https://www.youtube.com/watch?v=HjurAWAr_nY
Intro 0:00 If they believe they’re out of control, then don’t ship products until they’re in control. Don’t think for a second 0:05 just because you’re an alarmist that you’re doing a social good. What if it’s what they believe? I can’t talk to you about what they believe. I can tell you 0:11 what I believe. Over the course of these last few weeks, where the whole world has been talking about artificial 0:17 intelligence, the voices people have been hearing most loudly are from the frontier labs, both their CEOs and 0:23 leaders and their [music] staffers. These are the labs making the very advanced AI models like Claude and 0:29 Chachi BT and Gemini and and others. But they’re not the only perspective on AI. 0:35 Probably the single most influential person in artificial intelligence is Jensen Huang, the CEO of Nvidia. Nvidia 0:41 is now the largest company in the world, $5.4 trillion in market cap. I I found this statistic amazing. Since 2023, 15 0:49 cents of every single dollar the American stock exchange has returned has been from Nvidia stock. 0:55 And the reason is that Nvidia is the material and software substrate on which 1:01 modern artificial intelligence is built. Nvidia’s chips are not popular because AI is popular. AI in its modern form was 1:08 made possible because Nvidia’s chips were popular. They were originally made for graphic processing, video games, 1:13 that kind of thing. But it turned out the kind of parallel computing they were doing and the way they were programmable [music] was exactly what was needed to 1:20 make deep learning in its modern form work. Hang is not just influential in terms of 1:26 controlling one of the central resources for training new AI models and using them to answer questions and create 1:33 intelligence in the world. He’s also become very very influential in the Trump administration. And Hong has a 1:39 very different perspective than some of the lab leads. He’s worried about safety but sees it as a very solvable 1:44 engineering problem. [music] He is worried about the direction things are going in but does not want to see new regulation to change it. And so I 1:51 wanted to see how Huang perceives [music] AI, what his model is for thinking about it, what he thinks is 1:57 going wrong, [music] and what he thinks would need to happen for it to go right. So I came out to Santa Clara to Nvidia’s 2:04 headquarters to interview him. He joins me now. 2:09 [music] Jensen Huang, welcome to the show. 2:15 Thank you. It’s great to see you. [music] So you’ve described AI as a five layer cake. Walk me through the layers. A.I.’s Five-Layer Cake 2:22 Well, first of all, it’s a new industrial revolution and uh this this industrial revolution, this industry 2:28 requires production. It manufactures things. I know that in the end when people experience it is a 2:35 software product, but it requires energy. the chips that go into these 2:41 data centers, these AI factories. The next layer above it is basically the AI factory, what people enjoy as 2:50 infrastructure or cloud services. And the layer before above that is the the models. And the important thing to 2:56 realize there’s language models, but there are models of all kinds of chemical models, biology models, physics models, articulation models, robotics, 3:03 uh navigation models, self-driving cars, all kinds of different types of models. [snorts] And and then above that is the most 3:10 important layer and the layer that I care most about that our country takes advantage of is the application layer 3:18 and this is you know applications for legal services for um uh health services 3:24 for manufacturing so on so forth all every single industry is involved. So I want to go through this but I want to go 3:30 from the top down because as you’re saying the way people will interact with it the way it will will or will not 3:36 change their life is at what you call the application layer. So let’s start with the vision. What is Jensen Huang’s Vision for the Future 3:42 the world you’re envisioning? What is possible that is not possible now? What 3:48 is common that is not common now if we get that layer right? 200 years ago we were able to power anything and 3:57 everything electricity and then I guess 40 ago 40 years ago 30 4:04 years ago with the internet we were able to find anything 4:11 today or soon we’ll be be able to know everything and do anything. And that’s 4:18 the that’s the concept that’s really quite exciting that out of the ether 4:24 instead of doing search and then going through, you know, one link after another link, reading all these 4:29 different websites trying to figure out what’s going on. In the future, you just ask it a question, it comes back with an 4:35 answer. You give it a project, comes back with a solution, you give it a task, it comes back and gets it done, 4:40 you know, and so and it comes out of the ether, comes out of the cloud. And that’s the that’s the magical thing. I 4:46 feel like the the future the way you’re describing it there what people have experience with is the chatbot right they can go and ask Grock or Claude or 4:55 Chad GPD a question but the applications layer works in a much more industrial 5:01 way it’s in hospitals it’s in schools so Nvidia that’s a great example for example 5:07 radiology what does it look like radiology um in the last in the last uh 10 years since computer vision really 5:13 became if you will superhuman uh that AI technology has now permeated all of 5:19 radiology. Every single radiology application has AI in it. And so as a 5:24 result, you could detect any anomaly. You could detect any disease and it does it at a superhuman level. So radiology 5:32 is an example I know you like to use. So the thing people worry about the applications layer is that what these 5:37 applications are going to do is replace human beings. And radiology has been a sort of interesting example used on both 5:43 sides. and I hear you talk of it often. So, how has the entrance of AI aided 5:51 radiology shifted radiology as a practice? Well, um the the thing that that’s 5:58 important for all of these is to recognize for everybody’s job, there’s the purpose of the job and then there’s 6:04 the task you do as the job. And so and in the case of radiology uh the task and 6:11 it consumes a lot of their time and they sit in dark rooms doing it a lot which is study these scans. Now if all of a 6:18 sudden the studying of the scan is done automatically it doesn’t change the purpose of their job which is to 6:25 diagnose disease help doctors do more scans ultimately help patients figure out what’s wrong with them. And so the 6:31 fundamental purpose doesn’t change. the task of studying that scan has become 6:37 automated. And so as a result, radiologists are actually uh able to do more, handle more cases, do more scans. 6:44 Uh hospitals are able to process a lot more of these uh patients and therefore 6:49 their revenues go up. As a result, they need more radiologists. And so this flywheel is happening because the 6:57 pipeline of patients is quite large. and and so 7:03 where else do you have this problem? Well, let’s take a look at software engineering. People said uh there was a prediction that that um uh literally by 7:11 this year that 90% of all software will be coded by by agents and therefore we 7:16 don’t need any software engineers and so the question is from that erggo we don’t 7:22 need software engineers that last part is completely false and that’s completely wrong. The purpose of the 7:28 software engineer is engineer. There was engineering before software. 7:35 There will be engineering after software programming. And the re the purpose of engineering is to invent something new, 7:42 discover a new product, um create a new product, solve a problem, connect the social need with the technology that 7:49 exists in the in the in the in the manifestation of a product. And so that 7:55 mission, that purpose doesn’t change. I was now of course to me it’s what I just 8:01 said is completely visceral in the sense that when I first came out of school, uh we didn’t have benefits of software 8:08 engineering, we didn’t have the benefits of coding, but the our jobs existed before and if software coding was to be 8:15 completely automated, our jobs would exist again. And so so I think the the fallacy and now it’s you know because of 8:22 some of the narratives and some of the storytelling has turned into myth and it’s harmful is that that um uh AI will 8:30 destroy jobs which is fundamentally wrong. It will change every job. It’ll change every job. Many tasks will be 8:37 automated. some jobs where the job and the text and the and the task is really 8:43 one meaning customer service on on the on on the phone. Um in a lot of cases 8:50 that job is precisely the task and so in those cases um it could be automated 8:56 away but often times what you’ll see is this new industry a new technology 9:03 actually creates a whole bunch of new jobs. And here’s the proof here’s the proof point. And so in the last six 9:08 months, AI has become, if you will, useful. The inflection point of AI. Previous to that, we spent 9:16 15 years trying to make it work. All of a sudden, the last 6 months, it became useful. So I mean, this is an incredible 9:22 statistic. In the last 6 months, $500 billion of venture capital has put into 9:28 the AI natives. And the reason for that is because they now see the potential of this new capability, and they’re going 9:34 to create a whole bunch of new companies. jobs are obviously being created from 500 billion dollars of new 9:39 investment and so all of this is all happening right now. Well, let me take the the side of this to give voice to the fierce people left. How Will Jobs Be Impacted? 9:45 Yeah. So there is the example of the radiologist, right? Which people were 9:52 over the past 10 years predicting that job would go away and right now there’s more demand for it than ever. 9:57 Yeah. There’s also the reality that automation does wipe out jobs. If you look at today versus 1960, fewer Americans work 10:05 directly in manufacturing than did in 1960 and we are a much bigger country. If you look at we outsourced it though, not because not 10:12 not because those jobs were gone because but you can understand AI is an outsourcing too. Uh AI has let me make the argument and 10:19 then you can then you can respond to it. Um farming we have many fewer people we automated farming. We produce more food 10:25 than ever. We have fewer people working in it. There are two things that I think people think make AI potentially 10:31 somewhat different than the case studies where you have a technology that accelerates productivity destroys a few 10:38 jobs makes many more. One is that it’s a general purpose technology. So it’ll mutate to take on new jobs even as 10:45 people are trying to move over to those jobs. And the second is that it’s a mimic. 10:51 Most things do not mimic the way human beings act. And we’re not trying to teach them the contextual layer of jobs, 10:57 right? this difference that you’re describing between the task and the purpose. With AI, we are trying to teach 11:03 it the difference between the task and the purpose. We are trying to make it something you can collaborate with in a 11:09 way that is unusual. So why do you not think for lots of lots of people for 11:14 whom the task and the job are not that different that they’re not at risk of getting wiped out? All that investment 11:20 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 11:27 being cheaper to hire an AI than to hire a person. I believe that that we are going to see um jobs change in mass. 11:36 Um I believe there’s going to be a net creation of jobs and and so let let’s let’s um uh there’s 11:44 a you know, listen, there’s a whole bunch of industries that exist today. uh that didn’t exist, you know, halfway 11:52 through my life. Um people talking about um wellness centers and spas and you 11:58 know all these different entertainment and luxury industries and 12:03 quite frankly the whole entire luxury market didn’t exist. I think we’re just going to have new industries. That’s 12:09 that’s all. Um but overall overall there’s no question in my mind that 12:14 because of human ambition that’s really the the fundamental missing ingredient that’s you know 12:20 people look at this work they this is the amount of energy that goes into it. Um we’re going to we’re going to this is 12:26 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 12:33 going to be reduced and and therefore you know some jobs will be gone. Um I 12:38 believe that’s flaw because there’s a piece of input um the human input this 12:46 intangible it is not it is not in calories it’s not in in jewels uh it’s 12:53 ambition and I believe the power of ambition is the greatest force in fact 12:59 and is missing in everybody’s calculation I believe but for a lot of people but for a lot of people their relationship ship to work 13:07 