# Trailblazing the Future of AI, Marc Benioff and Jensen Huang Auto-transcribed by https://aliceapp.ai on Wednesday, 18 Sep 2024. Synced media and text playback available on this page: https://aliceapp.ai/recordings/mn4gCsdTG5kWprVh2ZFXJObNaK97ecjY * Words : 6,895 * Duration : 00:41:49 * Recorded on : Unknown date * Uploaded on : 2024-09-18 00:13:37 UTC * At : Unknown location * Using : Uploaded to aliceapp.ai ## Speakers: * Speaker A - 37.84% * Speaker B - 62.16% ---------------------------- Speaker A [00:00:03] Okay. Good afternoon, everybody. How y'all doing? All good. Do you like the keynote? Oh, wow. Fantastic. All right, all right. Are you gonna diy your AI? Oh, wow. Well, there we go. I guess I got through it, everybody. Fantastic. Good answer on that one. Wow, big energy. Okay, well, please welcome, uh, Jensen. Wong Jensen. Come on out. Great to have you. Great to have you. Glad you're here. Speaker B [00:00:41] All right, Agent Force. Speaker A [00:00:44] Okay. Speaker B [00:00:45] You know, our first product was called G Force. Speaker A [00:00:48] Oh, right, that's right. I forgot about that. Speaker B [00:00:51] G force made it possible. Speaker A [00:00:52] Do we owe you a royalty on that? Jeez. Speaker B [00:00:56] You're welcome to. Speaker A [00:00:57] I had no idea g force made. Speaker B [00:01:01] Possible cuda, which made possible AI, which made possible agent force. Speaker A [00:01:14] Congratulations on that. Speaker B [00:01:16] Look at all these people in the audience. Speaker A [00:01:17] You know, some of these people down. Speaker B [00:01:19] There, all these people. Speaker A [00:01:21] It's like a, uh. This is your life moment, right? I hope you want to tell this. Speaker B [00:01:25] Is your life moment. Speaker A [00:01:26] Should we tell them of the inside? Speaker B [00:01:27] I'm so young. This is not my life. I'm just getting started. Speaker A [00:01:30] Anyway, I just, uh, I invited because my neighbor, uh, in, uh, Hawaii, in, uh, Waimea, is David Kirk, who was a chief scientist. Speaker B [00:01:39] Chief scientist. My first chief scientist right there. Speaker A [00:01:41] And, uh, David and I put the AI in Waimea. David, stand up and just take a hand. And part of the team, I think, that created the GPU, if I remember correctly. Speaker B [00:01:56] Indeed. Speaker A [00:01:57] Amazing what a lineage and an incredible story Nvidia is. And I mean, the company now, more than four decades of success. And, uh, you know, when you started the company and looking back, um, you know, was this your vision that you would get to this point in technology or what was your initial vision for the company and what was your greatest hope, inspiration at that moment. Speaker B [00:02:25] We knew today was going to come all along. We called it to the day. Uh, we did a couple things right. We observed that general purpose computing, or CPU's, would be good at many things, but it wouldn't be good at everything. And that there would be interesting problems we could solve, uh, that wouldn't be practical without some other form of computing augmentation. Just as we're going to augment humans with agents, we augmented the cpu with a GPU with an accelerating function. And we were smart to observe that if we wanted to be a computing platform, it had to be architecturally compatible over time. And so, uh, we made several good decisions. We, uh, focused the company on accelerated computing. We focused the company on architecture. Uh, we dedicated ourselves to, uh, developers, uh, and creating tools and, uh, evangelizing our architecture. To as many people as possible. Uh, we were smart to select some good problems to solve. The first one we selected was one, uh, that was computationally intensive and also extremely high volume video games. Uh, that's right, GeForce. And uh, many of those video gamers eventually, uh, grew up to be computer scientists. We grew up to be uh, AI computer scientists who uh, eventually led to today. So um, we had it figured out all along. Speaker A [00:04:02] I remember when I first started building uh, software, which was in around 1970, 919 80, when we had the Atari 800, we had player missile graphics. Uh-huh and that was kind of the beginning of the idea that graphics could be built in to the computer processor. But you've taken it so much farther. And I remember when I bought my Nintendo switch, I was like, jesus, if it can do this, what else can it do? But in my wildest dreams, I never thought it could do what it's doing now. At what point did you realize, well, what we have built can do so much more? Speaker B [00:04:39] Well, you know, accelerated computing is about, about taking um, computationally, ah, intractable problems for CPU's and somehow doing it incredibly fast or cheaply, or with lower energy or at a higher scale. And so the fundamental tenant of the company was selecting problems that this computer architecture can solve. And so we're always out looking for uh, the next great computing