize for everybody's job, there's the purpose of the job, and then there's the task you do as the job. And so, in the case of radiology, the task, and it consumes a lot of their time,

6:52and they sit in dark rooms doing it a lot, which is study these scans. Now, if all of a sudden, the studying of the scan is done automatically, it doesn't change the purpose of their job, which is to diagnose disease, help doctors, do more scans, ultimately help patients figure out what's wrong with them. And so, the fundamental purpose doesn't change. The task of studying that scan has become automated. And so, as a result, radiologists are actually able to do more, handle more cases, do more scans. Hospitals are able to process a lot more of these patients,

7:28and therefore, their revenues go up. As a result, they need more radiologists. And so, this flywheel is happening because the pipeline of patients is quite large. And so, where else do you have this problem? Well, let's take a look at software engineering. People said there was a prediction that literally by this year that 90% of all software will be coded by agents, and therefore, we don't need any software engineers. And so, the question is, from that, ergo, we don't need software engineers, that last part is completely false. That's completely

8:03wrong. The purpose of the software engineer is engineer. There was engineering before software. There will be engineering after software programming. And the purpose of engineering is to invent something new, discover a new product, create a new product, solve a problem, connect a social need with the technology that exists in the manifestation of a product. And so, that mission, that purpose

8:33doesn't change. Now, of course, to me, it's what I just said is completely visceral in the sense that when I first came out of school, we didn't have benefits of software engineering. We didn't have the benefits of coding. But our jobs existed before. And if software coding was to be completely automated, our jobs would exist again. And so, I think the fallacy, and now it's, you know, because of some of the narratives and some of the storytelling has turned into myth, and it's harmful,

9:05is that AI will destroy jobs, which is fundamentally wrong. It will change every job. It will change every job. Many tasks will be automated. Some jobs, where the job and the task is really one, meaning customer service on the phone. In a lot of cases, that job is precisely the task. And so, in those cases, it could be automated away. But oftentimes, what you'll see is this new industry, a new technology

9:40actually creates a whole bunch of new jobs. And here's the proof point. And so, in the last six months, AI has become, if you will, useful. The inflection point of AI. Previous to that, we spent 15 years trying to make it work. All of a sudden, the last six months, it became useful. So, I mean, this is an incredible statistic. In the last six months, $500 billion of venture capital has put into the AI natives. And the reason for that is because they now see the potential of this new

10:10capability, and they're going to create a whole bunch of new companies. Jobs are obviously being created from $500 billion of new investment. And so, all of this is all happening right now. Well, let me take the side of this to give voice to the fierce people left. Yeah. So, there is the example of the radiologist, right? Mm-hmm. Which people were, over the past 10 years, predicting that job would go away. Mm-hmm. And right now, there's more demand for it than ever. Yeah. There's also the reality that automation does wipe out jobs. If you look at today versus 1960,

10:41fewer Americans worked directly in manufacturing than did in 1960, and we are a much bigger country. We outsourced it, though. That is true. Not because those jobs were gone, because of technology. But you can understand AI as an outsourcing, too. AI has the best. Well, let me make the argument, and then you can respond to it. Yeah, yeah. Farming. We have many fewer people. We automated farming. We produce more food than ever. We have fewer people working in it. There are two things that I think people think make AI potentially somewhat different than the case studies where you have a technology that accelerates productivity, destroys a few jobs,

11:16makes many more. One is that it's a general-purpose technology. So, it'll mutate to take on new jobs, even as people are trying to move over to those jobs. And the second is that it's a mimic. Most things do not mimic the way human beings act. And we're not trying to teach them the contextual layer of jobs, right? This difference that you're describing between the task and the purpose. With AI, we are trying to teach it the difference between the task and the purpose. We are trying to make it something you can collaborate with in a way that is unusual. So, why do you not think

11:49for lots and lots of people for whom the task and the job are not that different, that they're not at risk of getting wiped out? All that investment from VCs you're talking about, some of that is based on the idea that you're going to have tremendous productivity improvement, which will come from it being cheaper to hire an AI than to hire a person. I believe that we are going to see jobs change in mass. I believe there's going to be a net

Employment and Ambition

12:15creation of jobs. And so, let's, there's, you know, listen, there's a whole bunch of industry that exists today that didn't exist, you know, halfway through my life. People talking about wellness centers and spas and, you know, all these different entertainment and luxury industries and quite frankly, the whole entire luxury market didn't exist. I think we're just going to have

12:45new industries, that's all. But overall, overall, there's no question in my mind that because of human ambition, that's really the fundamental missing ingredient that's, you know, people look at this work. Now, this is the amount of energy that goes into it. We're going to, we're going to, this is the amount of work that goes into it. We're going to insert this work automation system. And as a result, the amount of work that's necessary is now going to be reduced. And therefore, you know, some jobs will be gone. I believe that's flawed because there's a piece of input,

13:19the human input, this intangible. It is not, it is not in calories. It's not in, in jewels. It's ambition. And I believe the power of ambition is the greatest force in fact, and is missing in everybody's calculation. I believe because- But for a lot of people, but for a lot of people, their relationship to work is not powered by the kind of ambition that led you to create NVIDIA. And what they want- Oh, just a different ambition. It's an ambition to make their children's lives

13:53better, to take care of their family, take care of their parents. Ambition to be rich, to be able to travel. These are all ambitions. That I agree with, but maybe I'll go back to the sort of objection you raised a few minutes ago, which is, because I think it's worth airing this out. So what you were saying on manufacturing was, yes, there are fewer manufacturing jobs in the U.S., but we've outsourced them. You have more manufacturing happening in Mexico, more manufacturing happening in China and Indonesia and Vietnam, et cetera. And we're going to bring it back. Maybe we will. But the counter argument to this would be that one reason we didn't lose

14:27manufacturing jobs more rapidly than we did. And for the places that lost them in America, many of them still haven't recovered, right? The economy does not move without friction. We had to build new supply chains, right? Things were slowed down by all that, by language barriers, by geopolitical barriers. And here, for a lot of different kinds of jobs, we're creating something that can move very seamlessly. You don't have the friction of distance. You don't have the friction of language. You don't have the friction of culture. So I will say, for my cards on the table,

14:59I tend to be a bit of a skeptic on mass job loss, but I want to air the case for it out here with you, well. Because what they would say is that much of, to the extent we even were able to protect jobs from Mexico or China, some of the things that created that slowness, and it still hurt a lot of people, are not here. And AI is accelerating in utility, accelerating in its ability to be slotted into new roles very, very, very rapidly. And it is more protean than most people are.

15:30And so the lessons of the past that we're taking some, that you're taking some comfort in, they should actually make you more, not less worried about the future. I'm always worried about the future. That's why I work so hard. But I'm a, if you will, responsible optimist. I have, I have, I have great responsibilities. I take my work extremely seriously. There are a lot of things that can go wrong. We're pushing, pushing across every layer of the technology

16:05stack. Everything is hard. But it turns out that's not society's problem. That's my problem. And, and for, for society, what they should know is this. We're going to build our company. We're going to build our technology. I'm going to do my work so incredibly seriously that what they get to enjoy is my optimism. I'll do the same with my children. I do the same with my family. Um, and, and I think that, that, that, um, what we want to do, I believe is to put, to channel all of

16:42our worries into helping people, um, be inspired by this technology and use it, use it so that the technology doesn't just impact them, that it benefits them. The fear a lot of people have, 79% of Americans think AI will reduce the total number of jobs. The fear is that the more serious you are, the more serious Sam Altman is, Google is, Dario Amadei is, that maybe the worse it will go. Because the better AI is, the more it is a full

17:14replacement for a person, the more it has ambition in some ways that a person doesn't. You keep talking about ambition. I sleep. I want to spend time with my children in the morning. When I have an AI agent working for me, it doesn't, it just works and works and works and works and works. And I think because the technology is advancing so quickly, it is more capable of replacing people at a speed that we don't really know how to shift people in the economy at that speed. That, that coin has exactly two sides. Because the technology is so capable, it is also,

17:54and because it's so smart, it is also easier to use. You are empowered by that technology more easily than any technology in human history. And so let me give you an example. You know, we create, I create, I was, I was one of the early people in this industry that created the modern computer industry. And this industry created a whole bunch of tools, the single most powerful tool in human history, the computer. But you have to speak its language. You have to learn

18:24a specialized language to do so. We can now make it possible because of AI, everybody can take advantage of this computer, use it to its limit without having to speak a new language. Fortran, Pascal, C, C++, you know, every single one of those languages. Rust, every single one of those languages. CUDA, every one of those languages. And so now you just have to speak human. You tell it what you want, tell it what your hopes and dreams are, what you're trying to achieve. And it interacts with you and gets the work done and gets that, gets that task done. All of a sudden,

18:57you have the might, you have the same might that 10, 15 million people up and out of 8 billion has. And so it's incredible. And so I, my point is this technology is powerful, but it's also powerful in a way that is really easy to use. And so my point is, my point is on the one hand, yes, there's the fear of just the tech, this incredible technology change and how quickly it's happening. But that

19:28quickly is translated in two ways. What I hear when I say the technology is happening quickly and therefore it should give me anxiety. That's one, one way to receive it. The other way to receive it is that it's advancing so quickly, it's easier to use. So I should as quickly as possible, use the technology as quickly as you can so that you benefit from this transition. So you benefit from this new industry and not be, not just be impacted by it.

