Jensen Huang on How to Use First-Principles Thinking to Drive Decisions

Full Transcript

Note: Transcripts are generated by machine and lightly edited by humans. They may contain errors.

Jensen Huang: If you send me something and you want my input on it and I can be of service to you and in my review of it, share with you how I reasoned through it, I’ve made a contribution to you. I’ve made it possible to see how I reason through something. And by reasoning, as you know, how someone reasons through something empowers you. You go, “Oh my gosh. That’s how you reason through something like this.” It’s not as complicated as it seems. This is how you reason through something that’s super ambiguous. This is how you reason through something that’s incalculable. This is how you reason through something that seems to be very scary. Do you understand? So, I show people how to reason through things all the time.

Shantam Jain: That was Jensen Huang, the CEO of NVIDIA. Jensen visited Stanford Graduate School of Business, as part of View From The Top, a speaker series where students, like me, sit down to interview leaders from around the world. I’m Shantam Jain, an MBA student of the class of 2024. In our conversation, we discussed the key pillars of Jensen’s leadership philosophy and how he breaks down generative AI using first-principles thinking.

You’re listening to View From The Top, the podcast.

Shantam Jain: Jensen, this is such an honor. Thank you for being here.

Jensen Huang: I’m delighted to be here. Thank you.

Shantam Jain: In honor of your return to Stanford, I decided we’d start talking about the time when you first left. You joined LSI Logic, and that was one of the most exciting companies at the time. You’re building a phenomenal reputation with some of the biggest names in tech, and yet you decided to leave to become a founder. What motivated you?

Jensen Huang: Chris and Curtis. I was an engineer at LSI Logic, and Chris and Curtis were at Sun. And I was working with some of the brightest minds in computer science at the time, of all time, including [unintelligible] and others building workstations and graphics workstations and so on and so forth. And Chris and Curtis one day said that they’d like to leave Sun, and they’d like me to go figure out where they’re going to go leave for.

I had a great job, but they insisted that I figure out with them how to build a company. So, we hung out at Denny’s whenever they dropped by, which was, by the way, my alma mater, my first company. My first job before CEO was a dishwasher, and I did that very well.

[Laughter]

Jensen Huang: So, anyways, we got together, and it was during the microprocessor revolution. This was 1993 and 1992 when we were getting together. The PC revolution was just getting going. You know that Windows ’95, obviously, which is the revolutionary version of Windows, didn’t even come to the market yet, and Pentium wasn’t even announced yet. This was all right before the PC revolution, and it was pretty clear that the microprocessor was going to be very important. And we thought, “Why don’t we build a company to go solve problems that a normal computer that is powered by general purpose computing can’t?” And so that became the company’s mission, to go build a computer, the type of computers that solve problems that normal computers can’t. And to this day, we’re focused on that.

And if you look at all the problems in the markets that we opened up as resolved, it’s things like computational drug design, weather simulation, materials’ design. These are all things that we’re really, really proud of — robotics, self-driving cars, autonomous software we call artificial intelligence. And then, of course, we drove the technology so hard that eventually the computational cost went to approximately zero, and it enabled a whole new way of developing software, where the computer wrote the software itself, artificial intelligence as we know it today. So, that was it; that was the journey.

Shantam Jain: Yeah. Thank you all for coming.

Well, these applications are on all of our minds today. Back then, the CEO of LSI Logic convinced his biggest investor, Don Valentine, to meet with you. He is obviously the founder of Sequoia. Now I can see a lot of founders here edging forward in anticipation. But how did you convince the most sought-after investor in Silicon Valley to invest in a team of first-time founders building a new product for a market that doesn’t even exist?

Jensen Huang: I didn’t know how to write a business plan. So I went to a bookstore, and back then, there were bookstores. And in the business book section, there was this book. And it was written by somebody I knew, Gordon Bell. And this book, I should go find it again, but it’s a very large book, and the book says, “How to Write a Business Plan.” That was a highly specific title for a very niche market. And it seems like he wrote it for 14 people, and I was one of them.

So, I bought the book. I should have known right away that it was a bad idea because Gordon is super smart. And super-smart people have a lot to say. I’m pretty sure Gordon wants to teach me how to write a business plan completely. So, I picked up this book, and it’s like 450 pages long.

Well, I never got through it, not even close. I flipped through it, a few pages. And I go, “You know what? By the time I’m done reading this thing, I’ll be out of business. I’ll be out of money. And Laurie and I only had about six months in the bank. And we had already Spencer, Madison and a dog. So, the five of us had to live off of whatever money we had in the bank, so I didn’t have much time.

So, instead of writing the business plan, I just went to talk to [Wilf Corey]. He called me one day, and he said, “Hey, you left the company. You didn’t even tell me what you were doing. I want you to come back and explain it to me.” And so, I went back and explained it to Wilf. And Wilf at the end of it said, “I have no idea what you said. That’s one of the worst elevator pitches I’ve ever heard.”

[Laughter]

Jensen Huang: And then he picked up the phone, and he called Don Valentine. He called Don, and he said, “Don, I’m going to send a kid over. I want you to give him money.” He’s one of the best employees LSI Logic ever had. And so, the thing I learned is you can make up a great interview. You can even have a bad interview. But you can’t run away from your past, and so have a good past. Try to have a good past.

And in a lot of ways, I was serious when I said I was a good dishwasher. I was probably Denny’s best dishwasher. I planned my work, I was organized, I was mise-en-place, and then I washed the living daylights out of the dishes, and then they promoted me to busboy. I was certain I’m the best busboy Denny’s ever had. I never left a station emptyhanded. I never came back emptyhanded. I was very efficient.

So, anyways, eventually I became a CEO. I’m still working on being a good CEO.

Shantam Jain: Talking about being the best, you needed to be the best among 89 other companies that were funded after you build the same thing. And then with six to nine months of runway left, you realize that the initial vision was just not going to work. How did you decide what to do next to save the company when the cards were so stacked against you?

Jensen Huang: Well, we started this company called [unintelligible] Computing. And the question is, what is it for? What’s the killer app? That became our first great decision. And this is what Sequoia funded. The first great decision was the first killer app was going to be 3D graphics. And the technology was going to be 3D graphics. And the application was going to be videogames. At the time, 3D graphics was impossible to make cheap. It was million-dollar image generators from silicon graphics. And so, it was a million dollars, and it’s hard to make cheap. And the videogame market was [zero billion dollars]. So, you had this incredible technology that’s hard to commoditize and commercialize. And then you have this market that doesn’t exist. That intersection was the founding of our company. And I still remember when Don, at the end of my presentation, one of the things he said to me, which made a lot of sense back then; it makes a lot of sense today, he said, “Startups don’t invest in startups or startups don’t partner with startups.” And his point is that in order for NVIDIA to succeed, we needed another startup to succeed, and that other startup was electronic arts.

And then on the way out, he reminded me that electronic arts is CTO, is 14 years old and had to be driven to work by his mom. He just wanted to remind me that that’s who I’m relying on.

Jensen Huang: And then after that, he said, “If you lose my money, I’ll kill you.” And that was kind of my memories of that first meeting. But nonetheless, we created something. We went on the next several years to go create the gaming market for PCs. It took a long time to do so. We’re still doing it today. We realized that not only do you have to create the technology and invent a new way of doing computer graphics so that what was a million dollars is now 3, 400, 500 dollars that fits in the computer, and you have to go create this new market. So, we had to create technology, create markets.

The idea that company would create technology, create markets defines NVIDIA today. Almost everything we do, we create technology, we create markets. That’s the reason people call it a stack, an ecosystem, words like that, but that’s basically it — a décor for 30 years when NVIDIA realized we had to do is in order to create the conditions by which somebody could buy our products, we had to go invent this new market, and it’s the reason why we’re early in autonomous driving. It was the reason why we were early in deep learning. It’s the reason why we’re early in just about all these things including computational drug design and discovery. All these different areas we’re trying to create the market while we’re creating the technology.

Okay. Then we got going, and then Microsoft introduced a standard called Direct 3D, and that spawned off hundreds of companies. And we found ourselves a couple of years later competing with just about everybody. The thing that we invented the company, the technology we invented 3D graphics with, that consumerized 3D with turns out to be incompatible with Direct 3D.

So, we started this company. We had this 3D graphics thing, a million-dollar thing. We’re trying to make it consumerized, and so we invented all this technology. And then shortly after, it became incompatible, so we had to reset the company or go out of business. But we didn’t know how to build it the way that Microsoft had defined it. I remember a meeting on a weekend, and the conversation was, “We now have 89 competitors. I understand the way we do it is not right, but we don’t know how to do it the right way.”

Thankfully, there was another bookstore, and the bookstore was called [Fry’s Electronics]. I don’t know if it’s still here. I think I drove Madison, my daughter, on the weekend to Fry’s, and it was sitting right there, the open GL manual, which would define how silicon graphics did computer graphics. So, it was right there; it was like $68.00 a book. I had a couple hundred dollars. I bought three books. I took it back to the office, and I said, “Guys, I found it. Our future.” I had the three versions of it. I handed it out. It had a big, nice centerfold. The centerfold is the open GL pipeline, which is the computer graphics pipeline. And I handed it to the same geniuses that I founded the company with. And we implemented the open GL pipeline like nobody had ever implemented the open GL pipeline, and we built something the world had never seen.

So, a lot of lessons are right there. That moment in time for our company gave us so much confidence. And the reason for that is you can succeed in doing something, inventing a future, even if you were not informed about it at all. And that’s kind of my attitude about everything now. When somebody tells me about something and I’ve never heard of it before, or if I’ve heard of it and don’t understand how it works at all, my first thought is always, “How hard can it be? And it’s probably just a textbook away. You’re probably one archive paper away from figuring this out.”

So, I spent a lot of time reading archive papers. And it’s true. Now, of course, you can’t learn how somebody else does something and do it exactly the same way and hope to have a different outcome. But you can learn how something can be done and then go back to first principles and ask yourself, “Given the conditions today, given my motivation, given the instruments, the tools, given how things have changed, how would I redo this? How would I reinvent this whole thing? How would I design it? How would I build a car today? Would I build it incrementally from 1950’s and 1900’s? How would I build a computer today? How would I write software today” Does that make sense?

So, I go back to first principles all the time, even in the company today, and just reset ourselves, because the world has changed. The way we wrote software in the past, it was monolithic, and it’s designed for supercomputers, but now it’s this aggregated so on and so forth. How we think about software today, how we think about computers today, just always cause your company, always cause yourself to go back to first principles, and it creates lots and lots of opportunities.

Shantam Jain: The way you apply this technology tends to be revolutionary. You get all the momentum that you need to IPO and then some more, because you grow your revenue nine times in the next four years. But in the middle of all of this success, you decide to [pip] it a little bit, the focus of innovation happening at NVIDIA based on a phone call you have with this chemistry professor. Can you tell us about that phone call and how you connected the dots from what you heard to where you went?

Jensen Huang: I remember at the core, the company was pioneering a new way of doing computing. Computer graphics was the first application. But we always knew that there would be other applications, so image processing came, particle physics came, fluids came, so on and so forth, all kinds of interesting things that we wanted to do.

We made the processor more programmable so that we could express more algorithms, if you will. And then one day, we invented programmable shaders, which made all forms of imaging and computer graphics programmable. That was a great breakthrough, so we invented that.

On top of that, we tried to look for ways to express more sophisticated algorithms that could be computed on our processor, which is very different than a CPU. So, we created this thing called a CG. I think it was 2003 or so. C for GPUs. It predated [CUDA] by about three years.

The same person who wrote the textbook that saved the company, Mark [Kilgard], wrote that textbook. And so, CG was super cool. We wrote textbooks about it. We started teaching people how to use it. We developed tools and such. And then several researchers discovered it. Many of the researchers here, students here at Stanford were using it. Many of the engineers that then became engineers at NVIDIA were playing with it.

A couple of doctors at Mass General picked it up and used it for CT reconstruction. So, I flew out and saw them and said, “What are you guys doing with this thing?” And they told me about that. Then a computational, quantum chemist used it to express his algorithms.

So, I realized that there’s some evidence that people might want to use this. And it gave us incrementally more confidence that we ought to go do this, that this form of computing could solve problems that normal computers really can’t and reinforced our belief and kept us going.

Shantam Jain: Every time you heard something new, you really savored that surprise, and that seems to be a theme throughout your leadership at NVIDIA. It feels like you make these bets so far in advance of technology inflections that when the apple finally falls from the tree, you’re standing right there in your black leather jacket waiting to catch it.

[Laughter]

Shantam Jain: How do you find the [conviction]?

Jensen Huang: It always seems like a diving catch. You do things based on core beliefs. We deeply believe that we could create a computer that solves problems that normal processing can’t do. There are limits to what a CPU can do. There are limits to what general purpose computing can do. And then there are interesting problems that we can go solve. The question is always — are those interesting problems only or can they also be interesting markets? Because if they’re not interesting markets, it’s not sustainable. And NVIDIA went through about a decade where we were investing in this future and the markets didn’t exist. There was only one market at the time; it was computer graphics.

For 10, 15 years, the markets that fuel NVIDIA today just didn’t exist. So how do you continue with all of the people around you, our company and NVIDIA’s management team and all of the amazing engineers that were there creating this future with me — all of your shareholders, your board of directors, your partners, you’re taking everybody with you, and there’s no evidence of a market. That is really, really challenging. The fact that the technology can solve problems, and the fact that you have research papers that are used, that are made possible because of it are interesting. But you’re always looking for that market. But nonetheless, before a market exists, you still need early indicators of future success.

We have this phrase in the company. There’s a phrase called “key performance indicators.” Unfortunately, KPIs are hard to understand. I find KPIs hard to understand. What’s a good KPI? A lot of people, when we look for KPIs, they go, “Gross margins.” That’s not a KPI; that’s a result. You’re looking for something that’s early indicators of future positive results and as early as possible. The reason for that is because you want that early sign that you’re going in the right direction.

So, we have this phrase that’s called, “EOIFS,” early indicators to EOIFS, early indicators of future success. And it helps people, because I was using it all the time, to give the company hope that, “Hey, look, we solved this problem, we solved that problem, we solved this problem.” The markets didn’t exist, but there were important problems, and that’s what the company’s about, to solve these problems. We want to be sustainable, and therefore, the markets have to exist at some point.

But you want to decouple the result from evidence that you’re doing the right thing, okay? So that’s how you kind of solve this problem of investing into something that’s very, very far away and having the conviction to stay on the road is to find as early as possible the indicators that you’re doing the right things. So, start with a core belief. Unless something changes your mind, you continue to believe in it. And look for early indicators of future success.

Shantam Jain: What are some of those early indicators that have been used by product teams at NVIDIA?

Jensen Huang: All kinds. I saw a paper. Long before I saw the paper, I met some people that needed my help on this thing called deep learning. At the time, I didn’t know what deep learning was. And they needed us to create a domain-specific language so that all of their algorithms could be expressed easily on our processors. And we created this thing called [Ku-DNN]. And it’s essentially the [SQL]. SQL is in-storage computing. This is neural-network computing, and we created a language, if you will, domain-specific language from that, kind of like the open GL of deep learning.

So, they needed us to do that so that they could express their mathematics. And they didn’t understand KUDO, but they understood the deep learning. So, we created this thing in the middle for them. And the reason why we did it was because these researchers had no money. This is kind of one of the great skills of our company, that you’re willing to do something even though the financial returns are completely non-existent or maybe very, very far out, even if you believed in it.

We ask ourselves, “Is this worthy work to do? Does this advance a field of science somewhere that matters?” Notice, this is something that I’ve been talking about since the very beginning of time. We find inspiration, not from the size of a market, but from the importance of the work, because the importance of the work is the early indicators of a future market. Nobody has to do a business case on it. Nobody has to show me a [PNL]. Nobody has to show me a financial forecast. Th only question is, “Is this important work?” And if we didn’t do it, would it happen without us?” Now if we didn’t do something and something could happen without us, it gives me tremendous joy, actually.

The reason for that is — could you imagine — the world got better, you didn’t have to lift a finger? That’s the definition of ultimate laziness. And in a lot of ways, you want that habit. And the reason for that is this — you want the company to be lazy about doing things that other people always do, can do. If somebody else can do it, let them do it. We should go select the things that if we didn’t do it, the world would fall apart. You have to convince yourself of that, “If I don’t do this, it won’t get done.” If that work is hard, and that work is impactful and important then it gives you a sense of purpose. Does that make sense? And so, our company has been selecting these projects. Deep learning was just one of them. And the first indicator of the success of that was this fuzzy cat that Andrew [Ang] came up with, and then Alex [Korchevsky] detected cats, not all the time, but successfully enough that it was, “This might take us somewhere.” And then we reasoned about the structure of deep learning, and we’re computer scientists, and we understand how things work. So, we convinced ourselves this could change everything. Anyhow, that’s an example.

Shantam Jain: So these selections that you’ve made, they’ve paid huge dividends both literally and figuratively. But you’ve had to steer the company through some very challenging times, like when it lost 80 percent of its market cap amid the financial crisis because Wall Street didn’t believe in your bet on ML. In times like these, how do you steer the company and keep the employees motivated at the task at hand?

Jensen Huang: My reaction during that time is the same reaction I had about this week. Earlier today, you asked me about this week. My pulse was exactly the same. This week is no different than last week or the week before that. So, the opposite of that, when you drop 80 percent, don’t get me wrong, when your share price drops 80 percent, it’s a little embarrassing, okay? You just want to wear a T-shirt that says, “It wasn’t my fault.”

But even more than that, you don’t want to get out of your bed, you don’t want to leave the house. All of that is true. But then you go back to just doing your job. I woke up at the same time. I prioritized my day in the same way. I go back to, “What do I believe?” You’ve got to gut check; always gut check back to the core — what do you believe? What are the most important things? Just check them off. Sometimes it’s helpful — family loves me? Okay, check, double check, right? So, you’ve just got to check it off, then you go back to your core and then go back to work. And then every conversation goes back to the core, keep the company focused back on the core. Do you believe in it? Did something change? The stock price changed, but did something else change? Did physics change? Did gravity change? Did all of the things that we assumed that we believed that led to our decision, did any of those things change? Because if those things changed, you’ve got to change everything. But if none of those things changed, you change nothing, keep on going. That’s how you do it.

Shantam Jain: In speaking with your employees, they say that —

Jensen Huang: And try to avoid the public.

Shantam Jain: [Laughs] In speaking with your employees, they’ve said that your leadership is —

Jensen Huang: Including the employees. I’m just kidding.

[Laughter]

Jensen Huang: Leaders have to be seen, unfortunately. That’s the hard part. I was an electrical engineering student, and I was quite young when I went to school. When I went to college, I was still 16 years old, so I was young when I did everything. So I was a bit of an introvert. I’m shy. I don’t enjoy public speaking. I’m delighted to be here. I’m not suggesting that. But it’s not something that I do naturally. So, when things are challenging, it’s not easy to be in front of precisely the people that you care most about. And the reason for that is because could you imagine a company meeting with just our stock prices dropped by 80 percent? And the most important thing I have to do as the CEO is this, to come and face you, explain it. Partly, you’re not sure why. Partly, you’re not sure how long, how bad. You just don’t know these things. But you’ve still got to explain it, face all these people and you know what they’re thinking. Some of them were probably thinking, “We’re doomed.” Some people are probably thinking, “You’re an idiot.” And some people are probably thinking something else. So, there are a lot of things that people are thinking, and you know that they’re thinking those things, but you still have to get in front of them and do the hard work.

Shantam Jain: Maybe you can give those things, but yet not a single person of your leadership team left during times like this.

Jensen Huang: Unemployable.

That’s what I keep reminding them.

I’m just kidding. I’m surrounded by geniuses, utter geniuses, unbelievable. NVIDIA is well-known to have singularly the best management team on the planet. This is the deepest technology management team the world’s ever seen. I’m surrounded by a whole bunch of them, and they’re just geniuses — business teams, marketing teams, sales teams, and it’s just incredible — engineering teams, research teams, unbelievable.

Shantam Jain: Your employees say that your leadership style is very engaged. You have 50 direct reports. You encourage people across all parts of the organization to send you the top five things on their mind. And you constantly remind people that, “No task is beneath you.” Can you tell us why you’ve purposefully designed such a flat organization? And how should we be thinking about our organizations that we design in the future?

Jensen Huang: To me, no task is beneath me because, remember, I used to be a dishwasher, and I mean that. I used to clean toilets. I’ve cleaned a lot of toilets. I’ve cleaned more toilets than all of you combined, and some of them you just can’t unsee.

I don’t know what to tell you. That’s life. So, you can’t show me a task that’s beneath me. I’m not doing it only because of whether it’s beneath me or not beneath me. If you send me something and you want my input on it and I can be of service to you and in my review of it, share with you how I reasoned through it, I’ve made a contribution to you. I’ve made I possible to see how I reason through something. And by reasoning, as you know, how someone reasons through something empowers you. You go, “Oh my gosh. That’s how you reason through something like this.” It’s not as complicated as it seems. This is how you reason through something that’s super ambiguous. This is how you reason through something that’s incalculable. This is how you reason through something that seems to be very scary. Do you understand?

So, I show people how to reason through things all the time — strategy things, how to forecast something, how to break a problem down, and you’re just empowering people all over the place. And so that’s how I see it. If you send me something and you want me to help review it, I’ll do my best, and I’ll show you how I would do it.

In the process of doing that, of course, I learned a lot from you. Is that right? You gave me a seed of a lot of information. I learned a lot, and so I feel rewarded by the process.

It does take a lot of energy sometimes because in order to add value to somebody and they’re incredibly smart as a starting point and I’m surrounded by incredibly smart people, you have to at least get to their plane, you know? You have to get into their headspace. And that’s really hard, and that takes just an enormous amount of emotional and intellectual energy, and so I feel exhausted after I work on things like that.

I’m surrounded by a lot of great people. CEOs should have the most of the reports by definition because the people that reports to the CEO requires the least amount of management. It makes no sense to me that CEOs have so few people reporting to them except for one fact that I know to be true. The knowledge, the information of a CEO is supposedly so valuable, so secretive, you can only share it with two other people or three.

And their information is so invaluable, so incredibly secretive that they can only share it with a couple more. Well, I don’t believe in a culture, in an environment where the information that you possess is the reason why you have power. I would like us all to contribute to the company. And our position in the company should have something to do with our ability to reason through complicated things, lead other people to achieve greatness, inspire, empower other people, support other people. Those are the reasons why the management team exists, in service of all of the other people that work at the company, to create the conditions by which all of these amazing people volunteer to come work for you instead of all of the other amazing, high-tech companies around the world. They elected, they volunteered to work for you. And so you should create the conditions by which they can do their life’s work, which is my mission.

You probably heard it. I’ve said it pretty clearly, and I believe that. What my job is is very simply to create the conditions by which you can do your life’s work. So, how do I do that? What does that condition look like? Well, that condition should result in a great deal of empowerment. You can only be empowered if you understand the circumstance; isn’t that right? You have to understand the context of the situation you’re in in order for you to come up with great ideas. And so, I have to create a circumstance where you understand the context, which means you have to be informed. And the best way to be informed is for there to be as little layers of information mutilation, right, between us. And so that’s the reason why it’s very often that I’m reasoning through things like in an audience like this. I say, first of all, these are the beginning facts. These are the data that we have. This is how we reason through it. These are some of the assumptions. These are some of the unknowns. These are some of the knowns. So, you reason though it.

Now you’ve created an organization that’s highly empowered. NVIDIA is 30,000 people. We’re the smallest large company in the world. We’re a tiny little company. But every employee is so empowered, and they’re making smart decisions on my behalf every single day. And the reason for that is because they understand my condition. I’m very transparent with people. And I believe that I can trust you with the information.

Oftentimes, the information is hard to hear, and the situations are complicated, but I trust that you can handle it. A lot of people hear me say, “You’re adults here. You can handle this.” Sometimes they’re not really adults, and they just graduated. I’m just kidding. I know that when I first graduated, I was barely an adult. But I was fortunate that I was trusted with important information. So, I want to do that. I want to create the conditions for people to do that.

Shantam Jain: I do want to now address the topic that is on everybody’s mind, AI. Last week, you said that generative AI and accelerated computing have hit the tipping point. So as this technology becomes more mainstream, what are the applications that you personally are most excited about.

Jensen Huang: Well, you have to go back to first principles and ask yourself, “What is generative AI? What happened?” What happened was we now have the ability to have software that can understand something. First of all, we digitized everything. Like, for example, gene sequencing — digitized genes. But what does it mean? That sequence of genes, what does it mean? We’ve digitized amino acids, but what does it mean?

So, we now have the ability to digitize words. We digitize sounds. We digitize images and videos. We digitize a lot of things. But what does it mean? We now have the ability through a lot of study and a lot of data and from patterns in relationships, we now understand what they mean. Not only do we understand what they mean, we can translate between them because we learned about the meaning of these things in the same world; we didn’t learn about them separately. So, we learned about speech and words and paragraphs and vocabulary in the same context. So, we’ve found correlations between them, and they’re all registered, if you will, registered to each other.

And so now not, only do we understand the meaning of each modality, we can understand how to translate between them. And so for obvious things, you could caption video to text; that’s captioning, text two images [mid journey], text-to-text, Chat GPT, amazing things. And so we now know that we understand meaning, and we can translate. The translation of something is generation of information. And all of a sudden, you have to take a step back and ask yourself, “What is the implication in every single layer of everything that we do?” So, I’m exercising in front of you, I’m reasoning in front of you, the same thing I did 15 years ago when I first saw Alex some 13, 14 years ago.

How I reasoned through it, what did I see? How interesting. What can it do? Very cool. But then, most importantly, what does it mean? What does it mean to every single layer of computing because we’re in a world of computing. So, what it means is that the way that we process information fundamentally will be different in the future. That’s when NVIDIA builds chips and systems. The way we write software will be fundamentally different in the future. The type of software we’ll be able to write in the future will be different, new applications. And then also the processing of those applications will be different. What was historically a retrieval-based model where information was prerecorded, if you will, almost. We wrote the text, prerecorded, and we retrieved it based on some recommender system algorithm. In the future, some seed of information will be the starting point. We call them prompts, as you guys know, and then we generate the rest of it. And so, the future of computing will be highly generated.

Well let me give you an example of what’s happening. For example, we’re having a conversation right now. Very little of the information I’m conveying to you is retrieved. Most of it is generated. It’s called intelligence. So, in the future, we’re going to have a lot more generative — our computers will perform in that way. It’s going to be highly generative instead of highly retrieval-based.

Then you go back and you’re going to ask yourself — now for entrepreneurs you’re going to ask yourself what industries will be disrupted? Therefore, will we think about networking the same way? Will we think about storage the same way? Will we be as abusive of Internet traffic as we are today? Probably not. Notice we’re having a conversation right now, and I don’t have to get in my car every question. So, we don’t have to be as abusive of transformation/information/transporting as we used to.

What’s going to be more? What’s going to be less? What kind of applications, etcetera, etcetera? So, you can go through the entire industrial spread and ask yourself what’s going to get disrupted, what’s going to be different, what’s going to get [new], so on and so forth.

And that reasoning starts from what is happening? What is generative AI? Foundationally, what is happening? Go back to first principles with all things. There was something I was going to tell you about organization. You asked the question, and I forgot to answer it. The way you create an organization by the way someday, don’t worry about how other companies’ org charts look. You start from first principles. Remember what an organization is designed to do.

The organizations of the past, there’s a king/CEO, and then you have all the royal subjects, the royal court and then east out. And then you keep working your way down. Eventually, they’re employees. The reason why it was designed that way is because they wanted the employees to have as little information as possible because their fundamental purpose of the soldiers is to die in the field of battle, to die without asking questions. You guys know this.

I only have 30,000 employees. I would like none of them to die. I would like them to question everything. Does that make sense? And so the way you organized in the past and the way you organize today is very different.

Second, the question is what does NVIDIA build? An organization is designed so that we can build whatever it is we build better. And so if we all build different things, why are we organized the same way? Why would this organizational machinery be exactly the same irrespective of what you built? It doesn’t make any sense. You build computers, you organize this way. You build healthcare services, you build exactly the same way. It makes no sense whatsoever. So you had to go back to first principles, just ask yourself, “What kind of machinery? What is the input? What is the output? What are the properties of this environment? What is the forest that this animal has to live in? What are the characteristics? Is it stable most of the time? Are you trying to squeeze out the last drop of water or is it changing all the time, being attacked by everybody?”

So you’ve got to understand, you’re the CEO. Your job is to architect this company. That’s my first job, to create the conditions by which you can do your life’s work, and the architecture has to be right, and so you have to go back to first principles and think about those things.

I was fortunate that when I was 29 years old, I had the benefit of taking a step back and asking myself, “How would I build this company for the future and what would it look like? What’s the operating system, which is called culture? What kind of behavior do we encourage, enhance, and what do we discourage and not enhance and so on and so forth? Anyways.

Shantam Jain: I want to save time for audience questions. But this year’s theme from you from the top is “Redefining Tomorrow.” And one question we’ve asked all of our guests is, Jensen, as the cofounder and CEO of NVIDIA, if you were to close your eyes and magically change one thing about tomorrow, what would it be?

Jensen Huang: Were we supposed to think about this in advance?

[Laughter]

Jensen Huang: I’m going to give you a horrible answer. I don’t know that it’s one thing. Look, there are a lot of things that we don’t control. There are a lot of things we don’t control. Your job is to make a unique contribution. Live a life of purpose, to do something that nobody else in the world would do or can do, to make a unique contribution so that in the event that after you are done, everybody says the world was better because you were here. So, I think, to me, I live my life kind of like this. I go forward in time, and I look backwards. So, you asked me a question that’s exactly from a computer vision pose perspective, exactly the opposite of how I think. I never look forward from where I am. I go forward in time and look backwards. And the reason for that is it’s easier. I would look backwards and kind of read my history. We did this and we did it that way and we [unintelligible] that problem down. Does that make sense?

So, it’s a little bit like how you guys solve problems. You figured out what is the end result that you’re looking for and you work backwards to achieve it. I imagine NVIDIA making a unique contribution to advancing the future of computing, which is the single most important instrument of all humanity. Now it’s not about our self-importance, but this is just what we’re good at, and it’s incredibly hard to do. And we believe we can make an absolute unique contribution. It’s taken us 31 years to be here, and we’re still just beginning our journey, and so this is insanely hard to do.

And when I look backwards, I believe that we’re going to be remembered as a company that kind of changed everything, not because we went out and changed everything through all the things that we said, but because we did this one thing that was insanely hard to do that we’re incredibly good at doing that we love doing, we did for a long time.

Female Voice: I’m part of the GSB lead. I graduated in 2023. So, my question is, how do you see your company in the next decade as what challenges do you see your company would face and how you are positioned for that?

Jensen Huang: First of all, can I just tell you what’s going on through my head? As you say what challenges, the list that flew by my head was so large that I was trying to figure out what to select. Now the honest truth is that when you asked that question, most of the challenges that showed up for me were technical challenges. And the reason for that is because that was my morning. If you had chosen yesterday, it might have been market creation challenges. There were some markets that, gosh, I just desperately would love to create. Can’t we just do it already? But we can’t do it alone. NVIDIA is a technology platform company. We’re here in service of a whole bunch of other companies so that they could realize, if you will, our hopes and dreams through them.

So, some of the things I would love, I would love for the world of biology to be at a point where it’s kind of like the world of chip design 40 years ago, computer-aided and designed, EDA that entire industry really made possible for us today. And I believe we’re going to make possible for them tomorrow. Computer-aided drug design — because we’re able to now represent genes and proteins and even cells now, very, very close to be able to represent and understand the meaning of a cell, combination of a whole bunch of genes. What does a cell mean? It’s kind of like, what does a paragraph mean? If we could understand a cell like we understand a paragraph, imagine what we could do.

So, I’m anxious for that to happen. I’m kind of excited about that. There are some that I’m just excited about that I know we’re around the corner on, for example, humanoid robotics. They’re very, very close around the corner. And the reason for that is because if you can tokenize and understand speech, why can’t you tokenize and understand manipulation? So these kind of computer science techniques, once you figure something out, you ask yourself, “Well, if I do that, why can’t I do that?” So I’m excited about those kinds of things. So that challenge is kind of a happy challenge.

Some of the other challenges of course are industrial and geopolitical and they’re social, but you’ve heard all that stuff before. These are all true, you know? The social issues in the world, the geopolitical issues in the world, why can’t we just get along, things in the world, why do I have to say those kinds of things in the world? Why do we have to say those things and then amplify them in the world? Why do we have to judge people so much in the world? All those things, you guys all know that. I don’t have to say those things over again.

Jose: My name’s Jose. I’m a Class of 2023 from GSB. My question is, are you worried at all about the pace at which we’re developing AI, and do you believe that any sort of regulation might be needed? Thank you.

Jensen Huang: The answer is yes and no. You know the greatest breakthrough in modern AI, of course, deep learning, it enabled great progress. But another incredible breakthrough is something humans know and we practiced all the time, and we just invented it for language models called grounding — reinforcement learning to human feedback. I provide reinforcement learning human feedback every day. That’s my job. And for the parents in the room, you’re providing reinforcement learning human feedback all the time, okay? Now we just figured out how to do that at a systematic level for artificial intelligence.

There are a whole bunch of other technologies necessary to guardrail, finetune, ground, for example, how do I generate tokens that obey the laws of physics? Right now, things are floating in space and doing things, and they don’t obey the laws of physics. That requires technology. Guard-railing requires technology. Finetuning requires technology. Alignment requires technology. Safety requires technology. The reason why planes are so safe is because all of the autopilot systems are surrounded by diversity and redundancy and all kinds of different functional safety and active safety systems that were invented.

I need all of that to be invented much, much faster. You also know that the border between cybersecurity and artificial intelligence is going to become blurrier and blurrier, and we need technology to advance very, very quickly in the area of cybersecurity in order to protect us from artificial intelligence. So, in a lot of ways, we need technology to go faster, a lot faster.

Regulation — there are two types of regulation. There’s social regulation; I don’t know what to do about that. But there’s product and services regulation; I know exactly what to do about that. So the FAA, the FDA, [NTSA], you name it, all the F’s and all the N’s and the FCCs, they all have regulations for products and services that have particular use cases, bar exams and doctors and so on and so forth. You all have qualification exams. You all have standards that you have to reach. You all have to continuously be certified, accountants and so on and so forth. Whether it’s a product or a service, there are lots and lots of regulations. Please do not add a super regulation that cuts across. The regulator who’s regulating accounting should not be the regulator that regulates a doctor.

I love accountants, but if I ever need open heart surgery, the fact that they can close books is interesting, but not sufficient. So I would like all of those fields that already have products and services to also enhance their regulations in the context of AI. But I left out this one very big one, which is the social implication of AI, and how do you deal with that? I don’t have great answers for that. But enough people are talking about it.

It’s important to subdivide all of this into chunks; does that make sense, so that we don’t become super-hyper-focused on this one thing at the expense of a whole bunch of routine things that we could have done, and as a result, people are getting killed by cars and planes. It doesn’t make any sense. We should make sure that we do the right things there, very practical things. May I take one more question?

Shantam Jain: Well, we have a set of rapid-fire questions for you as [unintelligible] [clinician].

Jensen Huang: Okay. I was trying to avoid that.

[Laughter]

Jensen Huang: All right. Fire away.

Shantam Jain: Well, your first job was at Denny’s. They now have a booth dedicated to you. What was your fondest memory of working there?

Jensen Huang: My second job was AMD by the way. Is there a booth dedicated to me there? I’m just kidding.

I loved my job there; I did. I loved it. It was a great company.

Shantam Jain: If there was a worldwide shortage of black leather jackets, what would we see you wearing?

Jensen Huang: No, I’ve got a large reservoir of black jackets.

I’ll be the only person who is not concerned.

Shantam Jain: You spoke a lot about textbooks. If you had to write one, what would it be called?

Jensen Huang: I wouldn’t write one.

You’re asking me a hypothetical question that has no possibility of …

Shantam Jain: That’s fair. Finally, if you could share one parting piece of advice to broadcast across Stanford, what would it be?

Jensen Huang: It’s not a word, but have a core belief. Gut check it every day. Pursue it with all your might. Pursue it for a very long time. Surround yourself with people that you love, and take ‘em on that ride. So, that’s the story of NVIDIA.

Shantam Jain: Jensen, this last hour has been a treat. Thank you for spending it with us.

Jensen Huang: Thank you very much.

Shantam Jain: You’ve been listening to View From The Top, the podcast, a production of Stanford Graduate School of Business. This interview was conducted by me, Shantam Jain, of the MBA Class of 2024. Lily Sloane composed our theme music. Michael Reilly and Jenny Luna produced this episode. Find this series on our YouTube channel or on our website at gsb.stanford.edu. Follow us on social media @stanfordgsb.