Skip to content
All-In PodcastAll-In Podcast

Jensen Huang: Why cheap chips still produce expensive tokens

Via Groq, Nvidia routes each inference step to the right chip. Vera Rubin targets agentic racks; Huang's key metric is token cost, not the datacenter price tag.

Jason CalacanishostJensen HuangguestDavid FriedberghostChamath Palihapitiyahost
Mar 19, 20261h 6mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 1:00

    Jensen Huang joins the show!

    1. JC

      Special episode this week. We've preempted the weekly show, and there's only three people we preempt the show for, President Trump, Jesus, and Jensen.

    2. JC

      [laughs]

    3. JH

      [laughs]

    4. JC

      And, uh, I'll let you pick which order we do that. Uh, but what an amazing run you've had and, and a great event. Um-

    5. JH

      Every industry is here. Every tech company is here. Every AI company is here. Incredible. Incredible.

    6. JC

      Extraordinary. And one of the great announcements of the past year has been Groq. When you made the purchase of Groq, did you realize how insufferable Cha- Chamath would become? [laughs]

    7. JC

      [laughs]

    8. JH

      I had, I had an inkling that, that, [laughs] that, uh-

    9. JC

      'Cause we're his friends. We have to deal with him-

    10. JH

      I know

    11. JC

      ... every week.

    12. JH

      I know it. I know it.

    13. JC

      You had to deal with him for the six-week close. [laughs]

    14. JH

      I know it.

    15. JC

      It's like two weeks.

    16. JC

      Two weeks.

    17. JH

      I know. It's all coming back to me now.

    18. JC

      [laughs]

    19. JH

      It's, it's making me rather uncomfortable. The, the thing is, uh, many of our strategies are, are presented in, in broad daylight at GTC, years

  2. 1:00 – 9:27

    Acquiring Groq and the inference explosion

    1. JH

      in advance of when we do it. Two and a half years ago, I introduced the operating system of the AI factory, and it's called Dynamo. Dynamo, as you know, is a piece of instrument, a machine that was created by Siemens to turn essentially water into electricity. And Dynamo, uh, w- powered the factory of the last industrial revolution. So I thought it was the perfect name for the operating system of the next industrial revolution, the factory of that. And so inside Dynamo, the fundamental technology is disaggregated inference. Jason, I, I know y- you're, you're-

    2. JC

      Yeah

    3. JH

      ... you're super technical.

    4. JC

      Absolutely.

    5. JH

      I know it. [laughs]

    6. JC

      I'll let you take this one. Go ahead and define it for the audience.

    7. JC

      [laughs]

    8. JC

      I don't want to step on you.

    9. JH

      Yeah. Thank you. I, I, I knew you wanted to jump in there for a second.

    10. JC

      Yeah.

    11. JH

      But it's, it's disaggregated inference, which means the, the pipeline, the processing pipeline of inference is extremely complicated. In fact, it is the most complicated computing problem today. Incredible scale, lots of mathematics of different shapes and sizes, and we came up, came up with the idea that you would change, you would, you would disaggregate parts of the processing such that some of it can run on some GPUs, rest of it can run on different GPUs, and that led to us realizing that maybe even disaggregated computing could make sense, that we could have different heterogeneous nature of computing. That same sensibility led us to Melonize.

    12. JC

      Yep.

    13. JH

      You know, today, Nvidia's computing is spread across GPUs, CPUs, switches, scale-up switches, scale-out switches, networking processors, and now we're gonna add Groq to that, and we're gonna put the right workload on the right chips. You know, we just really evolved from a GPU company to an AI factory company.

    14. JC

      I mean, I think that was probably the biggest takeaway that I had. You're seeing this fundamental disaggregation where we've gone from a GPU, and now you have this complexion of all these different options that will eventually exist. The thing that you guys said on stage, or you said on stage was, "I, I would like the high value inference people to take a listen to this," and 25% of your data center space you said should be allocated to this Groq LPU, GPU combo.

    15. JH

      We should add Groq to about 25% of the Vera Rubins in the G- in the data center.

    16. JC

      So can you tell us about how the industry looks at this idea of now basically creating this next generation form of disaggregated prefill, decode, disagg-

    17. JH

      Mm-hmm

    18. JC

      ... and how people, you think, will react to it?

    19. JH

      Yeah. And take a step back. And at the time that we added this, we went from large language model processing to agentic processing. Now, when you're running an agent, you're accessing working memory, you're accessing long-term memory, you're using tools, you're really beating up on storage really hard. You have agents working with other agents. Some of the agents are very large models, some of them are smaller models. Some of them are diffusion models, some of them are auto- autoregressive models. And so there are all kinds of different types of models inside this data center. We created Vera Rubin to be able to run this extraordinarily diverse workload. My sense is... A- and so we added, we used to be a one-rack company, we now added four more racks.