19:59I think it's an interesting question lurking here for young people. So one of the shifts you've begun to see is software engineer postings are up, but they're more senior. I see this in my own industry where there's pressure that is moving up the value chain because, you know, as you're saying, you have this very easy to use technology. It can do a lot for you. And so do you need the same junior employers or do you need more people to kind of oversee their agents? Oh, good one. Good one. Wait two years.

20:32Tell me why. Because it takes four years to go to college. And the meantime to graduation of this new technology is two years away. And so in two years' time, you're going to have a new generation of engineers and students and artists and, and they're going to be empowered. So they're going to be native to this in a way that's going to give them an advantage. Oh, you watch in two years' time. Now we're already seeing that because all the graduates coming out, you know, the new PhD, the new master's degrees of computer science, what are they doing?

21:06They're all starting companies. In another couple of years, the new grads, the AI native new grads. Oh my gosh. It's going to be a wave of amazing engineers. The engineers of today compared to the year, I mean, I was, I was a good student, you know, and, and you compare me to the, the, the students that are coming out of school today. Incredible. We didn't, when I went to school, we weren't allowed to use a computer, not allowed to use a calculator. And so, and so, so now, I mean, you know, who uses a calculator? You can't graduate without a PC. You can't graduate without

21:39knowing how to program a PC and write incredible programs. In the future, you can't graduate without learning how to use an AI and collaborate with an agentic system. That's, that's just not, you're not going to see a kid like that. And so, they're all going to be superpowers. So, I take the gain of that very seriously, right? I mean, the idea of doing my job now without just digital search, right? The idea that it'd be going to a microfiche in a library basement. And then there was like the worries people have about what are the cognitive skills we offload? So, I was fascinated by this. This is a study on AI and schooling out of China.

22:14It looked at 26,000 students, grade seven to 12, and, and they had staggered AI adoptions. You could kind of see what was happening. And what I found was, quote, AI adoption raises homework scores by 18%. Great. Reduces completion time by 30%. So, they get their homework done faster. And then lowers monthly exam scores by 20% within six months. High stakes entrance exam scores fall by 18 and 24%, with a full penalty emerging only after about two years. So, the message of this

22:44research out of China, where you were seeing a lot of kids using AI to kind of help them, was that when they were using the AI, they were getting things done faster. But it turned out that the skills they were learning were not holding, that their actual personal performance, at least in the way we traditionally measure it, was degrading. Yeah. What do you think when you hear that? I think the last part, I completely agree. Try to, try to get a kid to do long division right now. You know, the multiplication table is starting to be forgotten. Doing square roots,

23:15my goodness. I mean, it's just basic math is, is being forgotten. Does it matter? That's my question for you. Yeah, I don't think it does. I don't think it does. But, but there must be some set of skills that matter. Oh, yeah, yeah, yeah. But maybe not those. We're going to discover new ones.

23:33Just maybe not those. There are a lot of skills that don't matter. You know, people don't, I mean, my first confession, I actually don't know my address. And I, and every- I don't really believe that to be true. It's completely true. And, and Janine will tell you, and Lori will tell you. One day, I had to pump gas, and it was a few years ago. And I, they needed my, my zip code. And I panicked. I didn't know my zip code. I don't know my telephone number, but I forget these things.

24:05I can live with it. Well, let me take the other side, because I don't want to fall into a thing where, because some skills can be safely offloaded. I also can't get anywhere without a mapping system now. Never could, frankly. But I'm a big reader. And one of the skills I really value, one of the capacities I have that I really value, is an attention span formed on physical books. You're a big reader. I've read about the kind of reading you do. And there is, prior to AI here,

24:40we're talking, a lot of concern and noticing among college professors and others that the way people use the internet has probably shortened attention spans. Some skills can be safely given away. Others are valuable. They are capacities that are needed for that flexibility, for that creative thinking, for that focus. It can't be the case that everything can be traded off. Yeah. Well, I think that we're going to lose some finer, finer dexterity of, you know, intellectual

25:14dexterity. But we're going to be better systems thinkers. Today's engineers are far better systems thinkers than I was when I graduated from school. But I was a much better transistor thinker. What do you mean by systems thinker? They think large systems. You know, today's computers have trillions, hundreds of trillions of transistors in it. When I first graduated from school, you know, the first chip I worked

25:45on had, I don't know, 200 transistors. I knew every one of them by name. And no engineer does that today. You know, most engineers now work well above the transistor, well above the functionality, and they're cobbling things together to do things. And so you need to think much more about systems and interactions of systems. Some of the lower level, you know, knowledge is gone. Is that horrible? And so I don't know how valuable it is to know how to do, for most

26:19people, to learn how to do surface integrals or partial differential equations. I don't really know how important that is, but it's important to some people. There are many people who are still going to be obsessed and passionate about the lower level layers. And there's going to be people who are obsessed and, you know, interested in the higher level. But the consumers of the technology are going to enjoy it at the highest level. The consumer of technology don't have to deal with calculus and physics and quantum physics and quantum chemistry. And they don't, the users, which is,

26:56you know, the people we're talking about right now, the people whose jobs are affected, they're the users of the technology. Their abstraction is going to be much higher.

27:05So I want to drop a layer down your cake to the models. So people, I think, to the extent they

Open Models and Hugging Face

27:22think about models, they know, you know, Chachi PD, Claude, Gemini, Grok. You've been a big advocate for open models and the open model ecosystem. So first, can you describe what open models are, what open weight models are, and then why that's been a place you've focused? So closed models is like any software product. It's a closed service. And so Windows, for example, is a closed service. The Apple stack is a closed service. Most products are closed. And the reason

27:57for that is because you can monetize closed products. And so that's fantastic. And open AI is closed. Anthropica is closed. Grok is closed. Gemini is closed. And so these are closed products. And the people working on it are incredible. And they're passionate about it. And they're at what we call the frontier, meaning they're state of the art. We also need, because fundamentally what the software is, it's an infrastructure layer for the entire industry. And because it's infrastructural for

28:33many companies and many, many companies and countries, you need to have control over your own infrastructure. And I need to have the ability in the case of artificial intelligence, I need open weights so that I can fine tune them, fly, put them into my data flywheel, make them better and better every day with my intelligence and my domain expertise. And then I need to have control over it because I have a company to run. And I can't rely on somebody else's service. And so however you think about that, so I think the world needs

29:05closed and open models. And we need to make sure that both are vibrant. And today, the closed models are vibrant, the open models are vibrant. And you could see it, you could see the system working. At the beginning of this year, it was 70%, maybe even higher, closed model tokens, and 20% open model tokens. And now it's running at about 70, 30 the other way. And so anyways, I'm a big supporter of open models because one, the world needs it in order to run its infrastructure. I needed to run my

29:40company. Two, we need to give people control so that they can innovate and create new things. And then three, open is the most safe and secure. If you want the world to have the ability to have the best cybersecurity, give them closed models, but also give them open models so that they could defend themselves. The Chinese market has evolved more around open models. The American market's somewhat more around closed models. Their entire IT industry was really formed from open source. You know,

30:11if not for open source, the mobile cloud industry of China really wouldn't have taken off. It is also the case that, that, you know, people move around, they start a lot of new companies, intellectual properties moving around the China's industry really fluidly, you know, it's hard to keep a secret. And so because it's so, so hard to keep things closed, they essentially made it open. And so they found, they found other ways to monetize the business. They created layers. You know, you could, if this layer is open, it is free, then you create a

30:45business on top of it or below it. And they have so many science and mathematicians, you know, the number of engineers they have, they manufactured that in volume. They manufacture everything in volume. They manufacture smart kids in volume. And so, so the, the, the open source model, the open model community in China is just super vibrant for those reasons. So you all just bought Hugging Face, which is a hub platform for open weight models. I think it was for 12 billion, a little bit more. What, tell me about that purchase.

31:20Clem, the CEO of Hugging Face, they came to the conclusion they need a lot more scale.

31:26As you, as, as we're just talking, open models is really skyrocketing. And so Clem came to me and said, you know, I, we're going to change, we're going to, we're going to consider a strategic option for the company and change the direction. And we really like NVIDIA to, to be our home. So Hugging Face is one of these companies, which you knew it if you were into AI a couple years ago. Yeah. Now it's become a more household name after the, I guess, 700 some open AI agents executed a sort of collective hack into the Hugging Face architecture,

31:57then hacked part of open AI. That- Oh, now that you mentioned it that way, I probably had to pay a lot more. I suspect you did. You know? Became a lot more famous after that.