challenge. One of the earliest ones for CUDA was seismic processing. And uh, it's an inverse physics problem. Another one was molecular dynamics for virtual screening. Uh, all kinds of really, really complicated algorithms, uh, were ported onto cuda. Uh, and we were able to make, uh, the uh, problem much more tractable in a lot of ways. We put a supercomputer in everybody's PC. We democratize high performance computing. And that phenomenon led to, led to. Speaker A [00:05:39] Researchers, uh. Speaker B [00:05:41] Uh, reaching out to us, asking us to help them, uh, factor, uh, refactor, an algorithm called deep learning to run well on Cuda. And that was our earliest introduction to uh, deep learning. Uh, the thing that was really, really uh, a good insight of the company was recognize that deep learning was used for computer vision, but it's, its effectiveness should be scalable, that these deep networks should be possible to make deeper, it could be possible to make them larger with more parameters and train them with more data. Um, and that observation led us to um, imagine that maybe deep learning would change, uh, what software could do and how software would be done altogether. That maybe uh, maybe in the future it's going to become a universal function approximator, that we give it lots of examples and it would learn to function. And so that observation, I think, was a, was a, was a fortunate observation. And we had the, we had the, uh, conviction, uh, uh, to re, uh, engineer every single stack of computing as results. And so the way we designed our processors, our GPU changed, we added, uh, tensor cores and tensor processors to it. Uh, we realized that, uh, it would be incredibly large. So we created very large systems and the world's first, uh, DGX AI supercomputer. And I delivered the world's first DGX, uh, just right up the street to start a company at the time, OpenAI. And uh, we changed the way that software was done. We changed the way that networking would, uh, connect all these systems together and invented what is now known as modern AI superclusters and AI supercomputers. And so that was really, uh, stems from a first principle observation about why DNA work and how far can it scale. And if it scaled it to that level, what are the initiations to, uh, computer science to our industry, the type of problems we could solve. And so we, uh, just lead this so deeply, we just went all in on it. And just as Mark, when Salesforce was first founded, uh, uh, you felt so deeply about taking out and taking advantage of the scalable resources of the cloud, you put everything behind it. You were one of the first software companies to be cloud only. Not only just cloud first, but cloud only in a lot of ways. Uh, we saw the same thing, what we believed. And then you've got to, you know, if you believe something, the question what are you going to do about it? And so we just went all in. Speaker A [00:08:27] It reminds me a lot of one of our core philosophies and I'm sure they probably have a whole series of these at Nvidia as well. But, you know, we really believe that tactics dictate strategy over time. And I think that when you get into a winning tactic like deep learning, all of a sudden you're, you have this core technology, you're worried on these advanced computing possibilities. You're looking at what's happening in computing, what's happening at massive scale. We're looking at all these kind of interesting things at that exact point where there were massively parallel supercomputers. But then it was like, is this really where it's going to go? Or is it going to break apart into many pieces? That idea that deep learning, all these guys at Stanford are doing some of these incredible work. And then you decided to build the key infrastructure for deep learning. That was a key insight. But it was also a leadership moment too, for your perspective that you really said you were going to build that. I don't think anyone else in the industry, everybody is watching deep learning, everybody saw the vision breakthroughs and all of the different application breakthroughs. But you were the only company that said actually we're going to monetize this, we're going to really enable the next generation of it and help to then scale and move forward. Is that the right way to think about it? Speaker B [00:09:45] Yes, I think so. And I think at the core of it, Mark, when you see something um, uh, surprisingly effective, uh, what did you see and what can it do? Are of course uh, simple questions to ask. Um, but why is it um, what is its implication and how far can it go? What does this mean? Uh, so what uh, how far can this go and what could it therefore on the second derivative impact, you know, I think we were thoughtful about asking those questions and we um, methodically, uh, uh, analyzed what impact it would have in our industry and our adjacent industries. And the thing that was, that was really, really uh, uh, exciting when you think through this, is most