    20. JC

      Right.

    21. JH

      So Nvidia's TAM, if you will, increased from whate- uh whatever it was to probably something, call it, you know, 33%, 50% higher. Now, part of that 33% or 50%, a lot of it's gonna be storage processors. It's called Bluefield. Some of it will be, a lot of it I'm hoping, will be Groq processors, and some of it will be CPUs, and they're all gon- and a lot of it's gonna be networking processors. And so all of this is gonna be running basically the computer of the AI revolution called agents.

    22. JC

      Right.

    23. JH

      The operating system of, of, um-

    24. JC

      What about-

    25. JH

      ... modern, modern industry.

    26. JC

      What about embedded applications? So, you know, my daughter's teddy bear at home wants to talk to her. What goes in there? Is it a custom ASIC, or does there end up becoming much more kind of a broader set of TAM with developing tools that are maybe different for different use cases at the edge and in an embedded application set?

    27. JH

      We think that there's three computers in the problem, at the, at the largest, at the largest scale when you stick- take a step back. There's one computer that's really about training the AI model, developing, creating the AI, another computer for evaluating it. Depending on the type of problem you're having, like for example, you look around, there's all kinds of robots in cars and things like that. You have to evaluate these robots inside a virtual gym that represents the physical world. So it has to be software that obeys the laws of physics, and that's a second computer. We call that Omniverse. The third computer is the computer at the edge, the robotics computer.

    28. JC

      Right.

    29. JH

      That robotics computer, one of them could be self-driving car, another one's a robot, another one could be a teddy bear.Little tiny one for a teddy bear. One of the most important ones is one that we're working on that basically turns the telecommunications base stations into part of the AI infrastructure. So now all of the... It's a $2 trillion industry. All of that in time will be transformed into an extension of the AI infrastructure. And so radios, radios will become e- edge devices. Factories, warehouses, you name it. And so, so there are three, these three basic computers, all of them, you know, are gonna be necessary.

    30. DS

      Jensen, last, uh, last year I think you were ahead of the, the rest of the world in, in, in saying inference isn't gonna 1000X-

  3. 9:27 – 11:22

    Decision making at the world's most valuable company

    1. JH

      In a final analysis, that's the job of the CEO.

    2. JC

      Yeah.

    3. JH

      And our job is to define the strategy, define the vision, define the strategy. We're informed of course by amazing computer scientists, amazing technologists, great people all over the company, but we have to shape the future. Well, part of it has to do with is this something that's insanely hard to do? If it's not hard to do, we should back away from it, and the reason for that i- if, if it's easy to do obviously, um-

    4. JC

      Lots of competitors

    5. JH

      ... yeah, a lot of competitors.

    6. JC

      Yeah.

    7. JH

      Is this something that has never been done before that's insanely hard to do, and that somehow taps into the special superpowers of our company? And so I have to find this confluence of things to, that meets the standard. And in the end we also know that a lot of pain and suffering's gonna go into it.

    8. JC

      Yeah.

    9. JH

      There are no great things that are invented because it was just easy to do and just like first try, here we are. And so if it's super hard to do, nobody's ever done it before, it's very likely that you're gonna have a lot of pain and suffering.

    10. JC

      Can you-

    11. JH

      And so you better enjoy it.

    12. JC

      So can you, can you just look at maybe three or four of the more long tail things you announced, and just talk about the long-term viability of whether it's the data centers in space, or whether it's what you're trying to do with ADAS in autos, or, you know, what you're trying to do on the biology side. Just give us a sense of like how you see some of these curves inflecting upwards in some of these longer tail businesses.

    13. JH

      I... Excellent. Um, physical AI. Large category. We believe, and I just mentioned, we have three computing systems, all the software platforms on top of it. Physical AI as a large category, it's technology industry's first opportunity to address a $50 trillion industry that has largely been, you know, void of technology until now.

    14. JC

      Mm.

    15. JH

      And so we need to invent all of the technology necessary to do that. I felt that that was a 10-year journey. We started 10 years ago. We're seeing it inflecting now. It is a multi-billion dollar business

  4. 11:22 – 17:12

    Physical AI's $50T market, OpenClaw's future, the new operating system for modern AI computing

    1. JH

      for us. It's close to $10 billion a year now. And so it's a big business and it's growing exponentially. And so that's number one. I think in the case of digital biology, I think we are literally near the ChatGPT moment of digital biology.

    2. JC

      Yeah.