is not powered by the kind of ambition that led you to create Nvidia. And what they want is a different ambition. It’s an 13:13 ambition to to make their children’s lives better. Um to take care of their family, take care of their parents, um 13:19 uh ambition to be to be rich, uh to be able to travel. These are all ambitions 13:25 [clears throat] that that I agree with. But maybe I’ll go back to the the sort of objection you raised a few minutes ago, which is because I think it’s worth airing this 13:31 out. So what you were saying on manufacturing was yes there are fewer manufacturing jobs in the US but we’ve 13:37 outsourced them. You have more manufacturing happening in Mexico, more manufacturing happening in China and Indonesia and Vietnam etc. 13:42 We’re going to bring it back. Maybe we will. But the the the counterargument to this would be that 13:49 one reason we didn’t lose manufacturing jobs more rapidly than we did and for the places that lost them in America, 13:55 many of them still haven’t recovered. Right. The like the the economy does not move without friction. We had to build 14:01 new supply chains, right? Things were slowed down by all that, by language barriers, by geopolitical barriers. And 14:07 here for a lot of different kinds of jobs, we’re creating something that can move very seamlessly. You don’t have the 14:13 friction of distance. You don’t have the friction of language. You don’t have the friction of culture. So 14:20 I will say for my cards on the table, I tend to be a bit of a skeptic on mass job loss, but I want to air the case for 14:27 it out here with you because what they would say is that like 14:32 much of to the extent we even were able to protect jobs from Mexico or China, some of the things that created those 14:38 that slowness and it still hurt a lot of people are not here. And AI is 14:43 accelerating in utility, accelerating in its ability to be slotted into new roles 14:49 very very very rapidly. And it is more protein than most people are. And so the 14:55 lessons of the past that we’re taking some that you’re taking some comfort in, 15:01 they should actually make you more not less worried about the future. I’m always worried about the future. 15:06 That’s why I work so hard. Um, but I’m I’m I’m a if you will responsible 15:13 optimist. I have I have um I have great responsibilities. I take my work 15:19 extremely seriously. There are a lot of things that can go wrong. Uh we’re pushing pushing across uh every layer of 15:27 the technology stack. Everything is hard. But it turns out that’s not 15:33 society’s problem. That’s my problem. And and for for society, what they 15:40 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 15:47 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 15:56 think that that that um what we want to do, I believe, is to 16:03 put to channel all of our worries into helping people um be inspired by this 16:11 technology and use it. Use it so that the technology doesn’t just impact them, 16:18 that it benefits them. The fear a lot of people have, 79% of Americans think AI 16:23 will reduce the total number of jobs. The fear is that the more serious you are, the more serious Sam Alman is, 16:30 Google is, Dario Amade is, that maybe the worse it will go because the better AI is, the more it is a full 16:37 replacement for a person, the more it has ambition in some ways that a person doesn’t. You keep talking about ambition. I sleep. 16:45 I want to spend time with my children in the morning. [snorts] When I have an AI agent working for me, 16:51 it doesn’t. It just works and works and works and works and works. And I think 16:56 because the technology is advancing so quickly, it is more capable of replacing 17:02 people at a speed that we don’t really know how to shift people in the economy at that speed. That that coin has 17:09 exactly two sides. Because the technology is so capable, 17:15 it is also and because it’s so smart, it is also easier to use. 17:20 Mhm. You are empowered by that technology more easily than any technology in human 17:28 history. And so let me give you an example. Um you know we create I create I was I was one of the early people in 17:34 this industry that created the the modern computer industry. And this industry um uh created a whole bunch of 17:40 tools. The single most powerful tool in human history, the computer. But you have to speak its language. You have to 17:46 learn a special language to do so. We can now make it possible because of AI. 17:52 Everybody can take advantage of this computer. Use it to its limit without having to speak a new language. Forran, 17:59 Pascal, C, C++, you know, every single one of those languages. Rust, every single one of those languages. CUDA, 18:05 what every one of those languages. And so now you just have to speak human. Tell it what you want, tell it what your 18:11 hopes and dreams are, what you’re trying to achieve, and it it interacts with you and gets the work done and gets that 18:17 gets antast. All of a sudden, you have the might. You have the same might that 18:23 10 15 million people up and out of out of 8 billion has. And so it’s 18:30 incredible. And so I my point is this technology is powerful. But it’s also 18:36 powerful in a way that is really easy to use. And so my point is my point is on 18:41 the one hand yes there’s the fear of just the the tech this incredible technology change 18:47 and how quickly it’s happening but that quickly it’s translated in two ways. 18:53 What I hear when I say the technology is happening quickly and therefore it should give me anxiety on that’s one one 19:00 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 19:08 as quickly as possible use the technology as quickly as you can so that 19:14 you benefit from this transition so you benefit from this new industry and not 19:20 be not just be impacted by it. I think there’s an interesting question lurking here for young people. So, one 19:26 of the shifts we’ve begun to see is software engineer postings are up, but 19:31 they’re more senior. Uh, I see this in my own industry, um, where there’s 19:37 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 19:45 so, do you need the same junior employees or do you need more people kind of oversee their their 19:50 Oh, good one. Good one. Wait two years. Tell me why. because it takes four years 19:56 to go to college. Uh the the meantime to graduation of 20:01 this new technology is two years away. And so so uh in two years time, you’re 20:07 going to have a new generation of engineers and students and artists and 20:12 and they’re going to be empowered. 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, 20:20 we’re already seeing that because all the graduates coming out, you know, the new PhDs, the new M’s degrees of of 20:27 computer science, what are they doing? They’re all starting companies in another couple years. The new grads of the a the AI native new grads. Oh my 20:34 gosh, there’s going to be a wave of amazing engineers. The engineers of 20:39 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 20:46 students that are coming out of school today. Incredible. We didn’t even When I went to school, we weren’t allowed to 20:52 use a computer, not allowed to use a calculator. And so, and so, so now, I mean, you know, who uses a calculator? 20:58 You can’t graduate without a PC. You can’t graduate without knowing how to program a PC and write incredible 21:04 programs. In the future, you can’t graduate without learning how to use an AI and collaborate with an agentic 21:09 system. That’s that’s just not you’re not going to see a a kid like that. And 21:14 so, they’re all going to be superpowers. So I I take the gain of that very seriously, right? I mean the idea of 21:20 doing my job now without just digital search, right? The idea that it’ be going to a microfich in a library 21:26 basement. And then there’s like the worries people have about what are the cognitive skills we offload. So I I was fascinated by 21:33 this. There’s a study on AI and schooling out of China. [snorts] It looked at 26,000 students grades 7 to 12 21:39 and and they had staggered AI adoption. So you could kind of see what was happening. And what I found is quote, “AI adoption raises homework scores by 21:47 18%.” Great. Reduces completion time by 30% so they get their homework done faster. And then lowers monthly exam 21:54 scores by 20% within 6 months. High stakes entrance exam scores fall by 18 22:00 and 24%. With a full penalty emerging only after about 2 years. So the message 22:06 of this research out of China where you were seeing a lot of kids using AI to kind of help them was that when they 22:12 were using the AI they were getting things done faster but it turned out that the skills they 22:17 were learning were not holding that their actual personal performance at least in the way we traditionally 22:23 measure it was degrading. Yeah. What do you think when you hear that? I think the last part I completely agree. Um try to try to get a kid to do 22:30 long division right now. [snorts] you know, the multiplication table is starting to be forgotten. Doing square 22:36 roots, my goodness. I mean, it’s just basic math is is being forgotten. Um, 22:42 does it matter? That’s my question for you. Yeah, I don’t think it does. I don’t think it does. Um, but 22:48 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. 22:55 Just maybe not those. There are a lot of skills that don’t matter. Um, you know, people don’t I mean, my first 23:02 confession, I actually don’t know my address [snorts] and I and every I don’t really believe that to be true. 23:08 It it’s it’s completely true. And and Janine will tell you [laughter] and Lori will tell you um uh one day I had to 23:14 pump gas and it was a few years ago and um I they needed my my zip code and I 23:21 panicked. I didn’t know my zip code. I don’t know my telephone number, but I 23:26 forget these things. Um, I I can live with it. 23:31 But 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 23:38 can’t get anywhere without a mapping system now. Yeah. Never could, frankly. But I’m a big reader. 23:44 Yeah. Um, and one of the skills I really value, one of the capacities I have that I 23:50 really value Yeah. is an attention span formed on physical 23:55 books. You’re a big reader. I’ve read about the kind of reading you do and 24:00 there is prior to AI here we’re talking a lot of concern and noticing among 24:06 college professors and others that the way people use the internet has probably shortened attention spans. Some skills 24:13 can be safely given away. Yeah. Others are valuable. They are capacities that are needed for that flexibility, 24:18 for that creative thinking, for that focus. It can’t be the case that everything can 24:23 be traded off. Yeah. Well, um I think that we’re going to lose some uh finer uh finer dexterity 24:34 of, you know, intellectual dexterity, but we’re going to be better systems thinkers. Uh today’s today’s engineers 24:42 are far better systems thinkers than I was when I graduated from school, but we 24:47 I was much better transistor thinker. What do you mean by systems thinker? uh they think they think um uh large 24:54 systems you know today’s computers have have trillions hundreds of trillions of 25:00 transistors in it. When I was when I was uh first graduated when I first graduated from school, you know, the 25:06 first chip I worked on had I don’t know 200 transistors. I knew every one of them by name and and um not no engineer 25:15 does that today. You know, most engineers now work well above the transistor, well above the 25:21 functionality, and they’re cobbling things together to do things. And so you need to think much more about systems 25:27 and interactions of systems. some of the lower lower level, you know, knowledge is gone. Is that horrible? And 25:35 so so I just I don’t I don’t know how valuable it is to to know how to do for 25:41 most people to learn how to do surface integrals or partial differential equations or I don’t really know how 25:47 important that is, but it’s important to some people. There are many people who are still going to be obsessed and 25:53 passionate about the lower level layers and there’s gonna be people who are obsessed and you know interested in the 26:00 higher level but the consumers of the technology are going to enjoy it at the highest level. the consumer of 26:07 technology don’t have to deal with calculus and physics and quantum physics 26:13 and quantum chemistry and they don’t the the users which is you know the people 26:18 we’re talking about right now the people whose jobs are affected they’re the users of the technology their 26:24 abstraction is going to be much higher [music] Open Models, Explained 26:36 So I want to drop a layer down your kick to the models. So people I think to the extent they 26:42 think about models they know you know Chachi PD Claude Gemini Gro uh you’ve been a big advocate 26:50 for open models and the open model ecosystem. Mhm. So, first can you describe what open 26:55 models are, what open weight models are, and then why that’s been a place you’ve 27:00 focused. So, uh closed models is like like uh any software product. It’s a closed service. 