of the problems that, that um, are really hard to solve in the world, uh, are difficult to have a first principled fundamental scientific equation for. It's not like Newton's law, it's not like spring theory, it's not like Navier Stokes. It's hard to describe uh, these phenomenons, um, uh, using principled equations. And so uh, uh, we were excited about the possibility of using deep learning in so many different areas. Computer vision to speech understanding and so on and so forth. And when that happens, the question then becomes what new opportunities does it open up for the computer industry? And for the very first time, you know, we've been in, the two of us have been in the computer industry a long time. Our industry is, you know, call it about a trillion dollars and it's an industry of tools. You know, computers, software, um, they're industry of tools. For the very first time, this is going to be an industry of skills. And you capture that phrase and uh, you call it agents, but it's going to be for the very first time, agents sitting on top of tools, agents using tools. And the opportunity for agents is gigantic, uh, as we know, we now know. Speaker A [00:11:48] I'm hoping so, by the way. Speaker B [00:11:50] Uh, well, I'm hoping so too because you know, hopefully you're going to create hundreds of billions of dollars, billions of agents that we can hire to do, to do all kinds of things and. Speaker A [00:12:03] Augment all of our employees trying to get to our first billion first before I get to the hundreds of billions. Speaker B [00:12:07] You know, the thing about digital, as soon as you get to the first hundred, the next hundred thousand is easy. The next hundred million is easy. The next 100 billion is going to be easier than that. And so the um, uh, and you know, uh, it sounds insane, but here's the amazing thing. We're going to have agents that obviously, uh, understand the subtleties of the things that we ask it to do, but uh, it can also uh, use tools and it can reason and uh, it can reason with each other and collaborate with each other. And you know, we're going to give, we're going to give a problem to agent force and uh, uh, agents are going to go find other agents, ah, uh, that can help achieve this mission. They'll uh, work together, assemble together, work together and solve this problem. And of course, um, uh, agents will talk to agents and agents will talk to us and they'll come and ask us for feedback and we'll be able to monitor their conversation and uh, uh, improve maybe some of their dialogue so that we can, uh, improve their prompts, which improves the output. Just the same way that we lead organizations, uh, the way that we uh, uh, uh, uh, direct and fine tune the results of organizations, it's going to be pretty incredible. You uh, create the agents, I'll power it. Speaker A [00:13:28] Well, that's what's going on right now. I mean, without you, wouldn't be possible. I mean, that's the really cool thing. And I think that at Salesforce, when we look at our role in the industry, and I think kind of the symmetry between our two companies is when we looked at what's going on in the deep learning phenomenon, which was really happening, I would say when it started to get really exciting after 2010, at least, like that's when I was like, oh, this is very real. And then when we looked at what was happening at Stanford, it was just awesome. Some of the breakthroughs that were going on. And um, in 2018, Richard Socher, who was the head of our research, came in and said, hey, I've got this paper that I've written, prompt engineering. He was a deep learning researcher at Stanford with FeI, Fei Lihdenhe others. Now her husband Silvio, you know, runs our, um, research program. But I didn't have that same leap, you know, like you are going, oh, this is gonna like. I thought, you know, um, a lot of people have heard this story, but it was about ten years ago, 2013, 2014. I had an existential freak out around AI where I was like, oh, this is all gonna happen now. It's all gonna happen. And then we built Einstein. There he is on stage and we delivered on platform and yeah, he's right there, right there. And, and uh, we got into it and but uh, it was for us more about predictive and it was about machine learning. It was about machine learning, machine uh, intelligence. It was deep learning, but we didn't have the same level of, I would say, faith or commitment that it was about to transcend into this next area. But you did. So it was a leap of faith that you made at that point. Speaker B [00:15:13] Uh, we have an advantage. I mean, I have an advantage. My vantage point, of course, uh, starts all the way back to uh, working with every researcher on the planet because they need our computing infrastructure to do their research. And so I have the benefit of knowing a lot of researchers and seeing um, early indications of future success. Um, most of the um, uh, early observations and realizations uh, about deep learning and its