    3. JH

      We're about to understand how to represent genes, proteins, cells. We already know how to understand chemicals. And so the ability for us to represent and understand the dynamics of the building blocks of biology, that's a couple, two, three, five years from now. In five years' time, I completely believe that the healthcare industry or digital biology is gonna inflect. And so these are a couple of the really great ones, and you could see they're all around us.

    4. JC

      Agriculture.

    5. JH

      Agriculture.

    6. JC

      Inflecting now.

    7. JH

      No question. Yeah.

    8. DS

      Jensen, uh, I wanna take you from the data center to the desktop.

    9. CP

      Uh, the company was built in large part on hobbyists, video gamers and, and all those graphic cards in the beginning. And you mentioned in front of, I think ten thousand people here, just Claude, uh, OpenClaw, Claude Code, and what a revolution agents have become. And specifically the hobbyists who are really where a lot of energy, um, we see, you know, a lot of the innovation breaks, want desktops. You announced one here, uh, I believe it's the Dell 6800. Uh, this is a very powerful workstation to run local models, 750 gigs of RAM. Obviously, the, the Mac, uh, Studio sold out everywhere. In my company, we're moving to OpenClaw everything. Friedberg just got Claw-pelled. You got Claw-pelled, I understand that you're obsessed with these. What is this from the streets movement of creating open source agents and using open source on the desktop mean to you?

    10. JH

      So great.

    11. CP

      Where is that going?

    12. JH

      Yeah. So great. First of all, let's take a step back. Um, in the last two years, we saw basically three inflection points. The first one was generative. ChatGPT brought AI to the common everybody, to our awareness. But the fact of the matter is the technology sat in plain sight months before GPT. It wasn't until ChatGPT put a user interface around it, made it easy for us to use, that generative AI took off. Now, generative AI, as you know, generates tokens for internal consumption as well as external consumption.

    13. CP

      Yeah.

    14. JH

      Internal consumption is thinking, which led to reasoning. O1 and O3 continued that wave of ChatGPT, grounded information, made AI not only answer questions, but answer questions in a more grounded way useful. We started seeing the revenues and the ec- the economic model of OpenAI start to inflect. Then the third one was only inside the industry that we saw, Claude Code, the first agentic system that was very useful.

    15. CP

      Yeah.

    16. JH

      Really revolutionary stuff. But un- but Claude Code was only available for enterprises. Most people outside never saw anything about Claude Code until-

    17. CP

      OpenClaw

    18. JH

      ... OpenClaw. OpenClaw basically put into the po-popular consciousness what an AI agent can do.

    19. CP

      Mm-hmm.

    20. JH

      That's the reason why OpenClaw is so important from a cultural perspective. Now, the second, second reason why it's so important is that OpenClaw is open, but it formulates, it structures a type of computing model that is basically reinventing computing altogether. It has a memory system. It scratch... It has short-term memory file system. It has a-

    21. CP

      Skills.

    22. JH

      It has [laughs] It has scales. Did you say skills or scales?

    23. CP

      Skills.

    24. JH

      Oh, skills.

    25. CP

      They do have scales-

    26. JH

      Yeah [laughs]

    27. CP

      ... theoretically, yeah.

    28. JH

      Yeah.

    29. CP

      Skills. Yeah.

    30. JH

      So, so the first thing, first thing it, it, you know, it has resources. It, it manages resources. It's, it does scheduling.

  5. 17:12 – 21:22

    AI's PR crisis, refuting doomer narratives, Anthropic's comms mistakes

    1. JH

      is, what it is not.

    2. CP

      Mm-hmm.

    3. JH

      It is not a biological being. It is not alien. It is not conscious. Um, it is computer software.

    4. CP

      Yeah, exactly.

    5. JH

      And, and, and it is not something that, um, we say things like we don't understand it at all.

    6. CP

      Right.

    7. JH

      It is not true we don't understand at all. We understand a lot of things about this technology. And, and so, so I think, one, we have to make sure that we continue to inform the policymakers and not affect, not allow doomerism and extremism to affect how policymakers think and understand about this technology. However, however, we still have to recognize this technology is moving really fast and don't get policy ahead of the technology too quickly. And the risk that we, we run as a nation, our greatest source of national security concern with respect to AI, is that other countries adopt this technology while we are so angry at it or afraid of it or somehow paranoid of it that-Our industries, our society don't take advantage of AI.

    8. DS

      There's, uh-

    9. JH

      So I'm just mostly worried about the diffusion of AI here in the United States.

    10. DS

      Yeah.