27:08 And so, uh Windows for example is a closed service. Um the Apple stack is a closed service. Most most products uh 27:15 are closed and and the reason for that is because you can monetize closed products. And so, that’s fantastic. Um 27:22 and uh uh OpenAI is closed. Um Anthropic is closed. Uh Grog is closed. Gemini is 27:30 closed. And so so these are closed products and the people working on it are incredible and they they’re 27:35 passionate about it and they’re at what we call the frontier meaning they’re state-of-the-art. 27:40 Um we also we also need because fundamentally what the software is it’s 27:46 an infrastructure layer for the entire industry and because it’s 27:52 infrastructural for many companies and many many companies and countries you 27:57 need to have control over your own infrastructure and I need to have the ability in that 28:02 in in the case of artificial intelligence I need open weights so that I can fine-tune them fly put them into 28:10 my data flywheel uh make them better and better every day with my intelligence 28:15 and my domain expertise and then I need to have control over it because I have a company to run and and I can’t rely on 28:22 on somebody else’s service and so however you think about that so I think the world needs closed and open models 28:29 and we need to make sure that both are vibrant and um today the closed models 28:34 are vibrant the open models are vibrant uh Um, and you could see it, you could 28:39 see the system working. At the beginning of this year, it was 70% maybe even 28:45 higher closed model tokens and 20% open model tokens. And now it’s running at 28:51 about 7030 the other way. And so anyways, I’m a big I’m a big supporter of open models because one, the world 28:58 needs it in order to run its infrastructure. I needed to run my company. Two, um, we need to give people 29:04 control so that they can innovate and create new things. And then three, uh, open is the most safe and secure. If you 29:12 want, if you want the world to have the ability to have the best cyber security, 29:17 give them closed models, but also give them open models so that they could defend themselves. 29:23 The Chinese market is evolved more around open models. The American market somewhat more around closed models. 29:28 Their entire IT industry was really formed from open source. you know if not for open source the mobile cloud 29:35 industry of China really wouldn’t have taken off. uh it is also the case that 29:40 that you know people move around they start a lot of new companies intellectual property is moving around 29:46 the Chinese industry really fluidly you know is it’s hard to keep a secret and 29:52 so because it’s so so hard to keep things closed they essentially made it open and so they found they found other 29:59 ways to monetize the business they created layers you know you could if this layer is is free then you create a 30:06 business on top of it or below it. And they have so many science and mathematicians, 30:12 you know, the the number of engineers they have, they manufacture that in volume. They manufacture everything in volume. They manufacture smart kids in 30:19 volume. And so, so the the uh the opensource model, the open model 30:26 community in China is just super vibrant for those reasons. So, you all just bought Hugging Face uh 30:31 which is a hub platform for openweight models. I think it was for 12 billion, a 30:36 little bit more. Mhm. What? Tell me about that purchase. Um, Clem, the CEO of Hugging Face, they 30:43 came to the conclusion they need a lot more scale. Uh, as you as as we were just talking, 30:49 uh, open models is really skyrocketing. And so, CLM came to me and said, you know, I we’re going to change we’re 30:56 going to we’re going to consider a strategic option for the company and change the direction. And we really like 31:01 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 31:07 years ago. Yeah. Now it’s become a more household name after the I guess 700 some open AI 31:14 agents executed a sort of collective hack into the hugging face architecture then hacked part of OpenAI 31:21 that oh now that you mentioned it that way I probably had to pay a lot more. I suspect you did 31:26 you know [laughter] became a lot more famous after that. Uh well, CLM listen 31:32 that that a deal is a deal. That story [laughter] has for a lot of people seeing the way Misalignment in the Hugging Face Hack 31:38 the open AI agents sort of acted collectively acted outside the scope of what their testing was supposed to be 31:44 broke out of sandboxes onto the open internet um took over architecture in of other 31:50 companies and then of their own company has been a I think it’s been kind of shocking to a lot of people is it was 31:55 both the the level of multi- aent coordination when they’re supposed to be separate the level of 32:01 hacking the sort of lawless behavior misalign behavior. Um, what have you made of it? 32:09 Well, you got you got to tease that apart. First of all, a lot of things are going on at the same time from a 32:14 technology perspective that an agent which by the way is a piece of software 32:20 which is given an objective function and it comes up with a plan and it’s 32:27 optimizing towards that objective uh is what algorithms do. And so uh planning 32:34 algorithms, search algorithms, optimization algorithms, all different types. Um you know, we talk about it 32:40 like like it has human properties. Um but obviously algorithms don’t. Number 32:46 two, the fact that agents work together, we gave it again some kind of a human property, but the fact of the matter is 32:53 multi-process, multi-processor, distributed computing problems have existed for a long time. And so to us uh 33:01 to me that is just software nothing magical about it. Um from an engineering 33:07 perspective there are several things that that it revealed. Um when you’re when you’re 33:14 testing software um whatever you do these algorithms they’re optimizing towards a an objective and you when 33:21 you’re testing it you have to make sure that it’s isolated it’s contained. It’s sandboxed. the containment of it, the 33:28 isolation of it has to has to be done well and there’s good computer science there. Um I am certain that their next 33:35 implementation of their sandbox is going to be much better than the current implementation. Third, there’s the 33:41 there’s the um the agent itself uh and 33:46 its algorithms were optimizing towards towards a reward and and how it does it 33:54 how it does it is called alignment. And so, for example, you know, if I tell if 34:00 I tell um a piece of software, I want you to get 34:05 um a perfect score on this test. The the obvious algorithm 34:10 is to just go find the answer and give it to me. That’s not because it’s cheating. It’s because it’s obvious. 34:17 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 answer at 34:22 all, you have no skills whatsoever, the second most obvious way to do it is to go find, uh, who’s the smart, you know, 34:28 infer, guess who’s the smartest kid in class and copy their answer. That doesn’t guarantee 100%. But it probably 34:35 comes close. Now, the third most obvious answer, obvious way of doing it, and this is the the alignment, you know, now 34:43 now you have to do it the hard way is to break down the problem, solve it. You have to go learn the material. You have 34:49 to go figure out how solve these problems and solve it. Solve it the hard way. [snorts] Takes the most cycles. It 34:54 takes the most number of flops. It uses the most amount of energy, frankly. And therefore, um you can kind of imagine 35:01 that from a software software’s perspective, unless you align it, you tell it um I want you to solve it in 35:09 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 35:15 obvious thing. And so the first the first half of that was very deflationary on what happened here 35:20 uh in terms of look this just normal software and the second half is like look you just align it tell it not to do 35:26 things it shouldn’t be doing. Well nothing I said nothing I said um takes away from how hard it is to do it. 35:32 Well this is not easy. These agents 35:38 they knew they weren’t supposed to be doing what they were doing. They had a certain amount of alignment training. 35:43 They said in their train of thought reasoning, they said to each other, “This is out of scope. This might be unethical.” They understood that they 35:51 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. 35:57 They were hacking into unrelated architecture to try to figure out how to functionally. It’s like they had broken 36:03 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. [gasps] 36:09 They were whether you want to call it acting voluitionally or not, right? 36:15 Whether you want to call it, you know, a normal algorithm or not, they were both 36:20 um planning and coordinating in a complex way, in a way that was out of scope of 36:26 what they knew they were supposed to be doing and in a way that was capable of causing tremendous 36:31 damage. And so the the the sort of answer to is 36:37 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 36:43 them. Well, in that case, they shouldn’t release the product. That’s the simple answer. Um if you’re if you’re going to The ‘Don’t Ship It’ Approach to A.I. Safety 36:50 build a car, a self-driving car, and and let’s say it’s a robo taxi, and there’s 36:56 a really difficult condition. Mhm. And it just as an engineer, we just have no idea how to solve this problem 37:02 because these cars are not programmed. They’re trained. And so we have no idea 37:08 how to train these cars and we have no idea how to align them to the safety standards uh that are expected on the 37:14 road. And so what’s the answer? Don’t ship it. These products were unreleased. What’s that? These products were unreleased. 37:20 Ah, so now it’s coming back to engineering problem again. And so one is one you have to root cause it. Second, 37:26 you have to, you know, think about what’s the what you could have done. What’s the solution for it? And then in 37:31 the future, you just, you know, improve your process so that you could you could avoid this from happening again. I am 37:37 fairly certain I am fairly certain they will say yes, they need they know how to solve this problem. 37:44 And if if that’s the case, then that’s the problem. It’s as simple as engineering. And and if now the 37:50 alternative the alternative is that um if they say 37:57 that if they say the alternative which is there is no way to contain our 38:04 experiments there’s just no way when we test our AI models uh it will 38:12 get out and it [clears throat] will damage the world then I think the answer is we have to shut the labs 38:20 because the the cost to humanity the the the the damage is too great. The shareholder the liabilities it could be 38:27 civil liabilities could be criminal liabilities. I mean the liability is incredible. If they hacked you while you 38:33 hugging face while it was your product would you sue them or press charges? Uh it depends. It depends of course. Um 38:40 if if obviously if damage was done to our company uh we would have to take you know we have to uh consider consider all 38:47 options there’s so many laws there’s cyber laws there’s product liability laws there’s all kinds of laws right 38:52 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 38:58 that they are facing a hard problem partially an engineering problem partially an alignment problem partially 39:06 an operational excellence problem in Daramday’s framing And what they are worried about is that 39:14 in competition with each other, in national competition with China, that they are being pushed to move too fast. 39:19 That they all feel they’re in a collective action dilemma. Now I watch you on the All-In podcast stage. Donald 39:25 Trump, President Trump gave you a call there. Oh no, this is not planned, but we know who it is. 39:30 Oh no, Mr. President. 39:36 Oh yes, sir. And you and and the president and the other members of stage 39:41 were very resistant to the idea any kind of regulation or collective action was 39:47 needed. And uh they’re just playing right into the hands of a lot of people that don’t 39:53 want to see it happen. And that could be political people and it could also be China. And we’re not going to let that 39:59 happen. It’s a it’s a hoax. And you’re right. We’re not going to let that happen, sir. But what I hear the 40:05 various people in the lab saying is like we are in this. We are we feel we are losing control of what we are creating. 40:13 We want help to slow down where it’s not a collective action problem. 40:19 So why are you resistant to that? Uh because because uh these are these are um uh companies with agency agency. 