capabilities and limitations um, really stemmed around in my case, uh, 2011. And during that time, uh, there was quite a bit of work that was being done to uh, ah, uh, train one of the first larger models, ah, for uh, computer vision. And so I think the breakthrough uh, for me was at a moment when um, we uh, realized that unsupervised learning was going to be possible because the ability for humans to uh, be able to label data at scale, uh, that we would become the limiters of these digital AI to uh, expand uh, their capabilities. And uh, when unsupervised learning came along which allowed us to use language models to create language models which codify human prior knowledge, using that to now learn multimodality, multimodal, uh, uh, data. Uh, from that point forward, the scale was going to be exponential. Now, for everybody here, uh, this is an extraordinary time because uh, in no time in history has computer technology not only, uh, uh, moved faster than Moore's law. I mean, we're Moore's law. For example, over uh, the course of uh, over the course of a decade would be about 100 x. Um, we are probably advancing at somewhere near 100. Speaker A [00:17:07] That was our other neighbor in Hawaii, Gordon Moore. Speaker B [00:17:09] Oh, is that right? Speaker A [00:17:10] Great person, great leader, great person. Speaker B [00:17:12] And so we're at a stage now, we're in an era now where we're moving way faster than Moore's law. And um, uh, arguably easily Moore's law squared. And the reason for that of course, is at every single layer, computers went from cpu's to GPU's uh, from human engineered software to machine learning software. And now this feedback loop, um, that allows us to um, create uh, new AI's. And these new AI's are helping us create new computer systems. And these new computer systems are uh, advancing at such incredible rates, which allows us to create even better AI. Uh, that feedback, that flywheel is really m flying now. And so I think, um, uh, the progress that you're going to have with agents over the next year or two is going to be spectacular and surprising. Speaker A [00:18:13] I hope so. Now look at number one. Uh, m I think you've made the case. You went from tactics to strategy. You were in this advanced computing area, and then as you kind of crossed into this deep learning area, then you were able to take the leap of faith and say, we're going to invest in this to get this tremendous outcome. Now you've said we're kind of in the flywheel zone of that, that you see this going even faster than it was going faster before. Obviously, there's a huge amount of investment going in this area. Everybody's consciousness is around. Are we going to get to the next area of AI? Do you think that we're going to hit a limit on these LLMs? That that could be a ceiling, another AI ceiling? We've seen that a couple times in our industry before. Or do you think we're just breaking through that already? Speaker B [00:18:58] Well, it's hard to say. Um, it's hard to say. I think, uh, the problems that the challenges that we have in front of us are many and worthy to be tackled right away. Of um, course, some of the most, um, uh, inspiring advances has to do with fine tuning, supervised, um. Uh, guardrailing. And uh. Um, all of the work that surrounds safety, um, everything from using AI to curate the data to create safe curriculum to teach the AI, which then fine tunes the AI on particular skills that ground it on values. Then all of this wonderful AI technology to guardrail it, um, and guardrailing and then reflection, uh, using chain of thought to reflect on the quality of the answers that it's producing. It's no longer AI, this one shot producing information it's now reasoning about is the, the uh, answer I'm generating, uh, sufficiently safe and proper and is it the best possible answer I could be providing? It's reasoning with itself and uh, it's reflecting on the answers before it answers. And so I think the progress that we're making in safety is phenomenal. We're doing the same thing in robotics. Uh, AI, um, functional safety, active safety, all kinds of great technology that's being invented in safety. I think that's one, um, the progress that we're making. Speaker A [00:20:28] Those things are related too, right? Because that robot is going to have those same exact, uh, constraints. This is one of my goals for this conference, that exactly what you just said. And also to demystify these very critical technical concepts of things like grounding, uh, fine tuning, guardrails. These are three things that are built into our platform and that we demonstrated today so that basically anybody, even someone without a technical background, non programmatic background, business user, can get in there and rapidly build a high functioning autonomous agent. That's highly accurate with very little hallucination, but you taking advantage of these key concepts and execute these things. Speaker B [00:21:15] That's right. Speaker A [00:21:16] And this should