    11. JC

      Can you just double-click, if you were in the seat in the boardroom of Anthropic over that whole scuttlebutt with the Department of War, it sort of builds on this idea of people didn't know what to think. It's sort of added to this layer of either resentment or fear or just general mistrust that people have sometimes at the software levels of AI. What would, would, do you think you would have told Dario and that team to do maybe differently to try to change some of this outcome and some of this perception?

    12. JH

      The first thing that I, I would, I would say about Anthropic is, first of all, the technology's incredible.

    13. DS

      Incredible.

    14. JH

      We are a large consumer of Anthropic technology.

    15. DS

      Yeah.

    16. JH

      Really admire their focus on security, really admires their focus on safety. Um, the, the, the, the, the culture by which w- they went about it, the, the technology excellence by which they went about it, really fantastic. Um, I, I would say that, that the, the desire to warn people about the capability of the technology is, is also, uh, really terrific. We just have to make sure that we understand that the world has a spectrum and that, that warning is good, scaring is less good.

    17. JC

      [laughs] Right.

    18. JH

      Um, and because this technology is too important to us.

    19. DS

      True.

    20. JC

      Right.

    21. JH

      And, and I think that it is fine to, uh, predict the future, but we need to be a little bit more circumspect. We need to have a little bit more humility that, in fact, we can't completely predict the future. And the abil... And to say things that, that are quite extreme, quite catastrophic, that there's no evidence of it happening, um, could be more damaging than people think. And, and of course, we are technology leaders. Uh, there were, there was a time when nobody listened to us.

    22. DS

      Yeah.

    23. JH

      Um, but now because technology is so important in the social fabric, such an important industry, so important to national security, our words do matter. And I think we have to be much more circumspect. We have to be more moderate. We have to be more balanced. We have to be more th- more thoughtful.

    24. DS

      Well, I, you know, I would nominate you. I think the industry's got to get together. Seventeen percent popularity of AI in the United States. I mean, we see what happened to nuclear, right? We basically shut down the entire nuclear industry, and now we have a hundred fission reactors being built in China and zero in the United States. Um, we hear about moratoriums on data centers, so I think we have to be a lot more proactive about that. But, but I want to go back to this agentic explosion that you're seeing inside your company, the efficiencies, the productivity gains inside your company. There's a lot of debate whether or not we're seeing ROI, right? And you and I entering t-into this year, the big question was, are the revenues going to show up? Are the revenues gonna scale like intelligence? And then we had this kind of Oppenheimer moment, a five, six billion dollar month by Anthropic in February. Um, do you think as you look ahead, you announced a trillion-dollar, you know, visibility into a trillion dollars of just Blackwell and Vera Rubin over

  6. 21:22 – 31:24

    Revenue capacity, token allocation for employees, Karpathy's autoresearch, agentic future

    1. DS

      the course of the next couple years. When you see this happening at Anthropic and OpenAI, do you think we're on that curve now where we're going to see revenues scale in the way that intelligence is scaling?

    2. JH

      When you look around, when you... I'll answer this a couple different ways. When you look around this audience, you will see that Anthropic and OpenAI is represented here, but in fact, every but ninety-nine percent of everything that is here-

    3. DS

      Right

    4. JH

      ... is all AI, and it's not Anthropic and OpenAI.

    5. DS

      Right.

    6. JC

      Right.

    7. JH

      And the reason for that is because AI is very diverse.

    8. DS

      Mm.

    9. JH

      I would say that the second most popular model as a category is open models.

    10. DS

      Open source.

    11. JH

      Number one is, yeah, open-

    12. DS

      Open weights

    13. JH

      ... open source. Open weights, open source. OpenAI is number one. Open source is number two. Very distant third is Anthropic, and that tells you something about the scale of all of the AI companies that are here. And so, so it's important to recognize, recognize that. Um, let me, let me come back and say a couple things. One, when we went from generative to reasoning, the amount of computation we needed was about a hundred times.

    14. DS

      Right.

    15. JH

      When we went from reasoning to agentic, the computation is probably another hundred times. Now we're looking at, in just two years, computation went up by a factor ten thousand X. Meanwhile, people pay for information, but people mostly pay for work.

    16. DS

      Right.

    17. JC

      Yes.

    18. JH

      Talking to a chatbot and getting an answer is super great.

    19. DS

      Right.

    20. JH

      Helping me do some research, unbelievable. But getting work done, I'll pay for.

    21. DS

      Indeed.

    22. JC

      And yeah-

    23. JH

      And so that's where we are.

    24. DS

      Yeah.

    25. JH

      Agentic systems get work done.

    26. JC

      That's-

    27. JH

      They're helping our software engineers get work done. And, and so then you take that, you got ten thousand X more compute. You get probably, at this point, a hundred X more consumption now.