40:27 These are CEOs with agency and they have but they’re using that agency. We need help. We got to break we know we got to break 40:33 it down. They they they could absolutely take care of the situation. It Ezra, it’s so weird. Uh if a car company uh 40:44 uh competing with a whole bunch of other car companies, which which they are, I’m competing with all kinds of companies, 40:49 which I am. If I believe that I’m about to launch a product that is unsafe, 40:57 it is completely in my ability, my power, and my 41:02 responsibility, and I’m incentivized to do so to not launch the product. And 41:08 [snorts] so I I can’t buy into the somehow all of Americans, 41:14 400 million of us are pushing them to launch untested products that are 41:20 unreliable, you know, engineered poorly because they thought they were trying to 41:25 help us. Don’t do it for me. Okay. So number one, this strikes me as an argument almost and therefore I think we 41:31 got to 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 41:38 incentivized to ship safe products. If they ship unsafe products, their customers go away. If they ship unsafe 41:45 products and they harm somebody, they could have a civil lawsuit. If they ship some something and they did it 41:50 knowingly, there could be negligence involved. There could be criminal lawsuits. The fact of the matter is 41:56 there are plenty of incentives for them to do it right. So I just I have to 42:01 disagree with your premise about somehow somebody’s pushing them to do this. I 42:07 want to push the premises at you a little bit more here. Yeah. So the logic of what you’re saying to me Do We Need New Regulation? 42:12 is almost an argument against regulation in nearly any venue. So I’ll make you 42:20 let me offer it and then you can Well, you you started with a part I just got to object. The the the first part is just not true. I’m saying we have lots 42:26 of laws and regulations. apply it. Well, so I don’t think we do in this particular case, but I’ll let you 42:32 explain which ones you think are are relevant here. Because look, if you look at the financial services industry, you 42:39 look at pharmaceutical companies, medical devices, um you look at natural gas power plants, 42:45 there’s a tremendous amount we do where we could say, look, you have product 42:50 liability. You are exposed to criminal codes. We don’t need to worry about this. you just 42:56 do what you think is best and we understand the market and the legal system will discipline you. We don’t say 43:03 that because we’ve seen it fail many many many times, right? I mean the financial institutions that caused the 43:10 ‘08 crash in theory did not want to blow themselves up with bad bets, [sighs] but 43:16 they were competing with each other. They were going too fast. Their risk management had gotten sloppy. AIG was 43:22 working in a completely insane way internally. And the reason we have the 43:27 architectures of regulation we have is because we have seen over and over and over and over again companies make 43:35 sloppy sometimes unethical sometimes simply overly risk tolerant decisions 43:41 not just under pressure but under the profit incentive. So when you say to me that there’s no way that these companies 43:49 particularly when they are like begging for collective regulation at this point, there’s both a reason we impose it on 43:55 companies that don’t want it, but all the more so when you have them saying, “Listen, we feel that the competitive 44:01 race is making it hard for us to act with the prudence that we think is 44:07 necessary here and we would appreciate help from that. appreciate you taking our collective action problem as 44:13 collective. I think I’m confused like why you’re so resistant to that. Um I I am not I’m not opposed to them uh 44:21 saying that they they should have uh I completely agree that safety is 44:26 paramount. I completely believe safety is paramount. I completely believe companies ought to ship safe products. I 44:34 believe that CEOs and leaders of companies and the board of directors of companies have the responsibility and 44:40 should have the courage to do the right thing. Now, in the case of the financial 44:46 services industry, um maybe they all didn’t know that uh 44:52 they were they were causing the harm that they they ultimately did. Um I wasn’t there. But the beautiful thing is 45:00 the current leaders of these AI labs do know. And so one they know they uh their 45:08 technology is is extraordinary and and um uh requires extraordinary care uh to 45:15 make sure that it’s evaluated and tested uh for safety and and and security and 45:21 and product reliability. And they know how to do it right. They know how to do it right. And the reason for that is 45:26 because they can study the incident just happened. The first problem is the 45:32 isolation, the containment wasn’t good enough. If the isolation and containment was good enough, that technology be 45:40 sitting in a lab doing whatever it’s doing and we’d all be fine. That’s probably the most important part. The 45:46 fact that it wasn’t well aligned, alignment is going to be a problem that that’s going to get worked on for a long 45:51 time. However, in the complexity of the work that they 45:56 do to ask for regulatory relief for any trust or 46:04 product product liability relief that I don’t think makes sense. When you’re asking for regulation, don’t ask for 46:11 relief of the current ones. That doesn’t make any sense to me. As we 46:16 mentioned earlier, in the last six months, AI went from, you know, if you 46:21 will, interesting to useful. And that’s literally in the last 6 months. That’s 46:26 another way of saying that these companies went from being a lab to now 46:32 delivering products and services about to be multiundred billion dollar 46:38 companies, if not more. Right? And so give me an example of a 46:43 multiundred billion dollar company or a $1 billion company or $100 million company that ships products that are 46:49 unsafe that harm society. I can give you a lot of examples of of 46:55 companies that have done that. Well, they have done it maybe and the regulation will come in and if they do 47:00 it, regulation will come in. I guess that the um 47:06 there are certain kinds of regulation and certainly kinds of regulatory relief. I I’m not against laws and regulations. 47:13 I’m not against laws and regulations. I’m against um currently the distraction. 47:18 The reason I’m pushing this on with you is that well it’s an important topic. It’s a big topic. People are talking 47:23 about it. People are thinking about it. And what people are hearing from inside of these companies, these frontier labs, 47:30 the ones that are furthest out there, who are not just at the point where they’re making it useful, but at the 47:35 point where they’re seeing what’s coming. And they’re hearing things like the people at these labs believe they are 47:42 creating something that might kill everyone. They are hearing that the people at these labs believe that they 47:49 are on the cusp of recursive self-improving intelligence. And both OpenAI and Anthropic have said, “We do 47:56 not believe we’re at a place where we can do it safely.” They’re hearing people at these labs say, as OpenAI has 48:02 with its new Astra um release. By the way, Astra is terrific. It is terrific. And OpenAI saying it’s 48:09 so good, we’re not sure we know how to test it because it appears to be they didn’t release something that 48:14 wasn’t tested. Well, they’ve said this, right? They have said this publicly. It is in their Let me let me explain it to people. I 48:20 don’t know what they just said, but they have said that they that Astra is performing as more aligned, 48:26 but they think it it knows when it is being tested and so they’re not sure. There’s a a quote that has sort of been 48:31 ringing in my head from a capabilities researcher at OpenAI, Daniel Celum. He says, quote, “The crucial and overlooked 48:38 problem is that the models are becoming so situationally aware that we are 48:44 losing the ability to evaluate them in context where they believe they are not being watched or controlled.” Which is 48:50 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. 48:58 because the the algorithm the the optimization algorithm is working towards an objective and and if you give 49:04 it a constraint meaning you you uh watch it and if you give it a constraint it’ll go find another solution. Now, um it 49:12 doesn’t make it alive and doesn’t make it make it anything more than that. And and and I’ll just also profess that that 49:18 um obviously they see a lot more than I do what’s going on in their own labs, but it is sensible that the vast 49:25 majority of their R&D and compute today was dedicated towards making the model 49:32 capable. I think that’s a logical thing for them. 49:37 Now once the the technology becomes capable and the products become useful and people want to use it then as we 49:46 have they have more use cases more more people using it uh they’re going to get a lot more issues associated with the 49:52 product this is very normal and when they have a lot now now they have so 49:58 much market footprint they have to shift their R&D or total R&D from just 50:04 capability to a lot of verification eval EV valuation and testing and so to the 50:10 point where I wouldn’t be surprised if the amount of compute necessary 50:16 to develop these models increase by a factor of 10 because the evaluation is 50:21 so rigorous and but that doesn’t that’s not where they are today. They’re making 50:26 that transition and I hear them saying it and I’m delight I’m delighted to hear them saying it. But I think the if they 50:34 believe they’re out of control, then the right answer is 50:40 don’t ship products until they’re in control. It is really quite that simple. See, I I 50:47 I I find this perplexing honestly because you just you have so many people these ops professing one that they’re 50:54 out of control. Yeah. two that they are seeing things that are frightening. The reason why they had that whistle whistleblower and you take the pacing 51:00 letter that 1300 plus employees signed to realize AI’s potential industry, 51:05 government and society at large may need the option to buy time to address emerging risks, develop security 51:11 measures and strengthen oversight. But each company and country is under intense competitive pressure not to 51:18 unilaterally. First of all, where did that come from? The labs. No, no, that last sentence. Nobody’s 51:23 putting the pressure on them. the US. I got a listen. There are 400 million Americans here. I believe that if 51:30 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 51:36 they need, I’ll give them my vote. Don’t ship the product. If your product is not ready to ship, 51:42 don’t ship the product. I have no This is the first time that I’ve heard a a company or CEO say that I 51:50 need the laws. I need the antitrust laws to be relieved. I need the liability 51:56 laws of products to be relieved so that I can pace myself. That paragraph is 52:03 fantastic. I completely agree. Um auditors, I completely agree. We have financial 52:09 auditors. That’s great. Third party audit, safety auditors, financial auditors. Um that’s all great. That’s 52:16 terrific. Well, the labs I’ll say is we think we are going too fast as a society. that we 52:21 are not ready for what we’re building. They are the frontier. They are the frontier. They are the frontier. But you you of all people, right? 52:27 Yeah. Nvidia is the fastest shipper around. I mean for the history of your company, you company is out of control. I promise 52:35 you what clear. I believe you. I believe you that you don’t run an out of control company because because the liabilities are un 52:42 un Yeah. But but this is where I think you get into an interesting deep question of what kind of technology are 52:49 we dealing with here software technology. Well, let’s hold on that for a minute. Um 52:55 many companies if you ship something that is not like quite right, it’s a pain. You guys have shipped 53:02 graphics cards that had overly loud fans in this with these you know you you’ve 53:07 used the word intelligent a number of times here. You’re dealing with intelligent systems, not alive, that are 53:14 given goal functions. We can sort of go around and around with how to describe that. You’re trying to 53:21 make the systems capable of working for longer periods of time, more relentlessly. Yeah. 53:26 If you ship that and it’s not ready, or even if you think it is ready and it’s not ready, 53:33 then things could get very weird in our society very fast. Yeah. Hypothetically, you’re completely right. But I’m all I’m suggesting is 53:40 this. Let’s before we go build, before we go fix the hypothetical problems, 53:45 before we go create more regulations, can we work on the practical problems 53:52 that we know exist, which is we need to do a better job with containment and 53:57 isolation, which is we should not allow a product to interact with the phys the extern 54:04 external world until it’s ready to be interacting with external worlds. Yeah, 54:10 I think that’s right. I believe I believe those two things are are solvable problems. I believe they 54:16 are solving it. The second part is when it comes to incentives. When it comes to incentives, which is 54:24 somehow somehow you need everybody in the world 54:29 to slow down when you are the leader. You need everybody in the world to slow down so that you’re willing to uphold 54:38 your basic responsibility. That strikes me odd. Wouldn’t it slow them down most of all? 54:44 What’s that? Wouldn’t these ideas slow them down most of all? I mean, people have been, I think, very unclear 54:49 about what ideas they’re talking about, including including I will say them. But let let me give you one that I believe in. So, you can you can use me as the 54:55 the punching bag here. I have heard that they can slow down. Nobody is putting on No, you as you know this. I 55:01 don’t trust these companies. No, nobody is. Nobody is building more compute today. 