be very possible at this moment. Speaker B [00:21:19] And I love that you're demystifying it. There's a belief that AI is somehow this one giant wad of software numbers that's pulsing and growing and pulsing and growing, and it's under your house, right. But obviously, it's a system of AI's, and these systems of AI's are engineered by smart people, um, and hallucinations. Speaker A [00:21:44] It's so important, right? Isn't it so important that people can actually understand? Uh, this is what I call they need to get their hands in the soil. They need to understand the basic ideas to make this possible. And that's what I hope, you know, happens with the thousands of people here, that they can get in the code while they're here to really understand. So when they're going back to their offices, you know that exactly what you're saying, that we've demystified the AI, that, you know, they don't have to worry about how they're going to diy their AI, that they can somehow put all the hard work that your team has done right into some very practical use, because this leap of faith that you've had, the strategic motion of your company, has enabled so much of this. Now, how do we put it in the hands of everybody to, like, just get to work? And building an agent should not be some computer science, uh, fair project. It should be something that we can easily do, because it's going to be. Speaker B [00:22:43] A lot more like onboarding employees than writing software. It's going to be a lot more like introducing, welcoming a team member to help you do something. And you'll communicate with that person, with that agent, and explain what is the mission. Uh, show it examples of what the output would look like. Isn't that right? And we'll just go back and forth. And if you're not too clear about exactly what I'm asking you to help me do, m, you're going to prompt back to me and say, you're not exactly clear. And we're going to go back and forth, and I'm going to show you more examples. And so I think this idea, I. Speaker A [00:23:16] Wish I had called you before my keynote because I would have loved to use that phrase that was absolutely right. It's like onboarding employees, and it should be easy and understandable, and it shouldn't be. We should have to extract the complexity to get this to scale and go to get to your vision now of hundreds of billions of agents, you know, and the robots and I, blah, blah, blah, blah, all these things. Speaker B [00:23:39] Ah, we're going to have teammates. We're going to have agents that are excellent at particular skills. We're going to have certain agents that have access to particular types of information inside our company with access control. We're going to have agents, uh, that are more general purpose and are able to reason about things. It takes them a little longer to reason about missions and break it down into tasks. And they're going to be working together and they're going to be recruiting. You know, they're going to help us go find agents within our company who can be helpful to help them do their jobs. And so we're going to be agents working with agents and agents working with us. And, you know, we're going to supercharge the living daylights out of our company. Work is going to be done before we even think of it. What do you think about that? We're going to come to work tomorrow and a whole bunch of work that we didn't even realize needed to be done was done, huh? And then write us a nice report. It says, good job, CEO. Uh, we'll take all the credit. Speaker A [00:24:39] How many years have you been the CEO now? Speaker B [00:24:42] 30, 32 years. Speaker A [00:24:44] So you're definitely ready for that point for the work to be done before. Speaker B [00:24:48] You mark, I'm just getting started. These are, these are my finest hours. We're just getting started. This is the best. Nobody, nobody should miss the next decade. Would you agree with that? The next decade of technology advance everybody. Yeah. Don't. You're not going to want to miss this movie. Yeah, you're not going to want to miss this movie. I think the next ten years, uh, the breakthroughs that we're going to have in digital biology, the breakthroughs, uh, in just helping, uh, diagnosing, uh, disease, the breakthroughs in science. We're going to have so many scientific assistants in our company, we can't design our chips anymore without our AI. Um, digital engineer agents helping us. We, ah, have hundreds of thousands of them helping all of our engineers do our job. We can't write software without AI. We can't debug software anymore without AI. Um, all of the things that we do with, uh, supply chain planning can't be done without AI. The type of things that, you know, so whether it's in how we run companies, uh, the scientific breakthroughs, the way we educate, I, um, would think that it would be very hard, uh, to be able to go through school in the future without your own personal tutor helping you through things. And quite frankly, you know, I kind of have my own personal tutor today, right? I use chat GPT and help me, you know, think through and I reason through things with it and it, uh, helps me discover new knowledge, helps me break it down into understandable chunks in a lot of ways. Chad, GPT is my tutor, and I think a lot of people are going to, everybody is going to have their own personal tutor. Speaker A [00:26:25] I'm using it mostly as my therapist. How do you feel about that? Speaker B [00:26:29] Well, it's apparently working. You look pretty chill. And not $20 a month. Are you doing the free version or you do the. I'm doing the $20 a month version. If it's a tutor, I want to make sure I'm getting the highest quality tutorage. Speaker A [00:26:53] I don't like that. It never remembers me, you know, I don't like, every time I go to chat GPT, I feel like I'm starting over again. You know, it's like, needs a little more, uh, a little more context, a little more memory, a little more. Hey, do you remember me, what I talked about last time? Exactly. Did somebody give you an upgrade while I was away? What's happening? I like it. And then sometimes it gets me. I don't know if you ever tried this, who is Mark Benioff? And then it'll be like. I'm like, uh, no. And then there's nobody to talk to, to say, hey, that isn't right. Speaker B [00:27:25] You're not one of those guys that does search on yourself, are you? Speaker A [00:27:29] Oh, no, I never do. I never do. When none of us do, nobody does that. Speaker B [00:27:35] I'm going to go back. Uh, yeah. How great is Mark Benioff? And it's going to come back and says, the greatest man ever invented cloud computing. It's going to be great. Speaker A [00:27:52] You know, it's a funny thing, I, um, had a call from someone you know very well on Wednesday, Larry Ellison called me at 730 in the morning to talk to me. And then I was like, larry, you're up early. You know. And then I, um. And then I also heard from another friend of mine who's been at, uh, his job for much longer because it's hard to find. Larry's been doing this since 19, you know, 78. Speaker B [00:28:17] Wow. Speaker A [00:28:18] So if you think about Larry's career. Speaker B [00:28:20] That was before transistors. Close. Speaker A [00:28:25] Think about it, right? Speaker B [00:28:26] Close, yeah. Speaker A [00:28:27] And also. But kind of changed context. Neil Young also, like, is a friend of mine. And whenever I get a note from Neil or I get a call from Larry. And now I think I'm going to put you in this camp. You know, you've been doing this for not one decade, not two decades, you know, three decades. So when you do think about having longevity in your career and you're doing this for more than 30 years, and you have a passion now, you just said to go for 40 years or 50 years because you're so excited about the future, you know, where does that motivation come from? Where does that inspiration? Have you always had it? Are there moments when you have lulls in that level of energy and inspiration, or are you just self directed to the point? No, I'm going forward because we did hit a couple. We hit a few things. You talked about tactics dictate strategy in your life. You are able to take these leaps of faith like you did with deep learning. You look for these opportunities for flywheels of acceleration, like what you're seeing now in terms of where we are with AI. But how do you achieve that kind of velocity over time? Speaker B [00:29:33] Um, well, first of all, not being bored and not being fired are critical conditions for continued employment as CEO. So, um, those are my guardrails. And so, so far, so good. Uh, also, I guess, um. Uh, I can't really tell whether I'm running for food or running away from being food, but I'm running all the time and I don't know where that comes from. Mark. I don't know where that comes from. I think I need to sign up for the same therapy. Uh, maybe I just got to sign up and just self reflect. What are you doing? Is that what therapy does for you? Speaker A [00:30:29] I guess I'm asking what is the most important thing to you? Speaker B [00:30:33] For me, it changes from time to time. Sometimes it's because the technology, um, endeavor or the challenge is just too thrilling. Right now, it's just too thrilling. Don't you think the idea that we would have, we would be, uh, creating all these digital bots, you know, these agents, uh, that could help us do amazing things and that we're about to realize them in physical manifestations. With these, with software embodying these megatronic systems, uh, that could help people do things and help, you know, older people do things and, you know, people worry about, worry about labor, but the fact that it matters, we know there's not enough labor, the labor force is too small and it's causing inflation to happen and so on and so forth. And I, I think clearly the, the technology challenge and the technology realization of it is too thrilling. It's too exciting. And, and it's thrilling and exciting because. Speaker A [00:31:28] That your number one thing that's most important sometimes seems like it's bigger than that for you. Speaker B [00:31:34] Sometimes. I mean, you know, sometimes. And partly the excitement comes from the fact that it's almost, it's within your grasp, you know, sometimes when something. Speaker A [00:31:42] Individual achievement. Speaker B [00:31:44] Yeah. So, you know, we could do this, we can make a real contribution. Uh, sometimes it's because, uh, we realize, or I realize that, you know, our company is in a once in a lifetime position and once in a generation situation to be made, to be able to make a real contribution. We're the only AI, uh, company in the world that works with practically a team success itself. Speaker A [00:32:06] Yeah. Speaker B [00:32:06] And so we're, we're helping every company, we're helping every industry. We're helping life sciences, we're helping transportation and manufacturing. And so the ability to be able to help, uh, work with companies in Japan one day, the UK another day, Canada, you know, and be able to lift their society and lift their infrastructure and lift their capability, that's pretty exciting. And is that exciting? Part responsibility, part excitement, part, um, humility. I mean, it's all that. And then sometimes, and this is probably m more than most, uh, for me, is a sense of responsibility, that we're empowered with all of this capability and ability in our company and that responsibility to all of our employees, all of our partners, uh, the companies that we work with, the industries we serve, you know, that responsibility is great responsibility. And so I don't the responsibility, and. Speaker A [00:33:03] I would call that the Kuleana. So would you say, so it's the responsibility or it's the team success, or it's the levels of individual, individual achievement, or it's the thrill of technology itself. Speaker B [00:33:19] One of those four, it just depends on what. Speaker A [00:33:21] Do they rotate or is there one that is most important to you? Speaker B [00:33:24] You know, right now Mark is like all those four incredible forces. Agent forces are compressed in this one leather jacket, and I'm about to blow too much force. Yeah, I think you feel the same. You feel the same. I do. And most of my peers and colleagues around the industry, this is a once in a generation opportunity, and there are no time in history has technology, um, had the promise of being able to move the needle so much. And I think, um, uh, like I said, we don't want to miss this next decade. Speaker A [00:34:10] And you didn't have that same feeling ten years ago when you were in deep learning or 20 years ago when you were in advanced, you know, computing, you know, when we were talking about supercomputers and we moved into deep learning, and now we're in AI. Your feeling has changed in those three moments. Like now is the. You feel like now is the absolute pinnacle of that? Speaker B [00:34:34] Yeah, I would say. I would say that, um. Uh, you know, David will tell you that I'm, um, super fired up all the time, you know, and, uh, I don't know what that is about, but, you know, I have a deep need to do a good job, and maybe that. I don't know where that came from, but I just have a very deep need. Speaker A [00:34:59] So that's really the number one thing. Speaker B [00:35:00] Yeah. I just need to know that I left it on the field every day, that I couldn't have done anything more. Speaker A [00:35:06] Has that always been true just in your career, in your company? Or was it true before that when I college or in your. You know, I'm gonna pay your whole. Speaker B [00:35:15] I'm gonna give you $20 after our flight. Speaker A [00:35:18] Lie down, too. If you feel more comfortable, that's fine. Speaker B [00:35:24] I've always been like that. Speaker A [00:35:26] Maybe, um. Speaker B [00:35:27] I don't know why. I don't know why. You know, Mark, I'm gonna come to. I'm gonna come to Hawaii. We're gonna sit down and we're going to talk about it. If I don't put on some of my force field right now, you watch, I'm going to be in tears in about five minutes. All of my emotions bubbling up to the surface. They haven't been there for 40 years. Not that I'm 40 years old. I'm just saying. Speaker A [00:36:13] Well, I think we're all very fortunate that you have that revelation at these different points in your life and in your career, and also the passion to stay so fired up over such a long period of time. I use the analogy of Neil Young, but it's not so different than if you go to concert with him. And he's singing a song like old man or Ohio or heart of gold. It's like you feel like you're hearing him for the very first time. And I think maybe you have that same thing inside you that you're feeling that. I think the Japanese who are here, you have the word shoshin, which means the beginner's mind. In the beginner's mind, you have every possibility, uh, but in the expert's mind, you have few. But it seems like you're doing a good job of recultivating your beginner's mind on a regular basis, and it's reigniting your fire, and then you're kind of able to accelerate on each one of these critical points in your life and your career. Is that a fair understanding? Speaker B [00:37:08] Yeah, that's me right there. Exactly. I like the sound of that. Speaker A [00:37:12] So what? Speaker B [00:37:12] I'll pay for that. Speaker A [00:37:13] What is the way? How are you doing that? How are you rekindling your beginner's mind? Because none of us can stay, keep that fired up over a long period of time. Speaker B [00:37:23] Mark, I'm so excited. First of all, this is part the company we created, but I'm surrounded by amazing people. I mean, David Kirk down there, your neighbor, uh, one of the earliest people I knew that, um, uh, amazing, ah, amazing talent, amazing mind. Um, but I'm surrounded by people like that, and I'm always learning from them. Uh, I'm asking good questions. Um, I'm sufficiently humble to realize that I have much to learn to, uh, so I'm inquisitive, I'm curious about things, and when I connect these, the things that I learned, you, uh, know, when you learn something, it gets you fired up. When you see something for the first time, you go, aha. That gets you fired up. When you connect two random ideas that nobody realized could be connected, you get fired up. And, you know, and so, you know, my mind is, uh, just getting fired up by all of this stimulus from all this, you know, these people that I'm surrounded by. I think that's one. And then the other is just that, the endeavor, the endeavor of 30 years ago, wanting to reinvent computing, you know, don't forget, the computer system that we know today was largely described in 1964, the year after my birth, and it's remained basically the same for 60 years. And here comes, here we are. Speaker A [00:38:42] You know, were you born in 1964? Speaker B [00:38:44] I was 63. Speaker A [00:38:45] I'm 64. Speaker B [00:38:46] Okay. Okay. And here he is. Speaker A [00:38:49] So you pass the 63. Speaker B [00:38:50] Uh, here he is. Mark's giving me therapy, and so I think the ambition of wanting to reinvent computing, uh, to create what we now know is such a great endeavor. Um, so somewhere between the being, uh, inspired by all the stimulus, um, the incredible challenge ahead of us, uh, you know, it's kept us propelled, um, inspired and fired up for a long, long time. And now here we are, we now have the instruments, the tools, uh, this capability called artificial intelligence that goes, solve all of those other problems that we've been excited about ever since we were kids. And so, ah, that's pretty exciting times. Speaker A [00:39:35] The summation of that whole thing, right? Because now that I know, like, you were there when the TRS 80 model one was there, and we were at Radio Shack, and we're working on that, uh, you know, basic. And then we're moving into the 6502 computers and the commodores and the Ataris and the apples and all of that. And then all of a sudden we're into these mini computers and mainframes, and now all of a sudden we're in this, like, next generation systems that can do things that we could have never imagined. Speaker B [00:40:02] And that's right. Speaker A [00:40:02] If somebody had asked us when we were in high school to start programming on those things, we'd be like, whoa. Speaker B [00:40:08] How would that be? Unbelievable. Uh, times. And so I think for all of the industry, this is what we're all fired up about because we're now seeing it and we get the benefit of being inside the industry. Looking at the early indications of all this work. And I'm excited to be here today helping, uh, to celebrate the beginning of the enterprise AI era. This is a big deal. Don't forget, you know, we say enterprise is the name of an industry. Um, but the fact of the matter is, this is how society is built, this is how industries are built. And we're going to bring a level of automation capability that the world's never seen. And I love that you codified it as agents, uh, gave it a theme song. Um, it needed a theme song. It needed a theme. Yeah, of course. And so we're all going to be whistling, you know, secret agent man all the way home. And so anyways, I'm excited for you. Yeah, really great times. And you're going to create hundreds of billions of agents. Uh, we'll just be underneath, uh, powering them and, um, giving them energy, you know, making them do their jobs super fast. Speaker A [00:41:18] Well, we're so grateful to you, Jensen, for you, your leadership, your pioneering spirit, your desire and ability to continue to fire yourself up through every generation of computing. The incredible company that you built that's powering all of this. Because without your dedication and your leadership over the last four decades, we would not be where we are. We're so grateful to you. And thank you for everything that you've done for us. Speaker B [00:41:41] Thank you. Thank you.