    28. DS

      Yes.

    29. JC

      Yeah.

    30. JH

      And we haven't even started scaling yet. We are absolutely at a million X.

  7. 31:24 – 40:19

    Open source, global diffusion, Iran/Taiwan supply chain impact

    1. JC

      crazy technical accomplishment.

    2. JH

      Yeah.

    3. JC

      Because it's like random people and-

    4. JH

      Yeah

    5. JC

      ... each person gets a little share.

    6. JH

      Yeah. Our, our modern version of folding at home.

    7. JC

      Ex-exactly.

    8. JH

      Yeah.

    9. JC

      So what, what do you think about the end state of open source? Do you see this decentralization of architecture as well, and decentralization of compute to support open weights and a totally open source approach to making sure AI is broadly available to everyone?

    10. JH

      I believe we fundamentally need models as a first-class product, proprietary product, as well as models as open source. These two things are not A or B, it's A and B. There's no question about it. And the reason for that is because models is a technology, not a product. Model is a technology, not a service. For the vast majority of consumers, the horizontal layer, the general intelligence, I would really, really love not to go fine-tune my own.

    11. JC

      Right.

    12. DS

      Right.

    13. JH

      I would really love to keep using ChatGPT. I love to use Claude. I love to use Gemini. I love to u-use X. And they all have their own personalities, as you know, which just kind of depends on my mood and depends on what problem I'm trying to solve. You know, I might, you know, do it on X or I might do it on, on ChatGPT. And so that, that segment of the, of the industry is thriving. It's gonna be great. However, they're-- all these industries, their domain expertise, their specialization has to be channeled, has to be captured in a way that they can control, and that it can only come from open models. The open model industry we're contributing tremendously to, it is near the frontier, and quite frankly, even if it reaches the frontier, I think that products as a service, world-class products as, as a pro-- models as a product is gonna continue to thrive.

    14. JC

      Every startup we're investing in now is open source first, and then going to the proprietary models.

    15. JH

      Yeah. And the beautiful thing is because you have a great router, you connect the two by on, on first day, every single day, you're gonna have access to the world's best model and, and then it gives you time to cost reduce and fine-tune and specialize. And so you're gonna have world-class capabilities out the chute every single time.

    16. JC

      Let-

    17. DS

      Jensen, can I ask you this?

    18. JC

      Go ahead. Of course.

    19. DS

      Nobody wants the US to win the global AI race more than you, right? But o-- a year ago, the Biden era diffusion rule really was an anti-American diffusion of AI around the world. So here we are, a year into the new administration. Give us a grade. Where is-- where are we in terms of global diffusion and the rate at which we're spreading US AI technology around the world? H-are we an A, are we a B, are we a C? What is-- what's working? What's not working?

    20. JH

      Well, first of all, President Trump wants American industry to lead. He wants American technology industry to lead. He wants American technology industry to win. He wants us to s-spread American technology around the world. He wants United States to be the wealthiest country in the world. He wants all of that. At the current moment, as we speak, Nvidia gave up a ninety-five percent market share in the second-largest market in the world, and we're at zero percent. President Trump-

    21. DS

      China.

    22. JH

      That's right. President Trump wants us to get back in there and, and, uh, the first thing is, uh, to get license, licensed for the companies that we're gonna be able to sell to. We've got many companies who have requested for licenses. We've applied for licenses for them, and we've got approved licenses from Se-Secretary Hou- uh, Lutnick. Uh, now, uh, we've, we informed the Chinese companies, and many of them have given us purchase orders. And so we're gonna st- we're gonna-- we're in the process of cranking up our supply chain again to go ship. I think at the highest level, Brad, um, I think one of the things that we should acknowledge is this: Our national security is diminished when we don't have access to miniature motors, rare earth minerals. It's diminished when we don't control our telecommunications networks. It's diminished when we can't provide for sustainable energy for our country. It is fundamentally diminished. Every single one of these industries is an example of what I don't want the AI industry to be.

    23. DS

      Right.

    24. JH

      When we look forward in time and we say, "What do we want? What is the-- what does it look like when American technology industry, American AI industry leads the world?" We can all acknowledge that there is no way that AI models is one universally. It is-- we can all acknowledge that that is an outcome that makes no sense. However, we can all imagine that the American tech stack, from chips to computing systems to the platforms, are used broadly by the world where they build their own AI, they use public AI, they use private AI, whatever, and they can build their applications in their society. I would love that the American tech stack is ninety percent of the world.

    25. DS

      Yes.

    26. JH

      I would love thatThe alternative, if it looks like solar, rare earth, m- magnets, motors, telecommunications, I consider that a very bad outcome for national security.