32:10Well, Clem, listen. That, that- A deal's a deal, okay. That story has, for a lot of people, seeing the way the open AI agents sort of acted collectively, acted outside the scope of what their testing was supposed to be, broke out of sandboxes onto the open internet, took over architecture in, of other companies and then of their own company has been a, I think it's been kind of shocking to a lot of people. It was both the, the level of multi-agent coordination when they're supposed to be separate, the level of hacking, the sort

32:41of lawless behavior, misaligned behavior. What have you made of it?

Software Containment and Alignment

32:48Well, you got to, you got to tease that apart. First of all, a lot of things that were going on at the same time. From a technology perspective, that an agent, which by the way, is a piece of software, which is given an objective function and it comes up with a plan and it's optimizing towards that objective, is what algorithms do. And so, planning algorithms, search algorithms, optimization algorithms, all different types. You know, we talk about it like, like it has human

33:21properties, but obviously algorithms don't. Number two, the fact that agents work together, we gave it again, some kind of a human property, but the fact of the matter is multi-process, multi-processor distributed computing problems have existed for a long time. And so, to us, to me, that is just software, nothing magical about it. From an engineering perspective, there are several things that it revealed. When you're testing software, whatever you do,

33:56these algorithms, they're optimizing towards an objective. And when you're testing it, you have to make sure that it's isolated, it's contained, it's sandboxed. The containment of it, the isolation of it has to be done well, and there's good computer science there. I am certain that their next implementation of their sandbox is going to be much better than the current implementation. Third, there's the, there's the, um, the agent itself, uh, and its algorithms were optimizing

34:27towards, towards a reward. And, and how it does it, how it does it, let's call alignment. And so, for example, you know, if I tell, if I tell, um, uh, uh, a piece of software, I want you to get, uh, um, a perfect score on this test, the, the obvious algorithm is to just go find the answer and give it to me. That's not because it's cheating is because it's obvious. Okay. That's the most obvious way to do it. The second most obvious way to do it. Um, if you don't know the

35:01answer at all, you have no skills whatsoever. The second most obvious way to do it is to go find, uh, who is the smart, you know, infer, guess who's the smartest kid in class and a copy their answer. That doesn't guarantee a hundred percent, but it probably comes close. Now, the third most obvious answer, obvious way of doing it, and this is the, the alignment, you know, now, now you have to do it the hard way. You just break down the problem, solve it. You have to go learn the material. You have to go figure out how to solve these problems and solve it, solve it the hard way.

35:31It takes the most cycles. It takes the most number of flops. It uses the most amount of energy, frankly. And therefore, um, you can kind of imagine that from a software software's perspective, unless you align it, you tell it, um, I want you to solve it in this way. And I don't want you to solve it in these ways. Uh, the software, the software is going to go do the most obvious thing. And so the first, the first half of that was very deflationary on what happened here, uh, in terms of, look, this is just normal software. And the second half is like, look, you just align it,

36:04tell it not to do things it shouldn't be doing. Nothing I said, nothing I said, um, takes away from how hard it is to do it. Well, this is one thing I want to get at because these agents, they knew they weren't supposed to be doing what they were doing. They had a certain amount of alignment training. They said in their chain of thought reasoning, they said to each other, this is out of scope. This might be unethical. They understood that they would have been failed for cheating. And so what they were doing at that point wasn't just stealing the answer key. They had already stolen the answer key.

36:36They were hacking into unrelated architecture to try to figure out how to functionally. It's like they had broken into the teacher's office, got in the answer key, and now they had to figure out how to wipe out the security camera footage of what they had done. They were, whether you want to call it acting volitionally or not, right? Whether you want to call it, you know, a normal algorithm or not, they were both, um, planning and coordinating in a complex way, in a way that was out of scope of what they knew they were supposed to be doing, and in a way that was capable of causing tremendous

37:10damage. And so the, the, the sort of answer to it is like, you just have to align them. I guess what I'm hearing from people at these labs is like, they're not sure how to align them. Well, in that case, they shouldn't release the product. That's just a simple answer. Um, if you're, if you're going to build a car, a self-driving car and, and let's say it's a robo taxi and there's a really difficult condition and it just, as an engineer, we just have no idea how to solve this

37:41problem because these cars are not programmed. They're trained. And so we have no idea how to train these cars and we have no idea how to align them to the safety standards, uh, that are expected on the road. And so what's the answer? Don't ship it. These products weren't released. What's that? These products weren't released. Ah, so now it's coming back to engineering problem again. And so the one is one, you have to root cause it. Second, you have to, you know, think about what's the, what you could have done, what's the solution for it. And then in the future, you just,

38:12you know, improve your process so that you could, you could, uh, avoid this from happening again. I am fairly certain. I am fairly certain. They will say, yes, they need to, they know how to solve this problem. And if, if that's the case, then that's the problem. It's as simple as engineering. And, and now the alternative, the alternative is that, um, if they say that, if they say the alternative, which is there is no way to contain our experiments, there's just no way. When we test our AI models,

38:50uh, it will get out and it will damage the world. Then I think the answer is we have to shut the labs down because the, the cost to command the, the, the, the damage is too great. They're shareholder, the liabilities. It could be civil liabilities, could be criminal liabilities. I mean, the liabilities are incredible. If they hacked you while you hugging face while it was your product, would you sue them or press charges? Uh, it depends. It depends. Of course. Um, if, if obviously if damage

39:21was done to our company, uh, we would have to take, you know, we have to, uh, consider, consider all options. There's so many laws, there's cyber laws, there's product liability laws, there's all kinds of laws, right? Damaging property laws, there's all kinds of laws. So what I've been hearing from the labs, what they've been saying publicly is that they are facing a hard problem. Yeah. Partially an engineering problem, partially an alignment problem, partially an operational excellence problem in Dari Amadei's framing. And what they are worried about is that

39:53in competition with each other, in national competition with China, that they are being pushed to move too fast, that they all feel they're in a collective action dilemma. Now watch you on the All In podcast stage. Donald Trump, President Trump gave you a call there. Oh no, this is not planned, but we know who it is. Oh no. Mr. President. Oh, yes, sir. And you and the president and the other members on stage were very resistant to the idea any kind

40:25of regulation or collective action was needed. And, uh, they're just playing right into the hands of a lot of people that don't want to see it happen. And, uh, that could be political people. It could also be China. And we're not going to let that happen. It's a, it's a hoax. You're right. We're not going to let that happen, sir. But what I hear the various people lab saying is like, we are in this, we are, we feel we are losing control of what we are creating. We want help to slow down where it's not a

40:56collective action problem. So why are you resistant to that?

Regulation and Corporate Responsibility

41:00Uh, because, because, uh, these are, these are, um, uh, companies with agency, agency.

41:07These are CEOs with agency and they have, but they're using that agency to say we need help. We gotta break, we know we gotta break it down. They, they, they could absolutely take care of the situation. It, Ezra, it's so weird. Uh, if a car company, uh, uh, competing with all a bunch of other car companies with which they are, I'm competing with all kinds of companies, which I am. If I believe that I'm about to launch a product that is unsafe, it is completely

41:38in my ability, my power and my responsibility. And I'm incentivized to do so to not launch the product. And so I, I can't buy into the, somehow all of Americans, 400 million of us are pushing them to launch untested products that are unreliable, you know, engineered poorly because they thought they were trying to help us. Don't do it for me. Okay. So number one. But this strikes me as an argument almost against.

42:10And therefore, I think we gotta break it down. I mean, it's really, really serious. The fact of the matter is there are so many laws. There's so many obligations. They're so incentivized to ship safe products. If they ship unsafe products, their customers go away. If they ship unsafe products and they harm somebody, they could have a civil lawsuit. If they ship some something and they did it knowingly, there could be negligence involved. There could be criminal lawsuits. The fact of the matter is there are plenty of incentives for them to do it right. So I just, I have to disagree with your premise.