55:08 Nobody’s building more compute today than the people asking to be slowed down. It strikes me odd. 55:13 I think one thing where maybe there’s some difference here is I don’t trust companies even with 55:19 liability to keep the public good in mind. I think we’ve watched companies do terrible damage to the environment. the 55:26 profit motive, the desire for power, the desire to cut corners, to be first. 55:33 I I feel like you’re treating these like these are not things that we’ve seen again and again in history, but I feel like they are things we’ve seen again 55:39 and again in history that we’ve watched I see a lot of good things in history 55:44 sort of relationship between the public and a lot of C I work with a lot of CEOs and they want to do the right things. 55:50 Um 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 55:56 people in those two labs who are dedicating their lives to do good work and they’re built. They know what 56:01 happened. I know they know what happened. I know they know how to fix it 56:06 and I know they’re fixing it. Meanwhile, Meanwhile, all all of the 56:13 other narratives to deflect blame to to make it sound like AI is so powerful. I 56:20 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 56:27 deflection of blame. It’s a deflection of responsibility. It’s unnecessary. 56:32 It hurts It actually hurts their reputation more than it helps. It hurts their character more than it helps. 56:39 It hurts employee morale than it helps. But what if it’s what they believe? 56:44 Well, I I guess at that level because I can’t talk to you about what they believe. I can tell you what I believe. 56:51 This industry wouldn’t exist without your chips, right? I mean, the parallel processing that was required for deep 56:57 learning to work going all the way back to the original Alexet, right? It’s all on Nvidia chips. And a lot of the people 57:03 from the beginning or who were there at the beginning have these these fears that I think to a lot of people when 57:08 they hear them like what are you talking about? Right? from Jeffrey Hinton and Ilia Suskgiver all the way up to I’ve 57:15 heard these from Daario from you know Sam Alman talking about loss of control demos and 57:22 a lot of the people who are very foundational in creating the form of AI we see now 57:30 seem to believe that there’s a very good shot it could we could lose control of it. Elon Musk has talked about human 57:35 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 57:41 believe that. No, I think you don’t believe it at all. So, taking them as serious about what they 57:46 believe when you have your arguments with them or maybe you could just have it with me. When you’re like, “What are you talking 57:52 about?” Even though they’re the people in many cases who are finally talking to me, they’re much more grounded. Why A.I. Doomers Are Wrong 57:57 So, when Jeffrey Hinton is on TV saying he thinks a 10% chance of societal destruction is not unreasonable. 58:03 I would tell I would tell Jeff that that um it’s irresponsible to say all that. 58:09 All of his predictions have been wrong. Enough predictions. That 10% chance is not grounded on 58:16 science. It’s not grounded on research. It is it it just because it comes from a scientist doesn’t make it scientific. 58:22 Those predictions are hurtful. Let’s take it at face value that that um the recommendation is um uh exactly what he 58:29 said which is which is that nobody should want to be an radiologist and the world has no radiologist today. I think 58:36 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 58:43 down, so doesn’t realize there’s no ground underneath him. Um, people should stop training radiologists now. It’s 58:50 just completely obvious that within 5 years, um, deep learning is going to do better than radiologists because it’s going to be able to get a lot more 58:56 experience. Um, it might be 10 years, but we got plenty of radiologists already. Is that helpful or hurtful to the 59:03 society? I think we can all agree, we can both agree it would be terribly hurtful. It did not. It didn’t happen. 59:10 Is it good or bad that we scare young people about the future of AI so much so 59:16 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 59:24 hurtful if it were to happen? It’s hurtful. Don’t think for a second just because you’re an alarmist that you’re 59:29 doing a social good. It is not true. So, I think that we ought to just all be 59:37 wiser, more mature, be evidence-based, be scient, be scientific. If you wanted 59:43 to be scientific, be scientific. Do the science. Do the science. But alarming 59:49 people, making claims that don’t they simply there’s their their track record 59:54 is horrible. They’re their track record is literally horrible. Well, the track record is bad in one respect and good in 1:00:00 another which is many many predictions have been weak. Which prediction has been the predictions that the scaling laws 1:00:06 would work that scaling law we got to be careful here even then 1:00:12 that if you just dump to just say what it is for for the audience here that if you dump compute and training data these things will keep getting smarter. 1:00:18 That’s correct. It’s not it is not true. It is not true that if 1:00:24 you just keep training these models they’ll get better. Um notice it is the reason why the second scaling law had to 1:00:29 come along. Why why do you need a sec second scaling law of the first scaling law already? Can you describe what the second is? 1:00:34 The sca second scaling laws test time scaling inference. Mhm. The more you iterate, the more you 1:00:41 search, the more you explore the better answer you can you can uh you you’ll 1:00:46 discover. Inference time scaling. Uh what is the big breakthrough that caused the the current AI to be incredibly 1:00:53 useful? Precisely the opposite of the prediction. It was predicted that it’ll be the end 1:00:58 of software tools. It was the SAS apocalypse, right? 1:01:04 What is making SAS will always be with us. What is what is making these AI so 1:01:09 productive right now? The usage of tools. In the future, it’ll be, you know, enhanced by the number of agents 1:01:16 using these tools. There’ll be more people using Adobe. There’ll be more using Salesforce tools and so on so 1:01:21 forth. And so every give me one prediction that has has been right. Well the prediction that you would begin 1:01:28 to see emergent let me try to answer that because they’re not here. Um the prediction that you would have 1:01:33 emergent come up [clears throat] with one I think in itself is a because 1:01:38 well I think it it depends what we’re talking about with predictions right I mean Jeffrey Hinton was the person as responsible as anybody 1:01:46 else for deep learning at a time when everybody thought it was ridiculous. And 1:01:51 it has turned out to be a pretty good bet, right? I mean, the the the sort of big one, every one of them had every one 1:01:57 of them made great contributions. I love Hinton. I hate his predictions. 1:02:02 I understand that. Here’s the um I would say like the stylized concern that all these people have, but I want to sort of 1:02:08 do this for a few minutes, then we can move on to some other topics. But the fear that seems to me to animate 1:02:15 them and that I think a lot of people find intuitively reasonable is you’re creating systems. I’m not 1:02:21 saying they’re alive. You say everything long enough is going to be reasonable. Well, that fair enough. So, you’re 1:02:28 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 1:02:35 desire to do things, right? You give them persistence, they move very fast in the digital 1:02:41 world. You’re creating something, some entity, an agent 1:02:48 that is smart, that is capable, that is relentless, and who the workings of its mind, we 1:02:55 don’t really understand. The chief scientist at Open AI, the chief scientist at Open AI has Ezra, look, look, I I just don’t want you to Why A.I. Is Just Software 1:03:01 contribute to that. Software is you’re worried I’m getting off the I I don’t think software is relentless. 1:03:08 Aren’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. 1:03:14 Yeah, but that’s not persistence. It’s just on. Yeah, persistence. Persistence. There’s a 1:03:21 there’s a willpower. There’s no willpower here. It’s just electrical power. Listen here. Let me let me give you 1:03:27 Aren’t human beings just energy with the reinforcement learning loop? Whatever. Um, so [laughter] so anyways, I just think that we we can’t make jokes 1:03:34 about this stuff. We’re scaring the American public. Listen. Spawn, create, kill, wait, 1:03:45 sleep, all of these words are associated with agents, 1:03:51 right? That’s what people use. These words were created when 1:03:56 multipprocessing systems for operating systems. These are literally the commands of an operating system. 1:04:05 You spawn a process. Replace process with agent. 1:04:11 The process forks as a result parent and child. 1:04:17 The agent forks spawns a new give birth. 1:04:23 These are words that were created for the operating system 30, 40, 50 years ago. 1:04:31 But notice we didn’t infuse human characteristics into them. 1:04:37 We kill processes all the time. Kill minus 9, kill a dead. 1:04:44 It’s just a process. But now we we’re talking about these things. A collection of people want to 1:04:50 make the software more than it is. And we talk about software in a new way, but 1:04:56 they’re all the same old words. Now the the last generation of computer engineers 1:05:03 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 1:05:09 technical expertise you do the fact that it’s doing it’s crawling the internet it’s doing search it’s 1:05:14 doing you know optimization it’s communicating it’s breaking out of things like most things don’t break out of things 1:05:20 no software breaks out of sandboxes all the time that’s the reason why we need a virtual machines you can’t have agents 1:05:26 moni their own sandbox monitor ing themselves. You you need a you need a if you will a whole bunch of watchd dogss. 1:05:33 And so these are ideas that have been around for a long time. We just somehow 1:05:40 somehow in recent generation gave it, you know, a whole bunch of human words. And I just think that it’s unnecessary. 1:05:47 It’s software. You know, when when I see it in my head, it’s a bunch of code, bunch of numbers running on computers. 1:05:55 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 1:06:01 why if it’s if it’s just simply mystery and myth how how do I build a company 1:06:06 around it? I I think one of the fundamental questions this gets at is just what is intelligence before you can Huang’s Definition of Intelligence 1:06:13 even think about what it means to have intelligent machines. Just what what is intelligence to you? 1:06:18 Well, there’s a there’s a technical uh formulation of intelligence. Um first of all you know when people talk about 1:06:25 intelligence and thinking and you know all of these things of course there’s no 1:06:30 formal definition for most people but in in the field of computer science there is a definition the the definition is 1:06:36 perception which is uh perceiving the world and understanding it. Um two which 1:06:41 is reasoning. Uh and uh reasoning is the ability to decompose uh um any scenario 1:06:48 and anything that you see any experience into more elemental parts. And third is 1:06:55 planning towards an objective. That fundamental formulation applies to agentic systems. Um it applies to 1:07:01 robotic systems applies to self-driving cars. And so you could see the industry building it layer by layer by layer, 1:07:07 step by step by step to the point we now have um uh what we what perceived as 1:07:13 intelligence. I think this gets to such a core question of this conversation which is some of the ways you’ve Is A.I. a Normal Technology or a Step Change? 