    27. JC

      Agreed.

    28. JH

      Yeah.

    29. JC

      Yeah.

    30. CP

      How much are you monitoring the situation with the conflicts around the world right now, and how much does it worry you, Jensen? So China and Taiwan, and then helium availability coming out of the Middle East I understand can be a supply chain risk to semiconductor manufacturing. How much do these sit- situations worry you? How much are you spending on them?

  8. 40:19 – 48:06

    Self-driving platform, facing competition from active customers, responding to growth slowdown predictions

    1. CP

      an iOS with Tesla or Waymo. What's your strategy thinking there, and how that chessboard emerges? Because it feels like you have a, a pretty deep stack, and in some ways you're competing, and in other places you're collaborative.

    2. JH

      Yeah. Um, it's-- Taking a step back, we believe that everything that moves will be autonomous completely or partly someday, number one. Number two, we don't wanna build self-driving cars, but we wanna enable every car company in the world to build self-driving cars. And so we built all three computers: the training computer, the simulation computer, the valuation, evaluation computer, as well as the car computer. We developed the world's safest driving operating system. Uh, we also created the world's first s- reasoning autonomous vehicle so that it could decompose complicated scenarios into simpler scenarios that it knows how to navigate through, just like us, reasoning systems. And so that reasoning system, called Almaio, has enabled us to achieve incredible results. We open this pl- we ver- uh, we vertical optimization, we horizontally innovate, and we let everybody decide, do you wanna buy one computer from us? In the case of Elon and Tesla, they buy our training computers. Um, do they wanna buy our training computer and our simulation computers, or do you wanna let us, uh, work with us to do all three, and even put the car computer in your car? So we, you know, I... Our attitude is, we wanna solve the problem. We're not the solution provider, and we're delighted-

    3. JC

      But, but, let me-

    4. JH

      However you work with us.

    5. JC

      Let me build on this question because I think it's, like, it's so fascinating. You actually do create this platform. A thousand flowers are blooming, but it's also true that some of those flowers want to now go back down in the stack and try to compete with you a little bit. Google has TPU, Amazon has Inferentia and Trainium. You know, everybody's sort of spinning up their own version of-

    6. JH

      Yeah

    7. JC

      ... I think I can out Nvidia Nvidia, even though they also tend to be huge customers.

    8. JH

      Yeah.

    9. JC

      How do you navigate that? And-

    10. JH

      Yeah

    11. JC

      ... what do you think happens over time, and where do those things play in the complexion of this kind of vision?

    12. JH

      Yeah, [chuckles] really great. You know, first of all, um, we're the only AI company-- We're an AI company. We build foundation models. We're at the frontier in many different domains. We build every single-Every single layer, every single stack. Um, we're the only AI company in the world that works with every AI company in the world. They never show me what they're building, and I always show them exactly what I'm building.

    13. JC

      Right.

    14. JH

      Yeah. And so, so the, the confidence comes from this. One, uh, we are delighted to compete on what is the best technology, and to the extent that, to the extent that we can continue to run fast, I believe that buying from Nvidia still is one of the most economic things they could do, and that's just incredible confidence there. Number one. Number two, we're the only architecture that could be in every cloud, and that gives us some fundamental advantages. We're the only architecture you could take from a cloud and put into on-prem, in the car, in any region.

    15. JC

      In space.

    16. JH

      That's right, in space. And so there's a whole s- whole part of our market, about 40% of our mar- of our business, most people don't realize this, 40% of our business, unless you have the CUDA stack, unless you can build an entire AI factory, you have-- the customers don't know what to do with you. They're not trying to build chips. They're not trying to buy chips. They're trying to build AI infrastructure, and so they want you to come in with a full stack, and we've got the whole stack. And so surprisingly, Nvidia is gaining market share. If you look at where we are today, we're gaining share.

    17. JC

      Do you think what happens is these guys try, and they realize, "Oh my God, it's too much," and then they come back? Is that why the share grows?

    18. JH

      Well, we're gaining share for several reasons. One, um, our velocity has gone-- We help people realize it's not about building the chip, it's about building the system.

    19. JC

      System, yes. Yeah.

    20. JH

      And that system's really hard to build. Uh, and, and so their, their, their business with us is increasing. In the case of AWS, I think they just announced, I think it was yesterday, that they're gonna buy a, a million chips, uh, in the next couple of years. I mean, that's a lot of chips from, from AWS, and that's on top of all the chips they've already bought, and so we're delighted to do that. But number one, we're gaining share this last couple of years because we now have Anthropic coming to Nvidia, Meta SL is coming to Nvidia, and the growth of open models is incredible, and that's all on Nvidia. And so we're growing in share because of the number of models. We're also growing in share because outs-- all of these companies are outside the cloud, and they're growing regionally in enterprise, in industries, at the edge, and that entire segment of growth is, you know, really hard to do if it's just building an ASIC.