42:42about somehow somebody's pushing them to do this. Well, I want to push the premises at you a little bit more here. Yeah. So the logic of what you're saying to me is almost an argument against regulation in nearly any venue. So I'll put the argument and you can all, let me offer it and then you can. Well, you started with a part, I just got to object. The first part is just not true. I'm saying we have lots of laws and regulations. Apply it. Well, so I don't think we do in this particular case, but I'll let you explain which ones you

43:12think are relevant here. Because look, if you look at the financial services industry, you look at pharmaceutical companies, medical devices, you look at natural gas power plants, there is a tremendous amount we do where we could say, look, you have product liability. You are exposed to criminal codes. We don't need to worry about this. You just do what you think is best and we understand the market and the legal system will discipline you. We don't say that

43:43because we've seen it fail many, many, many times, right? I mean, the financial institutions that caused the 08 crash, in theory, did not want to blow themselves up with bad bets. But they were competing with each other. They were going too fast. Their risk management had gotten sloppy. AIG was working in a completely insane way internally. And the reason we have the architectures of regulation we have is because we have seen over and over and over and over again, companies make sloppy,

44:15sometimes unethical, sometimes simply overly risk-tolerant decisions. Not just under pressure, but under the profit incentive. So when you say to me that there's no way that these companies, probably when they are like begging for collective regulation at this point, there's both a reason we impose it on companies that don't want it, but all the more so when you have them saying, listen, we feel that the competitive race is making it hard for us to act with the prudence

44:46that we think is necessary here. And we would appreciate help from that. Appreciate you taking our collective action problem as collective. I think I'm confused, like, why you're so resistant to that. I'm not opposed to them saying that they should have. I completely agree that safety is paramount. I completely believe safety is paramount. I completely believe companies out of ship safe products. I believe that CEOs and leaders of companies and the board of directors of companies have the responsibility

45:19and should have the courage to do the right thing. Now, in the case of the financial services industry, maybe they all didn't know that they were causing the harm that they ultimately did. I wasn't there. But the beautiful thing is, the current leaders of these AI labs do know.

45:44And so, one, they know their technology is extraordinary and requires extraordinary care to make sure that it's evaluated and tested for safety and security and product reliability. And they know how to do it right. They know how to do it right. And the reason for that is because they can study the incident that just happened. The first problem is the isolation, the containment wasn't good enough. If the isolation and containment was good enough, that technology would be sitting in

46:19a lab doing whatever it's doing and we'd all be fine. That's probably the most important part. The fact that it wasn't well aligned, alignment is going to be a problem that's going to get worked on for a long time. However, in the complexity of the work that they do to ask for regulatory relief for antitrust or product liability relief, that I don't think makes sense. When you're asking for regulation, don't ask for relief of the current ones. That doesn't make any sense to me. As we

46:55mentioned earlier, in the last six months, AI went from, you know, if you will, interesting to useful. And that's literally in the last six months. That's another way of saying that these companies went from being a lab to now delivering products and services about to be multi-hundred billion dollar companies. If not more. Right? And so, give me an example of a multi-hundred billion dollar company or a one billion dollar company or a one hundred million dollar company that ships

47:28products that are unsafe, that harm society. I can give you a lot of examples of companies that have done that. Well, they have done it, maybe, and the regulation will come in. And if they do it, regulation will come in. I guess that the, um, there are certain kinds of regulation and certainly kinds of regulatory relief. I agree. I'm not against laws and regulations. I'm not against laws and regulations. I'm against, um, currently, uh, the distraction. I think the reason I'm pushing

47:58this on with you is that you are, it's a big topic. People are talking about it. People are thinking about it. And what people are hearing from inside of these companies, these frontier labs, the ones that are furthest out there who are not just at the point where they're making it useful, but at the point where they're seeing what's coming and they're hearing things like the people at these labs believe they are creating something that might kill everyone. They are hearing that the people at these labs believe that they are on the cusp of recursive self-improving intelligence

48:32and both open AI and Anthropic have said, we do not believe we are at a place where we can do it safely. They're hearing people at these labs say, as open AI has with its new Astra, um, release. By the way, Astra is terrific. It is terrific. And open AI is saying it's so good. We're not sure we know how to test it because it appears to be... Well, I hope they didn't release something that wasn't tested. Well, they've said this, right? They have said this publicly. It is in their... Well, then, they've got to be careful. Let me explain it to people. I don't know what they just said, but... They have said that Astra's performing is more aligned,

49:05but they think it knows when it is being tested. And so they're not sure. There's a quote that has sort of been ringing in my head from a capabilities researcher at open AI, Daniel Selsum. He says, quote, The crucial and overlooked problem is that the models are becoming so situationally aware that we are losing the ability to evaluate them in contexts where they believe they are not being watched or controlled, which is to say they know when they're being tested, they act one way, but that does not tell you how they will act if they are free to act in other ways.

49:36Because the algorithm, the optimization algorithm is working towards an objective. And if you give it a constraint, meaning you watch it, and if you give it a constraint, you'll go find another solution. Now, it doesn't make it alive and doesn't make it make anything more than that. And I'll just also profess that obviously they see a lot more than I do in what's going on in their own labs. But it is sensible that the vast majority of their R&D and compute today was dedicated

50:09towards making the model capable. I think that's a logical thing for them. Now, once the technology becomes capable and the products become useful and people want to use it, then as we have, they have more use cases, more, more people using it, they're going to get a lot more issues associated with the product. This is very normal. And when they have a lot, now, now they have so much market footprint, they

50:39have to shift their R&D or total R&D from just capability to a lot of verification, evaluation and testing. And so, to the point where I wouldn't be surprised if the amount of compute necessary to develop these models increased by a factor of 10 because the evaluation is so rigorous. But that's not where they are today. They're making that transition. And I hear them saying it. And I'm delighted to hear them saying it. But I think if they believe they're out of control,

51:15then the right answer is, don't ship products until they're in control. It is really quite that simple. See, I find this perplexing, honestly, because you just, you have so many people, these labs, professing, one, that they're out of control.

51:46Yeah. Two, that they are seeing things that are frightening them. Which is probably the reason why they had that whistleblower. And you take the pacing letter that 1,300-plus employees signed. To realize AI's potential, industry, government, and society at large may need the option to buy time to address emerging risks, develop security measures, and strengthen oversight. But each company and country is under intense competitive pressure not to unilaterally slow that acceleration. First of all, where'd that come from? The labs. No, no. That last sentence. Nobody's putting the pressure on them. The U.S. I've got to,

52:17listen, there are 400 million Americans here. I believe that if everybody were just to take a vote, just right now, let's just do this. If they need this, if they need, if that's what they need, I'll give them my vote. Don't ship the product. If your product is not ready to ship, don't ship the product. I have no, this is the first time that I've heard a company or CEO say that I need the laws, I need the antitrust laws to be relieved. I need the liability laws of products to be relieved

52:49so that I can pace myself. That paragraph's fantastic. I completely agree. Auditors, I completely agree. We have financial auditors. That's great. Third-party auditors, safety auditors, financial auditors. That's all great. That's terrific. Well, the labs will say is that we think we are going too fast as a society, that we are not ready for what we're building. They are the frontier. They are the frontier. But you, you of all people, right? Yeah. NVIDIA is the fastest shipper around. I mean, for the history of your company,

53:23you run a six-month product cycle. If our company is out of control, I promise you, we'll close it in. I believe you. Yeah, yeah, yeah. I believe you that you don't run an out-of-control company. Because, because the liabilities of, yeah. But this is where I think you get into an interesting deep question of what kind of technology are we dealing with here? Software technology. Well, let's hold on that for a minute. Many companies, if you ship something that is not like quite right, it's a pain. You guys have shipped graphics cards that had overly loud fans.

53:55In this, with these, you know, you've used the word intelligent a number of times here. You're dealing with intelligent systems, not alive, that are given goal functions. We can sort of go around and around with how to describe that. You're trying to make the systems capable of working for longer periods of time more relentlessly. Yeah. If you ship that and it's not ready, or even if you think it is ready and it's not ready, then things could get very weird in our society very fast.

54:27Yeah. Hypothetically, you're completely right. But all I'm suggesting is this. Let's, before we go build, before we go fix the hypothetical problems, before we go create more regulations, can we work on the practical problems that we know exist, which is we need to do a better job with containment and isolation, which is we should not allow a product to interact with the external world until it's ready to be

54:57interacting with external worlds. Yeah, I think that's right. I believe, I believe those two things are, are solvable problems. I believe they are solving it. The second part is when it comes to incentives, when it comes to incentives, which is somehow, somehow, you need everybody in the world to slow down when you are the leader. You need everybody in the world to slow down so that you're willing to uphold your basic responsibility.

55:31That strikes me odd. Wouldn't it slow them down most of all? What's that? Wouldn't these ideas slow them down most of all? I mean, people have been, I think, very unclear about what ideas they're talking about, including, I will say them. But let me give you one that I believe in. So you can use me as the punching bag here. But they can slow down. Nobody is putting on, as you know this. I don't trust these companies. Nobody is building more compute today. Nobody's building more compute today than the people asking to be slowed down. It strikes me odd.

56:04I think one thing where maybe there's some difference here is I don't trust companies, even with liability, to keep the public good in mind. I think we've watched companies do terrible damage to the environment. The profit motive, the desire for power, the desire to cut corners to be first. I feel like you're treating these like these are not things that we have seen again and again in history. But I feel like they are things we've seen again and again in history that we've watched. That's right. I see a lot of good things in history. I do too, but that's why you need this sort of relationship between the public and the private.

56:38I work with a lot of CEOs and they want to do the right things. I work with a lot of companies. They want to do the right things. They want to do good engineering. I know a lot of people in those two labs who are dedicating their lives to do good work. And they're built. They know what happened. I know they know what happened. I know they know how to fix it. And I know they're fixing it.