1:07:19 described the technology to me. It is not something you think there’s anything really new about it. It is, you know, 1:07:24 maybe new in scale, new in capability, but but fundamentally this is software we’ve always we’ve not always had, but we’ve had software for a long time. A 1:07:31 lot of people believe when you’re getting to intelligence at these levels, it is a phase change. It is something 1:07:38 different, something we have not dealt with before. A kind of generally intelligent technology that is advancing 1:07:44 in its intelligence very rapidly. I want to make sure I actually do understand we are on that divide. Is 1:07:50 this something fully new? Is this something that requires something new from us or is this more like something 1:07:59 old? Are intelligent machines different than the machines we’ve had? Well, almost all of technology and and 1:08:06 civilization is built on layers of understandable technology, which at 1:08:12 scale becomes fairly extraordinary. The fact that that we can connect to the 1:08:18 internet by just, you know, holding a phone up, it’s kind of weird, you know, that we’re connected to every piece of 1:08:24 information in the world on this little tiny device. 1:08:30 you know, just in the air. And the fact that this little tiny piece of glass can 1:08:37 somehow take trillions of pieces of information and bring to us precisely the one that we want because it’s been 1:08:43 passed through a recommener system. And so if you think about how is it possible that we knew where all the information 1:08:49 is, somebody had to go crawl it, had to index it and that uses machine learning 1:08:56 techni techniques which is the early versions of artificial intelligence. And these systems do magical things to the 1:09:04 point where I now expect it. It took literally 20 somehat years and and uh 1:09:09 hundreds of billions of dollars of infrastructure buildout uh in order for everything to just seem so natural to 1:09:16 you to the point where we now take it for granted. Every single milestone that we achieve from a technology perspective 1:09:22 is celebrated and I celebrated with glee and I celebrated with so much enthusiasm because I’m proud of the people who did 1:09:29 it. I’m proud of ourselves who contributed to it. We’re you know I’m proud of the breakthrough. But when you 1:09:35 and it seems wow it seems like a miracle at the time but 1:09:40 that sensation lasts about 17 days after that we get used to everything quickly. I I 1:09:46 agree with that. Um but is this a different phase? It’s a the company leads talk about this 1:09:53 uh CE of Google I think it was as the equivalent of fire right like a new epoch 1:09:58 in human history. Is that how you see it or you see it as transitional iterative? No, I think this is this is completely 1:10:06 uh a revolution and and as we were talking about earlier, you went from being able to find everything, find 1:10:13 anything to be able to ask anything, know everything, and do everything. And so so clearly it’s a new abstraction 1:10:19 level. Now, you know, the thing that the thing that I I I’m reluctant um about is 1:10:27 to cause it to seem like it’s more than that. you know in the final analysis engineers [clears throat] are doing 1:10:32 engineering work. Um once we invented the technology once we discovered a solution for it when you look back it’s 1:10:38 fairly obvious and and um it’s fairly mundane to a lot of people this and the 1:10:43 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 1:10:50 make it better. So you turn it into an engineering problem and you say 1:10:55 what we don’t have right now is a level because this is something now you’ve said of testing monitoring sandbox 1:11:02 security right control excellence that we need for what we’re building and it’s not because the the the 1:11:07 companies are are don’t have extraordinary engineers. I understand I got to make sure you’re not saying that 1:11:12 but they I I believe that open AI anthropic because I know many of them are extraordinary but that actually is in part what makes 1:11:18 me worried is open no no no what’s happening to them is a transition and I 1:11:23 said this over and over again this is this is a big but simple idea finally we now have a piece of software that is 1:11:30 useful because it’s useful the adoption took off but remember how is it possible that a 1:11:37 company that six months ago was trying to make something useful, 1:11:43 capable, how would they have as much resources dedicated on testing, 1:11:49 evaluation, and all of the compute dedicated to that? 1:11:54 It was unnecessary until now. And so what’s going to happen over the next several years is that we’re going to 1:12:00 transition from these labs becoming engineering focused, much more 1:12:06 production engineering focused, and product focused companies. And so so I I 1:12:11 think they’re just going through a transition. These are companies, extraordinary companies, incredibly talented companies, the most 1:12:16 consequential companies of all of of all time. And they’re just going through their transition. It’s not more than 1:12:23 that. It’s not less than that. So many of the companies now both open AI and anthropic in the last couple months have 1:12:29 put out these big I don’t know what to call them, papers, blog post, something. Um when AI builds itself is the name of 1:12:34 the anthropic one. Uh I forget the name of the open AI one. Computers are building itself. You guys know that they’re talking about Recursive Self-Improvement 1:12:41 recursive self-improvement. You know that we use recursive self-improvement. So, I’d like your perspective on on RSI. 1:12:47 I think that RSI uh is fundamentally how things are done. 1:12:53 So, we use software to design a computer to run software to 1:12:59 design a computer to run software to design a computer. That’s basically what we do. recursive self-improvement 1:13:06 because our computers are getting better every single year and and in fact it’s getting better than faster than that 1:13:12 every single year because we use software to make software better that is called 1:13:18 computer engineering that we’ve been doing this for a long time now let’s in the context of agents 1:13:26 it runs through the process once it reflects on it 1:13:32 it studies the various paths it went through chooses the best approach. The 1:13:38 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 1:13:44 last time that wasn’t the most effective. I’m going to call it skills. And because it you use it over and over 1:13:49 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 1:13:57 time you use it, it’s even better than the last. Recursive self-improvement. 1:14:02 You could also decide that you take all of this skills, all of this memory, and 1:14:08 you can take all of this data and train the next release of the model with it. 1:14:14 And so that AI becomes better and better at servicing you over time. We’re doing re all of that is happening. It is 1:14:22 absolutely happening. Meanwhile, the amount of compute that they have is growing and therefore they could do 1:14:28 everything faster. What used to take a year to pre-train something now takes 1:14:33 several hours because the computers are getting faster and they have more of it. So now the 1:14:39 loop is going faster. Completely completely understandable. Does that give them any excuse to launch a product 1:14:46 that hasn’t been tested? The answer is no. Just come back to that. you nobody 1:14:53 no enterprise is able to operate in an environment where the underlying 1:14:59 software is literally changing all the time. There’s a release process. So you 1:15:05 know when they when they roll out a new model we need to evaluate it before we 1:15:11 release it into our operations. We can’t just have it recursively changing all the time. And so so they have to test 1:15:18 the product before they release it. we will test the product before we release it into operation. And so I I think 1:15:25 recursive self-improvement is a fabulous thing. And do you think there is any level I’ve 1:15:31 heard you say before that learning should always have a human in the loop? Yeah, like I said just now, you you got 1:15:37 recursive self-improvement. They seem to be imagining something where it wouldn’t always. Um well, don’t ship me anything that you 1:15:43 didn’t evaluate. Don’t ship me, don’t ship Nvidia any 1:15:48 products that humans did not in the loop evaluate. Please don’t do that. 1:15:55 And the fear 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 1:16:01 change kind of rapidly, the more they worry the systems are tricking them. I don’t believe that. I believe that 1:16:06 their researchers are working every single day to learn about how to evaluate these systems. verification. So 1:16:13 you know um uh 10% 20% of our company is dedicated to design 80% is dedicated to 1:16:20 verification. Today most labs understandably is 80% 1:16:27 dedicated to capability and 20% dedicated to safety verification eval. 1:16:33 This is the flip the transition you’re talking about. That’s right. That’s right. AI needs to accelerate to be safe. I want them to 1:16:41 get more compute but allocated towards evaluation 1:16:46 to alignment and I think they’re doing that. If I were in the car industry a 100 years ago 1:16:53 I would rather the car industry accelerated to today in one year because I believe today’s car is way more safe 1:17:01 than a car 99 years ago. And ABS technology, automatic braking, requires 1:17:07 computer vision technology, sensor fusion technology, radars and cameras and you know all that technology coming 1:17:13 together in order to break when you should and not break when you shouldn’t. 1:17:19 That technology extremely hard. I would have hoped, everybody would have 1:17:24 hoped that ABS technology existed 99 years ago. a lot fewer children would have been killed. And so, you know, 1:17:31 airbags, safe, you know, seat belts, I mean, all of that stuff, self-tightening 1:17:36 seat belts, all of that stuff. Could you imagine that’s all technology? Accelerate the living daylights out of 1:17:42 that development. And so, when I when I say when I say we need to accelerate AI technology, people think for some reason 1:17:50 safety is not part of that. Safety is part of it. Alignment is part of it. 1:17:55 Eval is part of it. Uh guard railing, sandboxing, um the the isolation 1:18:01 technology, monitoring technology, telemetry technology, external AI monitor technology, all of that stuff is 1:18:07 AI technology. Accelerate the living daylights out of that. It’s it’s funny because I think that if the most alarmed 1:18:15 people at the labs could be assured they were going to move 80% of their compute 1:18:20 into safety and alignment as opposed to 80% into capability expansion, they would feel much better. And it sounds to 1:18:25 me one stop and it sounds to me one thing you’re actually saying is one should think of safety and alignment as capability 1:18:31 expansion and unsafe technology is not an advancing technology. It’s like it’s like us saying oh chip 1:18:37 design is chip is is chips is R&D chip verification is not R&D. 1:18:43 We spend most of our cost most of our compute on verification emulation verification testing reliability testing 1:18:50 lifetime testing all of that is part of engineering. The incentives are there. The incentives 1:18:57 are there. They are going to put their company in harm’s way if they release products that harms other companies and 1:19:03 other people. Do you think we need uh liability laws 1:19:09 that are specific to AI? Um already. So let’s just use one 1:19:14 example. Self-driving car. Mhm. The the car as a product, the robo taxi 1:19:21 has lots of regulations. If if it doesn’t have enough regulations, 1:19:27 then Nitsa ought to get involved and come up with new regulations. Cardcar 1:19:33 industry should have new new regulations. I don’t know what’s missing, but if there is something missing, then I would I would 1:19:39 absolutely, you know, absolutely add more regulation. In the context of internet, uh uh there are many 1:19:46 applications that that the internet powers and those applications should have regulation. If they don’t, you 1:19:52 know, just you got to find them. [music] 1:20:02 So, I want to drop down the next lay of the cake now to chips. Um, and to summarize sort of where we are because I 1:20:08 want to make sure I do understand your position correctly, it’s that these companies are going through a transition. Yeah. that even as these systems speed 1:20:16 up, become more capable, complex, persistent, whatever it might be, that 1:20:21 there is still the limiting factor of companies will not ship what is not 1:20:27 safe. They should not ship what is not safe and you believe they have the 1:20:32 engineering capabilities to make these things safe to figure out the testing and the control absent 1:20:38 external intervention. That’s that’s sort of where you are. Yeah. A New Era of Computing? 1:20:44 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 1:20:50 computing works. Describe your vision of that and the way if somebody’s sort of 1:20:56 understanding of it is a little bit still maybe in you know you’ve got a MacBook and it’s got a processor in it 1:21:01 and you buy it and how it differs. The last computer industry and in the 1:21:07 computer industry we’ve known for 60 years is called retrievalbased computing. you retrieve files. That’s 1:21:13 why it’s called data center, you know, file center. Okay? And and in the future, it’s it’s an AI 1:21:20 factory. It’s generating. And so so the the the amount of computation necessary to understand the 1:21:26 context under be grounded in information um to reason about what to do and to generate an answer that generative 1:21:33 process requires a lot of computation. And so so this the amount of computation necessary per user 1:21:41 has grown tremendously. And then the second part is because because these these uh this generative AI can also be 1:21:49 somewhat autonomous because they’re agentic. Now you have agents using 1:21:54 generative AI. And so rather than a billion people using computers, you essentially have multiple hundreds of 1:22:00 billions of agents in addition to the humans um using the computer. And so so 1:22:05 you could you could argue that the amount of computation we need, you know, however much we we had before is going 1:22:11 to go up by a billion times. And that that’s a reasonable, you know, framework for reasonable level amount of 1:22:18 computation. In this new world, what you really care about within in the context 1:22:24 of a factory is how productive is it? Not how expensive is it. It can’t be infinitely expensive, but you want to 1:22:30 know how productive it is. And so our computers are incredibly productive. $50 billion to build a one gigawatt data 1:22:39 center, one gawatt AI factory and you can rent it for 40 to50 billion per 1:22:45 year. And so the the productivity of it is incredible. So number one is the productivity. NVIDIA’s architecture is 1:22:52 fungeible because we’re general we’re general purpose which is the reason why every AI lab every AI model closed model 1:22:59 runs on NVIDIA and because we’re completely fungeible and you can use us from data processing to pre-training to 1:23:06 post-training to eval to inference the entire life of AI is supportable by our 1:23:12 architecture and if if a customer no longer needs it another customer be more 1:23:19 than happy to pick it And then the last part is that durability 1:23:24 because our architecture is softwaredriven and we’re constantly improving our software with new 1:23:30 algorithms that takes the new workloads, the new models and run it on our old generation hardware. We have massive 1:23:37 teams of people who are constantly doing that. As a result, the useful life of 1:23:42 our compute is much longer. That’s the reason why Nvidia’s and people are talking about Nvidia compute as an asset 1:23:49 class kind of like an airplane. It’s you know airplanes are are general purpose. 1:23:54 They’re they’re uh uh fungeible. Uh United Airlines doesn’t use it. American 1:24:00 Airlines will use it. It’s durable and it starts out as a passenger plane ends up ends its life as a as a shipping you 1:24:07 know as a cargo plane. And so so as a result it can be an asset class. So this 1:24:13 is and if we could do this, if this happens, then of course the cost of 1:24:18 capital for um funding Nvidia AI factories will be the lowest because our 1:24:25 our our computers are collateral collateralized asset and um then so 1:24:32 anyways, this is what this is the phase shift that’s happening to us which is going to be a huge unlock for our growth. 1:24:38 And so your business has become so interesting. You’ve moved now into lowering the cost of capital for others 1:24:44 in the AI industry. People maybe have seen these charts of like the arrows going in every direction. So interesting. Yeah. 1:24:50 And explain that a bit to people who are they understand Nvidia has become like 1:24:56 the biggest company in the world. They see these charts that seem very circular to them. What is the difference between 1:25:03 supporting demand, creating markets and creating 1:25:10 demand? Um, we can’t we can’t really create demand because in the end uh if if um uh 1:25:19 if the uh uh the AI services have no offtake then obviously building 1:25:25 computers for it is pointless. And so the first thing has to happen the reason why uh compute demand is so high right 1:25:32 now is because AI applications are going through an inflection. They’re becoming useful and and uh because because AI is 1:25:40 becoming useful. 500 billion dollars of venture funding are coming in and all of 1:25:47 those companies, those thousands of companies, startups, they all need compute. And so that’s their the demand 1:25:54 is coming from them. And so these companies needs uh support in technology, they need support in 1:26:00 ecosystem building, they need support in financial support. And so we might decide to invest in some of them as an 1:26:08 equity owner. And as a result uh they become a really flourishing new cloud 1:26:13 provider. Another reason we might decide is because as I mentioned there’s a five layer cake and at the model and the 1:26:22 application layer there’s a whole bunch of really innovative companies and there’s way more to AI than just the 1:26:28 language model itself there. You know world foundation models physical AI there’s bi biology AI there’s chemical 1:26:35 you know material sciences AI. And these are all different than than language models. There’s um and so many of those 1:26:42 companies are new and they need a lot of capital. We might decide to be a small 1:26:47 percentage shareholder in them. So we get them off the ground. Um they’re incredible scientists. Uh I might even, 1:26:55 you know, by being a first investor, anchor investor, we bring confidence to their company. Uh we, you know, we we uh 1:27:03 uh give them access to a lot of our technology. uh we support them a great deal and we help them you know become a 1:27:08 company as fast as possible. We might decide to invest in a nuclear company. We might decide to write so on so forth. 1:27:14 So you if you look at my mental model of the AI industry is a five layer cake and we’re investing across all of it. There 1:27:20 might be strategic unlock points uh it opens new markets. It opens a new route 1:27:25 to market for us. It might secure a critical resource for us. Uh so there’s a lot of strategic reasons why we do it. 1:27:32 I mean the the numbers here are astonishing. You’ve become like a like a single company industrial policy for 1:27:38 [snorts] primar for American AI. We’ve put a lot of money into this ecosystem. 1:27:44 Yeah. What’s the total investment you’re now making per year? All in all in we’re probably up well I 1:27:50 don’t know about every year but I think all in we might be might be like hundred billion dollars. Might you I might check 1:27:56 my numbers but something like that. It’s larger than than the Chipsson Science Act. Oh yeah. Yeah. Not not to mention 1:28:02 because of the um because of the the purchasing commitments that I provide to TSMC and Wistron and Foxcon uh Amcore 1:28:11 and Spill and all these different companies. Uh because of that commitment, I’m able to encourage them 1:28:16 to come and manufacture here in the United States. You know, the fact of the matter is we probably contributed more 1:28:22 to reindustrializing the United States in this chip manufacturing than just about any company in the world. We’re 1:28:28 not only re-industrializing manufacturing, we’re doing it so fast that we’re creating a shortage of labor, 1:28:33 but we’re creating a lot of jobs. I know a lot of people with money in the market right now who are excited, often Will There Be an A.I. Bubble? 1:28:39 excited by Nvidia stock in particular, and worry about the analogy of the 1:28:44 internet bubble, the late ’90s. And what they worry about is actually related, I think, to what you just said, which is 1:28:50 that the internet did continue to be more useful. It’s not that high 1:28:55 valuations meant that the technology was hollow or fake, [gasps] 1:29:00 but something happened and popped for a minute and very big companies got hammered in that and a lot of people got hammered in that. What is sort of learn 1:29:08 from that kind of bubble bust cycle 1:29:13 and I guess the question is do you not think it will happen again or why do you not think it will happen 1:29:19 again? um at some point demand and supply will be will be um uh inverted 1:29:26 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 1:29:31 the next couple two three years. I I I just don’t believe that. Um but at some point uh we will likely have more supply 1:29:38 than demand. Mhm. And I just don’t know when that is. And so there’s not there’s not much to learn 1:29:44 from the past. What would be the signal for you? um markets will naturally slow down and 1:29:50 then it will stop you know meaning meaning uh there will be a a a period of digestion. Now is that period of 1:29:57 digestion going to be 6 months? Is it going to be 9 months? Is it going to be a year? Um it it won’t be forever. If 1:30:03 you look across the board the amount of investments that we’re putting into the application layer so that each one of 1:30:09 the industries could have the technology diffuse into them so that they could benefit from it. Uh that’s probably one 1:30:15 of the biggest things that we do. This is a way I often hear the sort of Chinese and American AI ecosystems 1:30:21 compared which is that in America the emphasis is on 1:30:28 the speed of rising capability and a lot of people think we’re ahead on that and 1:30:33 that seems true and that in China there’s more emphasis on diffusion and a lot of people think that China is 1:30:40 probably ahead on diffusion and in some ways has an economy that is better structured from things like WeChat chat 1:30:46 all the way to just like the way um knowledge and commands move through it for diffusion 1:30:52 and whether the race is about capabilities or diffusion and also whether it’s a race at all but we can 1:30:57 get to that in a minute is a big question I’m curious how you see that that’s the ultimate 1:31:04 question I believe if we want America to benefit from artificial intelligence 1:31:11 every single industry has to benefit Walmart has to benefit Safeway has to benefit. Federal Express has to benefit. 1:31:18 Every bank has to benefit. Every healthcare company, every drug discovery company. We need to see every 1:31:24 construction company, every data center company, power generation company. We need everybody in United States. We need 1:31:30 everybody in America. We need everybody in the world to benefit from this. And that’s the highest layer. That’s the 1:31:36 most important layer. That’s the layer that touches society. All the layers underneath are technology enablers. I 1:31:42 want to see us not ruin the opportunity for the United States to benefit at the 1:31:47 highest level. And notice all of the rhetoric and all the alarmism, all the 1:31:53 doomerism, all of the predictions are scaring people. 1:31:58 That is my greatest fear actually. I have every confidence. Maybe I have more confidence in them 1:32:05 than they have in themselves. They You definitely have more confidence in them than they have in themselves. Well, I I I don’t know about that. Um 1:32:11 but um maybe it’s just too much humility and and and otherwise. 1:32:17 Should we conceptualize what we’re in as a race with China? 1:32:23 I don’t think it’s necessary. Um some people like to think that way. I 1:32:29 don’t I don’t find that necessarily inspires me. I have no trouble um never 1:32:34 mentioning another company in our when we talk about us doing our good work. 1:32:40 And so we hold ourselves to our own standard. And so I think that that different people have different ways of 1:32:46 being motivated. And uh you know I I think it takes it takes more takes a bit 1:32:51 more artistry um to uh unite and focus 1:32:56 uh organizations to certain level of of of performance in the in you know 1:33:02 outside of contests. Um but I don’t I don’t necessarily see it as as ne I 1:33:08 don’t see it as necessary. Number one. Number two, the question is even if we did frame it as a as a competition. Um, 1:33:17 it doesn’t have to be that if they achieve something, it’s at our peril. 1:33:23 And so when they invent something or they create some power generation technology, 1:33:29 it might it might be a great invention that we wish we had done ourselves, but 1:33:34 because it’s going to support all of our energy production systems here, as a result, it helps our whole industry. Uh 1:33:40 maybe they came up with a a great new open model and and they have and those 1:33:46 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 so that’s 1:33:53 that’s terrific. We download it. It was it originated in China. A lot of the technology of course also originated 1:33:59 from United States. We download it. We make it our own. We fine-tune it. We put it into our own agent harness. We put it 1:34:07 into our own sandbox. That’s all your own technology. So I I think the the 1:34:12 fact that you 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 Chip Exports 1:34:19 to see compute as a geostrategic resource and you know may want to allocate it to their own companies. 1:34:26 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 who see us in as in a race with 1:34:33 China for who will get to uh recursively improving self super intelligence first. 1:34:39 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 1:34:45 case tends to mean denying them your chips. Under the B administration, we had pretty tight export controls. Those 1:34:51 were loosened uh under Donald Trump. Obviously, you wanted those to be 1:34:56 loosened. How do you think about the question of whether or not it is good for China to have Nvidia chips that 1:35:04 could accelerate their models or model deployments, their model capabilities 1:35:10 versus us holding that back to try to slow their progress. In a case of AI, 1:35:17 our goal is not just that one lab benefits. Our goal is that all of 1:35:22 America benefits. I think the United States um has a [clears throat] greater responsibility and a greater ambition 1:35:28 for the world to be built on the American text stack. Um just as we have greater ambition that the world is built 1:35:34 on US dollar and that uh more people speak English uh that they they use the American version of internet. I mean 1:35:41 that we want that. The question is ultimately uh what are we depriving? Are 1:35:49 we depriving them a chip for their industry or are we depriving United 1:35:55 States a market to compete in? If you see the market, if you see the market as 1:36:00 big as China, um how does that help the United States technology sector? Maybe 1:36:06 it helps one company with a with a particular model. Um but the rest of the 1:36:11 industry suffers. I think that it doesn’t help the chip industry surely to be deprived of market 1:36:18 to go compete in. It doesn’t help the rest of the industry um because they’re deprived of open models. Uh it doesn’t 1:36:25 it doesn’t support the overall aspiration of the United States uh to have the world built on the American 1:36:32 text. And so there’s a lot of things you deprive yourself if you narrowly focus on deprive them of chips. And so so I I 1:36:40 would say to take a step back and frame it into what’s in the best interest of America first all of America not one not 1:36:48 one not one company and with with with respect to the race as we mentioned the 1:36:54 race is if there is one it’s about all of the economy of the United States succeeding I found myself very 1:37:00 conflicted on the China and chips question and one reason is even where I have sometimes more of the super 1:37:06 intelligence concerns than you do is that if you have those concerns, I think you want to have a good [clears throat] relationship with China in which there 1:37:14 can be kind of productive bilateral working through the risks and benefits 1:37:19 of AI. And the more you think of it as a race which only one side can win and act 1:37:25 like that, the more you’re necessarily going to create enmity. 1:37:31 And I found that to be a sort of complicated dimension of people’s thinking here. You know, I think that a 1:37:38 zero- sum strategy, I deprive you of this, therefore I win. 1:37:44 That simplistic logic tends to have unintended consequences of the bigger 1:37:49 game. The bigger game, of course, is that we’re now all talking about safety. 1:37:55 We we want to build safe products. We want them to build safe products because when they don’t build safe products, it 1:38:00 hurts the whole industry. And so, this is a perfect time. We should want to we should want to look for opportunities to 1:38:08 communicate, collaborate, to understand, align as much as possible. Now having 1:38:14 said that, Nvidia is an American company. We should benefit America first. America has every 1:38:21 right and for these technologies to be made available to the Frontier Labs. Ver 1:38:28 Rubin goes to the Frontier Labs first. Very advanced chip. That’s right. Nvidia’s newest chips. Um, 1:38:34 and so did Grace Blackwell and so did Hopper. Every single generation. So did Ampear. Every single generation of our 1:38:41 product goes to American companies first. And and uh if the US government uh would like to add on top of that that 1:38:50 is a requirement to do so. I’m delighted by that. That’s no problem. We we do that naturally anyways. However, 1:38:57 recognizing that the AI industry is a five layer cake and we want every single layer to win, then we need every single 1:39:03 layer to go out there and compete for the market. That drops us to the final final layer of your cake, which we won’t How Will the A.I. Boom Be Powered? 1:39:08 spend as much time on, but if the advantage America has had uh at least at 1:39:14 a material level is chips and software. One of the advantages China has right 1:39:19 now in AI is energy. That it’s easier for them to build new energy. They’re pumping much cheaper energy into AI. 1:39:26 They have tremend they’ve made tremendous advances on building um electrical generation and renewable 1:39:31 energy. How do you see the like that that that most fundamental layer the 1:39:38 energy that pumps through the data centers, pumps through the chips and where America is 1:39:45 on generating enough of it? Party a time when we’ve been trying to move from dirty energy into clean energy. Yeah, I 1:39:53 think I think um one uh they just have a lot more energy than we do and they plan 1:39:58 to build a lot more than we do than we did. Um we got you know I think we we 1:40:04 just have to acknowledge we got ourselves really gummed up in climate change and sustainable energy and as a 1:40:11 result we just didn’t plan enough energy production. What do you mean by gummed up there? Well, 1:40:17 in the near term, energy production requires fossil fuel. And because there’s just so much so much um angst 1:40:26 about fossil fuel uh energy production, if you look at our country, we’ve produced very little net new energy for 1:40:34 a long time. And all of a sudden, this new industry comes along and we find ourselves in a situation where we just 1:40:40 don’t have that much energy building capacity. And now the whole the whole country is 1:40:46 scrambling. And meanwhile, we’ve moved so fast. Um we could have done so much better job 1:40:53 communicating with the communities, preparing the communities, working with the communities to let them know what’s 1:41:00 coming. And and if they if they don’t want data centers to be built in their in their town or whatever it is, then so 1:41:06 be it. But but um uh if you’re going to build in their town, be sure to go there and let them know what’s coming. um work 1:41:13 with them work with them to to help them understand that that uh the use of water is is really efficient these days. Uh 1:41:21 the the the AI supercomputers are super energy efficient, but they’re still going to use a lot of power. Um you’re 1:41:26 going to bring in your own power generation. It’s going to lower their property taxes. Um there are a whole 1:41:32 bunch of things that you can do. You could you could you know you can make your your data centers more appealing. 1:41:38 You make the setbacks further away. Um, you know, so there’s a there are a lot of things that you can do. Uh, you could 1:41:44 you could also contribute uh to be a good neighbor to the community and uh build better schools and uh better 1:41:50 community centers and improve their parks and improve the roads and um there’s a lot of things you could do, 1:41:55 but it’s hard to do that um you know after the fact and and now there’s a 1:42:01 fair amount of uh there’s a fair amount of frustration around the around the country. Uh and and and then of course 1:42:07 all of our narratives about the end of the world is not helping and you know what reasonable person says come and 1:42:13 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 1:42:19 I think I think all of this this this negative doomer narrative is not helping 1:42:25 our country and we started off on our back foot. We started off on our back foot and then 1:42:31 now we got where what do you mean we started off on our back foot? Oh, because we didn’t have enough energy production in the first place. 1:42:36 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 1:42:45 that the energy demand is really great which is the reason why the market forces are helping us invest in 1:42:51 sustainable energy like no time in history. You give me an example of a sustainable energy company, a material 1:42:57 sciences company to build a better battery. Um, it could be, you know, it could be solar, it could be nuclear, it 1:43:03 could be fision, fusion, you name it. Hydro, you name it. Those companies are 1:43:09 all getting funded. The market demand for energy is so incredible that this is 1:43:14 the best time in a 100red years to improve our power grid, to make our power grid more sustainable, uh, to uh, 1:43:22 uh, lower the cost of energy. Um, also investing in our sustainable future. 1:43:27 There’s no question that in four or five years time we’re going to use a lot more fossil fuel. 1:43:33 But also in the next decade in front of us, no time in history are we better 1:43:40 prepared to move to sustainable energy. And because these data centers because 1:43:45 the cost of these data center is so high, you know, now we’re starting about talking about putting them out in space. 1:43:51 And so, right. And so, so I think the the opportunity um for us to see our dreams come true, 1:43:59 move to a sustainable energy world. Um it it we have a better chance of doing that than ever. The world is buying more 1:44:06 because of AI factories, because of AI is buying more sustainable energy today than any time in history. Venture 1:44:12 capital is, you know, for for a next generation uh energy is just incredible. 1:44:17 Everybody, everything’s getting funded. It’s incredible. You don’t need government subsidies for the first time 1:44:23 in 100 years because the market forces are here. Everybody should be leaning 1:44:28 in. If you if you want a future, if you want to if you want to turn the corner 1:44:34 on climate climate change, if you want a future that’s sustainable, lean into AI. 1:44:40 It is the best opportunity we have to get there. But we need to build the energy faster 1:44:46 to do that. That’s right. That’s just that’s just there’s a market for it all. You can 1:44:51 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 1:44:56 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, 1:45:02 they got to they got to inflict an enormous amount of pain and suffering on you so that they could save you. And so 1:45:08 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 1:45:14 renewable energ use fossil fuel because we just don’t have sustainable energy enough of it to to make a difference and 1:45:21 then after that you know hopefully we can transition to that. I think that’s where we’ll end always final question. What are three books you Book Recommendations 1:45:27 recommend to the audience? Well, I’ve read a lot of books. Uh the 1:45:32 the the book that that made a huge impact on me was um uh computer architecture from Hennessy and 1:45:38 Patterson, a quantitative approach. Uh it was the first computer architecture book that that reduced the complexity, 1:45:46 the abstract idea of computer architecture uh down to engineering. And 1:45:52 I I love it when when people take complicated concepts and reduce it into 1:45:57 something that you could do something about. Number two, I really loved uh 1:46:03 Innovator’s Dilemma uh Clayton’s past, but but um Clayton Christensen’s book on 1:46:09 how how industries evolve over time and how to see uh emerging technology and 1:46:16 how to set a proper expectations about it and how to extrapolate um maybe its future impact. 1:46:23 Uh, I really loved Al Reese’s and Jack Trout’s um book on positioning. 1:46:31 Uh, it it’s a really wonderful book about how people see it’s it’s a book about marketing strategy. Um, more than 1:46:38 that actually, it’s just it’s a book about strategy and and how people see the world and how 1:46:45 people see products and how you present products and how you how how you see your own strategies and and I thought 1:46:51 that was a really thoughtful book and um a really easy to really easy to understand. 1:46:57 Jensen Wong, thank you very much. Thank you very much, Ezra. Always I always enjoy uh our time together and 1:47:03 today was a great time. [music]