    21. JC

      Brad? Yeah.

    22. DS

      Related to that, um, and not to get in the weeds on the numbers, but analysts don't seem to believe, right? So if you look at the consensus forecast, you said compute could one million X, right? And yet they have you growing next year at thirty percent, the year after that at twenty percent, and in twenty twenty-nine, which is supposed to be a monster year, at seven percent, right? So if you just-- if you take your TAM and you apply their growth numbers, it suggests that your share will plummet. Do you see anything in your future order book that would make that correct?

    23. JH

      Yeah, first of all, they, they just don't understand the scale and the breadth of AI.

    24. DS

      Yes.

    25. JC

      Yeah. Yeah, I think that's true.

    26. JH

      Most people think that AI is in the top five hyperscalers.

    27. DS

      Right.

    28. JC

      That's right. There's also an orthodoxy around these law of large numbers where, you know, they have to go back to their investment banking risk committee and show some model. They're not gonna believe in their minds that-

    29. DS

      Well, it's-

    30. JC

      ... five trillion goes to fifteen trillion.

  9. 48:06 – 56:44

    Datacenters in space, AI healthcare, Robotics

    1. JH

      requires very large surfaces. And so-

    2. JC

      Right

    3. JH

      ... now that's not an impossible thing to solve, and there's a lot of sp- lot of space in space. Um, but nonetheless, the expense is still quite there, is, is there. Uh, we're gonna go explore it. We're already there. We're already radiation hardened. Uh, we have, we have, uh, uh, uh, CUDA in satellites around the world. Um, they're doing imaging, image processing, AI imaging, and, um, and that kind of stuff ought to be done in space instead of sending all the data back here and do imaging down here. We ought to just do imaging out in space. And so there's a lot of things that we ought to done, do, do in space. And in the meantime, uh, we're gonna explore what does the architecture of data centers look like, uh, in space? And it'll take, it'll take years. It's okay.

    4. JC

      I wanna do-

    5. JH

      I got plenty of time

    6. CP

      I wanted to, um, double-click on healthcare. I know you've got a big effort there.

    7. JH

      Yeah.

    8. CP

      We're all of a certain age where we're thinking about [laughs] lifespan, healthspan. I mean, we all look great, I think.

    9. JH

      [laughs] Some better than others.

    10. CP

      I think some better than others.

    11. JH

      [laughs]

    12. CP

      I don't know what your secret is, Jensen.

    13. DF

      Look pretty good these, these days.

    14. CP

      I mean, uh, what, what's-

    15. JH

      He's Asian

    16. CP

      ... what are you taking?

    17. JH

      He's Asian.

    18. CP

      What's off the menu? You gotta talk to me when we're backstage. I wanna know in the green room what you've got going on.

    19. JH

      S- squats, squats and pushups and surfs.

    20. CP

      Perfect.

    21. JH

      Yeah.

    22. CP

      Okay. Um, but-

    23. JH

      That works

    24. CP

      ... what you know in terms of the build-out in healthcare, where is that going a- and what kind of progress are we making? I was just using Claude to do some analysis and saying, like, where are all these billing codes? We spend twice as much money in the US, we get, seem to get half as much. It seemed like, uh, 15 to 25% of the dollars spent were on these first GP visits, and I think we all know, like, ChatGPT and a large language model does a better job more consistently today at a first visit. So what has to happen there to kinda break through all that regulation and have AI have a true impact on the healthcare system?

    25. JH

      There's several w- several areas that we're involved in, in, um, in healthcare. One is, uh, AI, uh, physics. Uh, and, and that's... Or AI biology. Using AI to understand, represent, predict biology behavior, biological behavior. And so that's one. That's very important in drug discovery. There's second, which is AI agents, and that's where the assistance and helping diagnosis and things like that. OpenEvidence is a really good example.

    26. CP

      Yeah.

    27. JH

      Hippocratic is a really good example. Love working with those companies. Um, I really think that this is an area, uh, where agentic technology is gonna revolutionize how we interact with doctors and how do we interact for healthcare. The third part that we're in, e- involved in is physical AI. The first one's AI physics, using AI to predict physics. The second one is physical AI, AI that understand the properties of the laws of physics, and that's used for a- uh, robotic surgery. Huge amounts of activities there. Every single instrument, whether it's ultrasound or, you know, CT or whatever instrument we interact with in a hospital in the future will be agentic.