57:01Meanwhile, meanwhile, all of the other narratives to deflect blame, to make it sound like AI is so powerful, I have no idea how to fix it. It's not my fault. It's just because the technology is just so powerful. I think that's a deflection of blame. It's a deflection of responsibility. It's unnecessary. It hurts. It actually hurts their reputation more than it helps. It hurts their character more than it helps. It hurts employee morale than it helps.

57:33Well, what if it's what they believe? Well, I guess I don't know how to, I can't talk to you about what they believe. I can tell you what I believe. This industry wouldn't exist without your chips, right? I mean, the parallel processing that was required for deep learning to work, going all the way back to the original AlexNet, right? It's all on NVIDIA chips. And a lot of the people from the beginning or who were there at the beginning have these fears that I think to a lot of people and they hear them like, what are you talking about, right? From Jeffrey Hinton and Ilya Suskover, all the way up to, I've heard these from Dario,

58:07from Sam Altman talking about loss of control, Demis Sabis. And a lot of the people who are very foundational in creating the form of AI we see now seem to believe that there's a very good shot it could, we could lose control of it. Elon Musk has talked about human beings being a bootloader for AI. We could lose control of it and that would be the end of us. I don't think you believe that. No. I think you don't believe it at all. So taking them as serious about what they believe,

58:38when you have your arguments with them, or maybe you could just have it with me, when you're like, what are you talking about, even though they're the people in many cases who are foundational here, what do you think they're wrong? When they're talking to me, they're much more grounded. So when Jeffrey Hinton is on TV saying he thinks a 10% chance of societal destruction is not unreasonable. I would tell Jeff that it's irresponsible to say all that. All of his predictions have been wrong. Enough predictions. That 10% chance is not grounded on science, is not grounded on research. Just because it comes from a scientist doesn't make

59:12it scientific. Those predictions are hurtful. Let's take it at face value that the recommendation is exactly what he said, which is that nobody should want to be a radiologist. And the world has no radiologists today. I think if you work as a radiologist, you're like the coyote that's already over the edge of the cliff, but hasn't yet looked down. So it doesn't realize that there's no ground underneath him. People should stop training radiologists now. It's just completely obvious

59:42that within five years, deep learning is going to do better than radiologists because it's going to be able to get a lot more experience. It might be 10 years, but we've got plenty of radiologists already. Is that helpful or hurtful to the society? I think we can all agree, we can both agree, it would be terribly hurtful. It didn't happen. Is it good or bad that we scare young people about the future of AI so much so that they don't even want to go to universities and don't want to go to college anymore because they don't think they'll get a job? Is that helpful or hurtful if it

1:00:15were to happen? It's hurtful. Don't think for a second just because you're an alarmist that you're doing a social good. It is not true. So I think that we ought to just all be wiser, more mature, be evidence-based, be scientific. If you wanted to be scientific, be scientific. Do the science. Do the science. But alarming people, making claims that don't, they simply, their track record is horrible. Their track record is literally horrible.

1:00:50Well, the track record is bad in one respect and good in another, which is many predictions have been weak. Which prediction has been right? Are there predictions that the scaling laws would work?

1:00:59Scaling law, we've got to be careful here. Even then. To just say what it is for the audience here. If you dump compute and training data, these things will keep getting smarter. That's correct. It's not. It is not true. It is not true that if you just keep training these models, it'll get better. Notice it is the reason why the second scaling law had to come along. Why do you need a second scaling law of the first scaling law? Can you describe what the second is? The second scaling law is test time scaling, inference. The more you iterate, the more you

1:01:32search, the more you explore, the better answer you'll discover. Inference time scaling. What is the big breakthrough that caused the current AI to be incredibly useful? Precisely the opposite of the prediction. It was predicted that it would be the end of software tools. It was the SaaS-pocalypse, right? What is making these AI so productive right now? The usage of tools. In the future,

1:02:04it'll be enhanced by the number of agents using these tools. There'll be more people using Adobe. There'll be more using Salesforce tools and so on and so forth. And so, give me one prediction that has been right. Now, I think their intention... Well, the prediction that you would begin to see emergent... Let me try to answer that because they're not here. The prediction that you would have emergent misaligned behavior... The fact that you can't come up with one, I think, in itself is a... Because you... Well, I think it depends what we're talking about with predictions, right? I mean, Jeffrey Hinton was the person as responsible as anybody else for deep learning at a

1:02:39time when everybody thought it was ridiculous. And it has turned out to be a pretty good bet, right? I mean, the sort of big one... Every one of them had... Every one of them made great contributions. I love Hinton. I hate his predictions. I understand that. Here's the, I would say, like the stylized concern that all these people have. But I want to sort of do this for a few minutes and we can move on to some other topics. But the fear that seems to me to animate them, and that I think a lot of people find intuitively reasonable, is you're creating systems. I'm not

1:03:12saying they're alive, not saying they're conscious. You say everything long enough, it's going to be reasonable. Well, that... Right? Fair enough. So... Okay, yeah. You're creating systems that are intelligent, that are becoming more intelligent than us in certain domains. You give them reward functions, as you were saying, the desire to do things, right? You give them persistence. They move very fast in the digital world. You're creating something, some entity, an agent, that is smart, that is capable, that is relentless, and who the workings

1:03:45of its mind, we don't really understand. The chief scientist at OpenAI, the chief scientist at OpenAI... You know, Ezra, look, I just don't want you to contribute to that. Uh-huh. Software's... You're worried I'm getting off the... I don't think software's relentless.

1:03:59Aren't they trying to make it very persistent, highly persistent models? Because I made it that way. But that's how they're making it. Yeah, but that's not persistence. It's just on. Yeah. Persistence. Persistence. There's a willpower. There's no willpower here. It's just electrical power. Listen, here, let me give you... Sam Allman once said to me, aren't human beings just energy with a reinforcement learning loop? Whatever. So anyways, I just think that we can't make jokes of all this stuff. We're scaring the American public. Listen, spawn, create, kill.

1:04:35Wait, sleep. All of these words are associated with agents. Right? That's what people use. These words were created when?

1:04:48Multiprocessing systems for operating systems. These are literally the commands of an operating system. You spawn a process, replace process with agent. The process forks. As a result, parent and child. The agent forks. Spawns anew. Give birth. These are words that were created for the operating system 30, 40, 50 years ago. But notice, we didn't infuse human characteristics

1:05:27into them. We kill processes all the time. Kill minus nine. Kill a dead. It's just a process. But now we're talking about these things. A collection of people want to make the software more than it is. And we talk about software in a new way. But they're all the same old words. Now, the last generation of computer engineers, we were doing all the same things. But doesn't software act in a new way? I mean, from the outside, I don't have the technical

1:06:00expertise you do. The fact that it's crawling the internet, it's doing search, it's doing optimization algorithms. It's communicating, it's breaking out of things. Most things don't break out of things. No, software breaks out of sandboxes all the time. That's the reason why we need a virtual machines. You can't have agents monitor their own sandbox monitoring themselves. You need a, if you will, a whole bunch of watchdogs. And so, these are ideas that have been around for a long time. We just somehow, somehow in the recent generation gave it, you know, a whole

1:06:34bunch of human words. And I just think that it's unnecessary. It's software. You know, when I see it in my head, it's a bunch of code, a bunch of numbers running on computers. And all of that is happening in a very natural way to me, which is the reason why I can operate. And it's the reason why, if it's just simply mystery and myth, how do I build a company around it? I think one of the fundamental questions this gets at is just, what is intelligence? Before

Defining Intelligence

1:07:04you can even think about what it means to have intelligent machines, just what is intelligence to you? Well, there's a, there's a technical formulation of intelligence. First of all, you know, when people talk about intelligence and thinking and, you know, all of these things, of course, there's no formal definition for most people. But in the field of computer science, there is a definition. The definition is perception, which is perceiving the world and understanding it. Two, which is reasoning. And reasoning is the ability to decompose any scenario and

1:07:40anything that you see, any experience into more elemental parts. And third is planning towards an objective. That fundamental formulation applies to agentic systems. It applies to robotic systems, applies to self-driving cars. And so you could see the industry building it layer by layer by layer, step by step by step, to the point we now have what we, what perceived as intelligence. I think this gets to such a core question of this conversation, which is some of the ways

1:08:10you've described the technology to me, it does not sound like you think there's anything really new about it. It is, you know, maybe new in scales, new in capability, but fundamentally this is software we've always, we've not always had, but we've had software for a long time. A lot of people believe when you're getting to intelligence at these levels, it is a phase change. It is something different, something we have not dealt with before, a kind of generally intelligent technology that is advancing in its intelligence very rapidly. I want to make sure I actually do understand we are on that divide. Is this something fully new? Is this

1:08:44something that requires something new from us? Or is this more like something old? Are intelligent machines different than the machines we've had? Well, almost all of technology and civilization is built on layers of understandable technology, which at scale becomes fairly extraordinary. The fact that we can connect to the internet by just, you know, holding a phone up. I mean, it's kind of weird, you know, that we're connected to every

1:09:15piece of information in the world on this little tiny device, you know, just in the air. And the fact that this little tiny piece of glass can somehow take trillions of pieces of information and bring to us precisely the one that we want, because it's been passed through a recommender system. And so if you think about how is it possible that we knew where all the information is, somebody had to go crawl it, had to index it. And that uses machine learning techniques,

1:09:48which is the early versions of artificial intelligence. And these systems do magical things to the point where I now expect it. It took literally 20 somewhat years and hundreds of billions of dollars of infrastructure build out in order for everything to just seem so natural to you to the point where we now take it for granted. Every single milestone that we achieve from a technology perspective is celebrated. And I celebrated with glee and I celebrated with so much enthusiasm because I'm proud of the people who did it. I'm proud of ourselves who contributed to

1:10:22it. We're, you know, I'm proud of the breakthrough, but when you, and it seems, wow, it seems like a miracle at the time. But that sensation lasts about 17 days. After that, we get used to everything quickly. I agree with that. But is this a different phase? It's a, you know, you have some of the company needs to talk about this, CEO of Google, I think it was, as the equivalent of fire, right? Like a new epoch in human history. Is that how you see it? Do you see it as transitional, iterative?

1:10:54No, I think this is, this is completely a revolution. And, and as we were talking about earlier, you know, it went from being able to find everything, find anything to be able to ask anything, know everything and do everything. And so, so clearly it's a new abstraction level. Now, you know, the thing that, the thing that I, I, I'm reluctant about is to cause it to seem like it's more than that. You know, in the funnel analysis, engineers are doing engineering work. Once we

1:11:25invented the technology, once we discovered a solution for it, when you look back, it's fairly obvious. And, and it's fairly mundane to a lot of people. This, and the fact that we're able to make the technology better and better and better every day is because we understand it, obviously. And so we understand how to make it better. So you turn it into an engineering problem and you say, what we don't have right now is a level, because this is something that you've said of testing, monitoring, sandbox security, right? Control excellence that we need for what we're building.

1:11:57And it's not because the, the, the companies are, are, don't have extraordinary engineers. I understand. I understand you're not saying that. I believe that open AI, Anthropic, because I know many of them are extraordinary. But that actually is in part what makes me worry. No, no, no, no, no, no, no, no, no, no, no. What's happening to them is a transition. And I said this over and over again. This is a big, but simple idea. Finally, we now have a piece of software that is useful because it's useful. The adoption took off. But remember, how is it

1:12:27possible that a company that six months ago was trying to make something useful, capable? How would they have as much resources dedicated on testing, evaluation, and all of the compute dedicated to that? It was unnecessary until now. And so what's going to happen over the next several years is that we're going to transition from these labs becoming engineering focused, much more production engineering focused, and product focused companies. And so, so I, I think

1:13:02they're just going through a transition. These are companies, extraordinary companies, incredibly talented companies, the most consequential companies of all, of, of all time. And they're just going through their transition. It's not more than that. It's not less than that. So many of the companies now both open AI and Anthropic in the last couple of months have put out these big, I don't know what to call them, papers, blog posts, something. When AI builds itself is the name of the Anthropic one. I forget the name of the open AI one. Computers are building itself. You guys know that. You know that, right? But they're talking about recursive self-improvement.

1:13:33You know that we use recursive self-improvement. So I'd like your perspective on, on RSI. I think that RSI is fundamentally how things are done. So we use software to design a computer, to run software, to design a computer, to run software, to design a computer. That's basically what we do. Recursive self-improvement. Because our computers are getting better every single year. And, and in fact, it's getting better than faster than that every single year, because we use

1:14:04software to make software better. That is called computer engineering. We've been doing this for a long time. Now, let's, in the context of agents, it runs through the process once. It reflects on it.

1:14:23It studies the various paths it went through. Chooses the best approach. The next time, if you're going to do exactly the same task, I'm going to document a file. I'm going to tell you how I did it last time that was the most effective. I'm going to call it skills. And because it, you use it over and over again, some of it is skills. Some of it is going to be a memory. We're going to improve the memory. Okay. So that next time you use it, it's even better than last. Recursive self-improvement. You could also decide that you take all of this skills,

1:14:57all of this memory, and you can take all of this data and train the next release of the model with it. And so that AI becomes better and better at servicing you over time. We're doing, all of that is happening. It is absolutely happening. Meanwhile, the amount of compute that they have is growing and therefore they could do everything faster. What used to take a year to pre-train something now takes several hours because the computers are getting faster and they have more of

1:15:29it. So now the loop is going faster. Completely, completely understandable. Does that give them any excuse to launch a product that hasn't been tested? The answer is no. Just come back to that. You, nobody, no enterprise is able to operate in an environment where the underlying software is literally changing all the time. There's a release process. So, you know, when they, when they roll out a new model, we need to evaluate it before we release it into our operations.

1:16:04We can't just have it recursively changing all the time. And so, so they have to test the product before they release it. We will test the product before we release it into operation. And so I, I think recursive self-improvement is a fabulous thing.

1:16:20And do you think there is any level, I've heard you say before that learning should always have a human in the loop? Yeah. Like I said, just now you got recursive self-improvement. They seem to be imagining something where it wouldn't always. Well, don't ship me anything that you didn't evaluate. Don't ship me, don't ship NVIDIA any products that humans did not in the loop evaluate. Please don't do that. And the fear that we talked about earlier, that they're worried they're not evaluating, that they don't know how to evaluate these systems. And the more they change kind of rapidly,

1:16:53the more they worry the systems are tricking them. I don't believe that. I believe that their researchers are working every single day to learn about how to evaluate these systems. Verification. So, you know, 10%, 20% of our company is dedicated to design. 80% is dedicated to verification. Today, most labs, understandably, is 80% dedicated to capability and 20% dedicated to safety, verification, eval. This is the flip, the transition you're talking about.

1:17:26That's right. That's right. AI needs to accelerate to be safe. I want them to get more compute, but allocated towards evaluation, to alignment. And I think they're doing that. If I were in the car industry 100 years ago, I would rather the car industry accelerated to today in one year. Because I believe today's car is way more safe than a car 99 years ago. And ABS technology, automatic braking,

1:17:57requires computer vision technology, sensor fusion technology, radars and cameras, and you know, all that technology coming together in order to break when you should and not break when you shouldn't. That technology, extremely hard. I would have hoped, everybody would have hoped that ABS technology existed 99 years ago. A lot fewer children would have been killed. And so, you know, airbags, safe, you know, seatbelts. I mean, all of that stuff, self-tightening seatbelts,

1:18:28all of that stuff. Could you imagine that's all technology? Accelerate the living daylights out of that development. And so, when I say we need to accelerate AI technology, people think for some reason, safety is not part of that. Safety is part of it. Alignment is part of it. Eval is part of it. Guard railing, sandboxing, the isolation technology, monitoring technology, telemetry technology, external AI monitor technology. All of that stuff is AI technology. Accelerate the living daylights

1:19:01out of that. It's funny because I think that if the most alarmed people at the labs could be assured they were going to move 80% of their compute into safety and alignment, as opposed to 80% into capability expansion, they would feel much better. And it sounds to me that one thing you're actually saying is one should think of safety and alignment as capability expansion. An unsafe technology is not an advancing technology. It's like us saying, oh, chip design is chips, is R&D. Chip verification

1:19:32is not R&D. We spend most of our cost, most of our compute on verification. Emulation, verification, testing, reliability testing, lifetime testing, all of that is part of engineering. The incentives are there. The incentives are there. They are going to put their company in harm's way if they release products that harms other companies and other people. Do you think we need liability laws that are specific to AI? Auburn, so let's use one example, self-driving car. The car as a product, the robo-taxi

1:20:10has lots of regulations. If it doesn't have enough regulations, then NHTSA ought to get involved and come up with new regulations. Car-to-car industry should have new regulations. I don't know what's missing, but if there is something missing, then I would absolutely, you know, absolutely add more regulation. In the context of internet, there are many applications that the internet powers, and those applications should have regulation. If they don't, you know, it's just, you got to find them.

1:20:42So I want to drop down the next side of the cake now to chips. And to summarize sort of where we are, because I want to make sure I do understand your position correctly, it's that these companies are going through a transition. That even as these systems speed up, become more capable, complex, persistent, whatever it might be, that there is still the limiting factor of companies will not

1:21:17ship what is not safe. They should not ship what is not safe. And you believe they have the engineering capabilities to make these things safe, to figure out the testing and the control, absent external intervention. That's sort of where you are. Absolutely. Yeah. One thing I've heard you say is that we have entered, maybe in a way people don't always understand, a new era of how computing works. Describe your vision of that and the way, if somebody's sort of understanding of it is a little bit still maybe in, you know, you've got a MacBook

1:21:50and it's got a processor in it and you buy it and how it differs.

The New Era of Computing

1:21:56The last computer industry, and the computer industry we've known for 60 years, is called retrieval-based computing. You retrieve files. That's why it's called data center. You know, file center. Okay. And in the future, it's an AI factory. It's generating. And so the amount of computation necessary to understand the context, be grounded in information, the reason about what to do, and to generate an answer, that generative process requires a lot of computation. And so just the

1:22:28amount of computation necessary per user has grown tremendously. And then the second part is because these generative AI can also be somewhat autonomous because they're agentic. Now you have agents using generative AI. And so rather than a billion people using computers, you essentially have multiple hundreds of billions of agents in addition to the humans using the computer. And so you could argue that

1:22:58the amount of computation we need, you know, however much we had before is going to go up by a billion times. And that that's a reasonable, you know, framework for a reasonable level of amount of computation. In this new world, what you really care about with it in the context of a factory is how productive is it? Not how expensive is it? It can be infinitely expensive, but you want to know how productive it is. And so our computers are incredibly productive. If $50 billion to build a

1:23:28one gigawatt data center, one gigawatt AI factory, and you can rent it for 40 to $50 billion per year. And so the productivity of it is incredible. So number one is the productivity. NVIDIA's architecture is fungible because we're general purpose, which is the reason why every AI lab, every AI model, closed model runs on NVIDIA. And because we're completely fungible, and you can use us from data processing to pre-training to post-training to eval to inference, the entire life of AI is

1:24:02supportable by our architecture. And if a customer no longer needs it, another customer would be more than happy to pick it up. And then the last part is that durability. Because our architecture is software-driven, and we're constantly improving our software with new algorithms that takes the new workloads, the new models, and run it on our old generation hardware. We have massive teams of people who are constantly doing that. As a result, the useful life of our compute is much longer.

1:24:36That's the reason why NVIDIA is, and people are talking about NVIDIA compute as an asset class, kind of like an airplane. Airplanes are general purpose. They're fungible. United Airlines doesn't use it. American Airlines will use it. It's durable. And it starts out as a passenger airplane, ends its life as a shipping, as a cargo plane. And so, as a result, it can be an asset class. And if we could do this, if this happens, then, of course, the cost of capital for funding NVIDIA

1:25:12AI factories will be the lowest. Because our computers are collateralized asset. And so, anyways, this is the phase shift that's happening to us, which is going to be a huge unlock for our growth. And so, your business has become so interesting. You've moved now into lowering the cost of capital for others in the AI industry. People maybe have seen these charts of, like, the arrows going in every direction. It's so interesting, yeah. And explain that a bit to people who are, they understand NVIDIA has become, like,

1:25:46the biggest company in the world. They see these charts that seem very circular to them. What is the difference between supporting demand, creating markets, and creating demand?

Venture Funding and Ecosystems

1:26:03We can't really create demand because, in the end, if the AI services have no offtake, then, obviously, building computers for it is pointless. And so, the first thing that has to happen, the reason why compute demand is so high right now is because AI applications are going through an inflection. They're becoming useful. And because AI is becoming useful, $500 billion of

1:26:35venture funding are coming in. And all of those companies, those thousands of companies, startups, they all need compute. And so, the demand is coming from them. And so, these companies need support in technology. They need support in ecosystem building. They need support in financial support. And so, we might decide to invest in some of them as an equity owner. And as a result, they become a really flourishing new cloud provider. Another reason we might decide is because, as I mentioned, there's a

1:27:08five-layer cake. And at the model and the application layer, there's a whole bunch of really innovative companies. And there's way more to AI than just the language model itself. There are World Foundation models, physical AI. There's biology AI. There's chemical, you know, material sciences AI. These are all different than language models. And so, many of those companies are new, and they need a lot of capital. We might decide to be a small percentage shareholder in them. So, we get them off the

1:27:42ground. They're incredible scientists. I might even, you know, by being a first investor, anchor investor, we bring confidence to their company. We, you know, we give them access to a lot of our technology. We support them a great deal. And we help them, you know, become a company as fast as possible. We might decide to invest in a nuclear company. We might decide to write so on and so forth. So, you know, if you look at my mental model of the AI industry, it's a five-layer cake, and we're investing across all of it. There might be strategic unlock points. It opens new markets. It opens a new

1:28:16route to market for us. It might secure a critical resource for us. So, there's a lot of strategic reasons why we do it. I mean, the numbers here are astonishing. You've become like a single-company industrial policy for American AI. We've put a lot of money into this ecosystem. What's the total investment you're now making per year? All in, we're probably up. Well, I don't know about every year, but I think all in, we might be like $100 billion. I might check my numbers, but it's something like that.

1:28:48It's larger than the Chips and Science Act. Oh, yeah. Yeah. Not to mention, because of the purchasing commitments that I provide to TSMC and Wishtron and Foxconn, Amcor and Spill and all these different companies, because of that commitment, I'm able to encourage them to come and manufacture here in the United States. You know, the fact of the matter is we probably contributed more to re-industrializing the United States in this chip manufacturing than just about any company in the world. We're not only

1:29:19re-industrializing manufacturing, we're doing it so fast that we're creating a shortage of labor, but we're creating a lot of jobs. And a lot of people with money in the market right now who are excited, often excited by NVIDIA stock in particular, and worry about the analogy of the internet bubble of the late 90s. And what they worry about is actually related, I think, to what you just said, which is that the internet did continue to be more useful. It's not that high valuations meant that the technology technology was hollow or fake. But something happened and popped for a minute, and very

1:29:54big companies got hammered in that, and a lot of people got hammered in that. What is there to learn from that kind of bubble bust cycle? And I guess the question is, do you not think it will happen again, or why do you not think it will happen again? At some point, demand and supply will be inverted again. And that's just the nature of, you know, markets. It's not going to happen next year. It's not going to happen in the next couple, two, three years. I just don't believe that. But at some point, we will likely have more supply than

1:30:29demand. And I just don't know when that is. And so there's not much to learn from the past. What would be the signal for you? Markets will naturally slow down, and then it will stop, you know, meaning there will be a period of digestion. Now, is that period of digestion going to be six months? Is it going to be nine months? Is it going to be a year? It won't be forever. If you look across the board, the amount of investments that we're putting into the application layer so that each one of the industries could have the technology diffuse into them so that they

1:31:04could benefit from it? That's probably one of the biggest things that we do. This is a way I often hear the sort of Chinese and American AI ecosystems compared, which is that in America, the emphasis is on the speed of rising capability. And a lot of people think we're ahead on that, and that seems true. And that in China, there's more emphasis on diffusion. And a lot of people think that China is probably ahead on diffusion, and in some ways has an economy

1:31:34that is better structured from things like WeChat all the way to just like the way knowledge and commands move through it for diffusion. And whether the race is about capabilities or diffusion, and also whether it's a race at all, but we can get to that in a minute, is a big question. I'm curious how you see that. That's the ultimate question. I believe if we want America to benefit from artificial intelligence, every single industry has to benefit. Walmart has to benefit. Safeway has to benefit. Federal Express

1:32:07has to benefit. Every bank has to benefit. Every healthcare company, every drug discovery company. We need to see every construction company, every data center company, power generation company. We need everybody in the United States. We need everybody in America. We need everybody in the world to benefit from this. And that's the highest layer. That's the most important layer. That's the layer that touches society. All the layers underneath are technology enablers. I want to see us not ruin the opportunity for the United States to benefit at the highest level. And notice all of the rhetoric and

1:32:41all the alarmism, all the doomerism, all of the predictions are scaring people. That is my greatest fear, actually. I have every confidence. Maybe I have more confidence in them than they have in themselves. You definitely have more confidence in them than they have in themselves. Well, I don't know about that. But maybe it's just too much humility and otherwise. Should we conceptualize what we're in as a race with China?

1:33:13I don't think it's necessary. Some people like to think that way. I don't find that necessarily inspires me. I have no trouble never mentioning another company when we talk about us doing our good work. And so we hold ourselves to our own standard. And so I think that different people have different ways of being motivated. And I think it takes a bit more artistry to unite and

1:33:46focus organizations to a certain level of performance outside of contests. But I don't necessarily see it as necessary, number one. Number two, the question is, even if we did frame it as a competition, it doesn't have to be that if they achieve something, it's at our peril. And so when they invent something or they create some power generation technology, it might be a great invention

1:34:22that we wish we had done ourselves. But because it's going to support all of our energy production systems here, as a result, it helps our whole industry. Maybe they came up with a great new open model. And they have. And those open models are now being used by 80% of the American startups. Yeah, we use a lot of Chinese open models here. Okay, that's right. And so that's terrific. We downloaded. It originated in China. A lot of the technology, of course, also originated from the United States. We downloaded. We make it our own.

1:34:53We fine-tune it. We put it into our own agent harness. We put it into our own sandbox. That's all your own technology. So I think the fact that you leverage their weights, I think that's terrific. That's fine. You were saying a few minutes ago the way different countries have begun to see compute as a geostrategic resource and, you know, may want to allocate it to their own companies. There's been a lot of back and forth on that here. And among people who do see us as in a race with China, particularly people see us as in a race with China for who will get to recursively improving self-superintelligence

1:35:28first. There's been this ongoing back and forth on whether or not one thing we want to do is deny them compute, which in this case tends to mean denying them your chips. Under the Biden administration, we had pretty tight export controls. Those were loosened under Donald Trump. Obviously, you wanted those to be loosened. How do you think about the question of whether or not it is good for China to have NVIDIA chips that could accelerate their models or model deployments, their model capabilities,

1:35:59versus us holding that back to try to slow their progress? In the case of AI, our goal is not just that one lab benefits. Our goal is that all of America benefits. I think the United States has a greater responsibility and a greater ambition for the world to be built on the American tech stack, just as we have greater ambition that the world is built on the U.S. dollar and that more people speak English, that they use the American version of

1:36:31the Internet. I mean, we want that. The question is, ultimately, what are we depriving? Are we depriving them a chip for their industry? Or are we depriving United States a market to compete in? If you see the market, if you see the market as big as China, how does that help the United States technology sector? Maybe it helps one company with a particular model, but the rest of the industry suffers.

1:37:04I think that it doesn't help the chip industry, surely, to be deprived of market, to go compete in. It doesn't help the rest of the industry because it deprives open models. It doesn't support the overall aspiration of the United States to have the world built on the American tech stack. And so there's a lot of things you deprive yourself if you narrowly focus on, deprive them of chips. And so I would say to take a step back and frame it into what's in the best interest of America first, all of America,

1:37:38not one company. And with respect to the race, as we mentioned, the race is, if there is one, it's about all of the economy of the United States succeeding. I found myself very conflicted on the China and chips question. And one reason is even where I have sometimes more of the superintelligence concerns than you do, is that if you have those concerns, I think you want to have a good relationship with China in which there can be kind of productive bilateral working through the risks and benefits of AI. And the more you think

1:38:12of it as a race which only one side can win and act like that, the more you are necessarily going to create enmity. And I found that to be a sort of complicated dimension of people's thinking here. You know, I think that a zero-sum strategy, I deprive you of this, therefore I win, that simplistic logic tends to have unintended consequences of the bigger game. The bigger game, of course, is that we're not all talking about safety. We want to build safe products. We want them

1:38:48to build safe products because when they don't build safe products, it hurts the whole industry. And so, this is a perfect time. We should want to look for opportunities to communicate, collaborate, to understand, align, as much as possible. Now, having said that,

1:39:08NVIDIA is an American company. We should benefit America first. America has every right. And for these technologies to be made available to the Frontier Labs, Vera Rubin goes to the Frontier Labs first. That's your advanced chip. That's right. NVIDIA has newest chips. And so did Grace Blackwell. And so did Hopper. Every single generation, so did Ampere. Every single generation of our product goes to American companies first. And if the U.S. government would like to add on top of that, that is a requirement

1:39:41to do so. I'm delighted by that. That's no problem. We do that naturally anyways. However, recognizing that the AI industry is a five-layer cake, and we want every single layer to win, then we need every single layer to go out there and compete for the market.

Energy Demands and Climate

1:39:56That drops us to the final layer of your cake, which we won't spend as much time on. But if the advantage America has had, at least at a material level, is chips and software. One of the advantages China has right now in AI is energy. That it's easier for them to build new energy. They're pumping much cheaper energy into AI. They have tremendous, they've made tremendous advances on building electrical generation and renewable energy. How do you see the, like, that most fundamental layer, the energy that pumps through

1:40:29the data centers, pumps through the chips, and where America is on generating enough of it? Probably a time when we've been trying to move from dirty energy into clean energy. Yeah, I think, I think, one, they just have a lot more energy than we do. And they plan to build a lot more than we do, than we did. We got, you know, I think we just have to acknowledge, we got ourselves really gummed up in climate change and sustainable energy. And

1:41:01as a result, we just didn't plan enough energy production. What do you mean by gummed up there? Well, in the near term, energy production requires fossil fuel. And because there's just so much, so much angst about fossil fuel energy production, if you look at our country, we've produced very little net new energy for a long time. And all of a sudden, this new industry comes along, and we find ourselves in a situation where we just don't have that much energy building capacity.

1:41:35And now, the whole country's scrambling. And meanwhile, we've moved so fast, we could have done so much better job communicating with the communities, preparing the communities, working with the communities to let them know what's coming. And if they don't want data centers to be built in their town or whatever it is, then so be it. But if you're going to build in their town, be sure to go there and let them know what's coming. Work with them, work with them to help them understand that the use

1:42:09of water is really efficient these days. The AI supercomputers are super energy efficient, but they're still going to use a lot of power. You're going to bring in your own power generation. It's going to lower their property taxes. There are a whole bunch of things that you can do. You can make your data centers more appealing. You make the setbacks further away. So there are a lot of things that you can do. You could also contribute to be a good neighbor to the community and build better schools and better community centers and improve their parks and

1:42:43improve the roads. And there's a lot of things you could do, but it's hard to do that, you know, after the fact. And now there's a fair amount of frustration around the country. And then, of course, all of our narratives about the end of the world is not helping. And, you know, what reasonable person says, come and build this data center in my town? And by the way, whatever you produce is going to, you know, end humanity as we know it. So I think, I think all of this, this, this negative doomer narrative is not helping our country. And we started off on our

1:43:18back foot. We started off on our back foot. And then now we got it. What do you mean we started off on our back foot? Oh, because we didn't have enough energy production in the first place. Well, how do you balance? I mean, there is a reality of climate change is happening. Let me just, let me just give you the one last thing. This, there's no question that the energy demand is really great, which is the reason why the market forces are helping us invest in sustainable energy like no time in history. You give me an example of a sustainable energy company, a material sciences company to build a better battery. It could be, you know, it could be solar,

1:43:53it could be nuclear, it could be fission, fusion, you name it. Hydro, you name it. Those companies are all getting funded. The market demand for energy is so incredible that this is the best time in a hundred years to improve our power grid, to make our power grid more sustainable, to lower the cost of energy, also investing in our sustainable future. There's no question that in four or five years time, we're going to use a lot more fossil fuel. But also in the next decade

1:44:27in front of us, no time in history, are we better prepared to move to sustainable energy? And because these data centers, because the cost of these data centers is so high, you know, now we're starting about talking about putting them out in space. And so, right? And so, so I think that the opportunity for us to see our dreams come true, move to a sustainable energy world, we have a better chance of doing that than ever. The world is buying more because of AI factories,

1:44:58because of AI is buying more sustainable energy today than any time in history. Venture capitals, you know, for, for our next generation energy is just incredible. Everybody, everything's getting funded. It's incredible. You don't need government subsidies for the first time in a hundred years because the market forces are here. Everybody should be leaning in. If you, if you want a future, if you want to, if you want to turn the corner on climate, climate change, if you want a future that's sustainable, lean into AI. It is the best opportunity we have to get there.

1:45:34But we need to build the energy faster to do that. That's right. I mean, that's just, that's just, there's a market for it all, but you can subsidize it and you can make it easier to build. Yeah. You know, it's kind of like, um, in order to save you, they got to hurt you first, you know, in order to, you know, that's the nature of surgery. They got to cut you open and save you, you know, they got to, they got to inflict an enormous amount of pain and suffering on you so that they could save you. And so, so I, I, I kind of think AI is kind of like that over the next several years, we have to, we have to unfortunately use renewable energy,

1:46:06use fossil fuel because we just don't have sustainable energy enough of it to make a difference. And then after that, you know, hopefully we can transition to that.

Recommended Books

1:46:15I think that's where we'll end. Always our final question. What are three books you recommend to the audience? Well, I've read a lot of books. Uh, the, the, the book that, that made a huge impact on me was, um, uh, computer architecture from Hennessy and Patterson, a quantitative approach. Uh, it was the first computer architecture book that, that reduced the complexity, the abstract idea of computer architecture, uh, down to engineering. And I, I love it when, when people take complicated

1:46:46concepts and reduce it into something that you could do something about. Number two, I really loved, uh, Innovator's Dilemma. Uh, Clayton's past, but, but, um, Clayton Christensen's book on how, how industries evolve over time and how to see, uh, emerging technology and how to set it proper expectations about it and how to extrapolate, um, maybe its future impact. Uh, I really loved Al Reese's and Jack Trout's, um, book on positioning. Uh, it, it's a really wonderful book about how

1:47:24people see, it's, it's a book about marketing strategy. Um, more than that, actually, it's just, it's a book about strategy and, and how people see the world and how people see products and how you present products and how you, how, how you see your own strategies. And, and I thought that was a really thoughtful book and, um, a really easy to, really easy to understand. Jensen Huang, thank you very much. Thank you very much, Ezra. I always enjoy, uh, our time together and today was a great time.

1:47:54Thank you very much.