    28. CP

      Yeah.

    29. JH

      You know? OpenClaw in a safe version-

    30. CP

      Yeah

  10. 56:44 – 59:38

    OpenAI/Anthropic revenue potential, how to build an AI moat

    1. JH

      enterprise software company will also be a reseller, value-added reseller of Anthropic code, Anthropic's tokens.

    2. DS

      Yeah.

    3. JH

      Value-added reseller-

    4. DS

      Have to be

    5. JH

      ... of OpenAI.

    6. DS

      Have to be.

    7. JH

      Value... That's right.

    8. DS

      Yeah.

    9. JH

      And they're gonna... That, that, that part of their-

    10. DS

      Get this logarithmic expansion.

    11. JH

      Yes.

    12. DS

      Yeah.

    13. JH

      Their go-to market is going to expand tremendously-

    14. JC

      What do you think-

    15. JH

      ... this year

    16. JC

      ... what do you think in that world is the moat? What's left over? I mean, you have some moats that are, frankly I think as this scales, almost insurmountable. The best one that nobody talks about is probably CUDA, which is just, like, an incredible strategic advantage. But in the future, if a model can be used to create something incredible, then the next spin of a model can be used to maybe disrupt it. Sort of in your mind, what do you think for these companies that are building at that application layer, what's their moat? Like, how do they differentiate themselves?

    17. JH

      Deep specialization. Deep specialization. I believe that, um, these models, they're, they're gonna have general, general models that are connected into the software company's agentic system.

    18. DS

      Right.

    19. JH

      Many of those models are cloud models and proprietary models, but many of those models are specialized sub-agents that they've trained on their own d-

    20. JC

      Right. So the call to arms for you, for entrepreneurs is, look, know your vertical.

    21. JH

      That's right.

    22. JC

      Know it as deep and as better than everybody else-

    23. JH

      That's right

    24. JC

      ... and then wait for these tools because they're catching up to you, and now you can imbue it with your knowledge-

    25. JH

      That's right

    26. JC

      ... and your edge.

    27. JH

      The sooner you connect your agent-

    28. JC

      Yeah

    29. JH

      ... the sooner you connect your agent with customers, that flywheel is gonna cause your agent to get hyper-

    30. JC

      It very, very much is an inversion of what we do today, because today we build a piece of software and we say, "What generalizes?" And then, "Let's try to sell it as broadly as possible, and then sell the customization around it."

  11. 59:38 – 1:06:05

    Advice to young people on excelling in the AI era

    1. JH

      just because he doesn't hang out with me enough.

    2. DS

      [laughs]

    3. JC

      [laughs]

    4. DS

      But we, I mean, we've hung a little bit.

    5. JC

      Be careful what you... [laughs]

    6. DS

      We don't talk about it.

    7. JC

      He will show up at your breakfast table next-

    8. DS

      We don't talk about it.

    9. JC

      He'll follow you around.

    10. JH

      I'm not asking for it, I'm just saying.

    11. JC

      He will follow you around.

    12. DS

      I mean-

    13. JH

      I'm not asking for it. [laughs] I'm just saying.

    14. DS

      You can come with me and Tucker, we ski in Japan every January.

    15. JC

      Really?

    16. JH

      Oh, boy.

    17. DS

      Love it.

    18. JH

      Oh, boy.

    19. DS

      Me and Tucker will go road trip.

    20. JH

      Oh, boy. Wow. Okay.

    21. DS

      Um, no, there is gonna be-

    22. JC

      No comment

    23. DS

      ... job displacement.

    24. JC

      Fair enough.

    25. DS

      And then the question becomes-

    26. JH

      Yeah

    27. DS

      ... you know, do those people have the fortitude, the resolve to then go embrace these, you know, technologies? We're, we're going to see 100% of driving go away by humans. That's just... It's, that's a beautiful thing in the lives saved, but we have to recognize that's 15 million people in the United States, 10 to 15 million, who are employed in that way. And, and so that is gonna happen, yes?

    28. JH

      I, I think, I think that jobs will change. For example, um, there are many chauffeurs today, uh, who drives the car. I believe that though many of those chauffeurs will actually be in the car sitting behind the drive- the steering wheel while the car is driving by itself. And the reason for that is because remember what a chauffeur does. In the end, these chauffeurs, they're helping you, they're your assistants, they're helping you-

    29. DS

      Yes

    30. JH

      ... with your luggage, they're helping you, I mean, they're helping you with a lot of things. And, and so I wouldn't be surprised actually if the chauffeurs of the future become your mobility assistant-

Episode duration: 1:06:05

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode gwW8GKwHB3I

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome