Nvidia’s Jensen Huang on the AI revolution, job losses and what drives him — Full interview

[00:00:00] I hope to die on the job. The narrative that connects AI to job loss, for many of the CEOs [00:00:07] that are doing it, it is just too lazy. Well, China is going to be everybody's greatest rival. [00:00:13] Jensen, thank you so much for joining us. Victoria, it's great to be with you. [00:00:18] Yeah. You know, in today's world, it's hard to imagine AI without you and NVIDIA. [00:00:23] So, can you share with us your vision of this AI revolution and where will we be five years from now? [00:00:31] AI is a revolutionary technology, no question. Just like the invention of information technology with IBM [00:00:42] and the personal computer with Microsoft and Intel, and of course, Taiwan was central to that, [00:00:49] the internet, mobile cloud, and now here we are with artificial intelligence. [00:00:54] The big breakthrough of artificial intelligence is on multiple levels. On the one hand, it's a brand new [00:01:02] technology that is able to enable a whole new class of applications that were impossible before [00:01:08] because you can now understand information, you can reason, you can plan to come up with an action plan, [00:01:16] and even now with agentic AI, you can take action, use tools. [00:01:21] And so, the applications of artificial intelligence, just like humans, are quite vast [00:01:26] and almost boundless in potential. [00:01:30] But artificial intelligence is more important than that even. [00:01:33] You know, unlike many new technologies, artificial intelligence opens up a whole new industry. [00:01:40] And a good framework to think about for artificial intelligence is that AI is not just a model, AI is really a five-layer cake. [00:01:50] To think about it from the lowest level, it requires energy because AI is produced, generated in real time. [00:01:59] Just like you and I right now, we need calories, we need energy so that we can engage each other and produce, generate intelligence, generate answers based on the question you just asked me. [00:02:11] And I can generate it for the very first time in this way based on the fact that I'm with the audience of CNA and I'm here in Taipei. [00:02:20] So the context is different, the circumstances are different, so the answer is different. [00:02:26] And so, we need energy at the lowest level for this industry. [00:02:30] The next level is chips. [00:02:31] This is where NVIDIA is and where a lot of the companies in Taiwan are in. [00:02:36] And the layer above that is infrastructure, land, power, data center, as well as the cloud service software that turns the chips into a cloud service, into a data center. [00:02:52] The layer above that is where most people think about when they think about AI, which is models. [00:02:57] ChatGPT was revolutionary. [00:03:00] Cloud Code is now incredibly successful. [00:03:03] But AI is much more than just human language. [00:03:07] AI can process information of almost any kind from English and Chinese all the way to videos and images, but very importantly, proteins and chemicals and three-dimensional geometry, which is really important for manufacturing, which is important here in Taiwan. [00:03:27] And so, that's the model layer. [00:03:30] Then above that, ultimately, this is the most important, which is how does all of this AI technology impact society in a positive way? [00:03:40] And so, that would be applications for information workers like ourselves. [00:03:45] And I use AI every day. [00:03:47] You probably use AI every day. [00:03:49] But it's much more than that. [00:03:50] It's, you know, software engineers and chip designers and manufacturing managers and plant managers, all the way to, of course, self-driving cars and robotics and healthcare. [00:04:01] Just about every large industry in the world is going to be impacted by this. [00:04:05] And so, the thing that's really helpful for me is to take a step back and look at AI, not in the context of just the model, but look at it in the context of the whole industry. [00:04:19] And when you do that, you come to realize that AI is reinventing every industry, from energy all the way to all of the applications above. [00:04:31] And you've talked about agentic AI. [00:04:34] Now, in simple terms, what is it and how will it change the game? [00:04:39] Agentic AI is called agentic AI for its agency. [00:04:44] Agency simply means it has the ability to do things autonomously with very little supervision or a lot of supervision, just like people. [00:04:55] And so, agentic pipeline has several stages. [00:05:00] One, it starts with what is the context, what's the environment, what's the circumstance, and what is being asked of? [00:05:08] What are you asking this AI to do? [00:05:11] And so, you start with you are an excellent news reporter, you know, or you're an excellent software engineer. [00:05:20] And I would like you to access these files, use this as an example, and I would like you to write the software, for example, or write the story with me. [00:05:31] And so, that's the context and the request. [00:05:34] From that, it has to understand the context. [00:05:37] It has to plan its work. [00:05:40] It has to use whatever tools that it uses. [00:05:43] It could be a browser. [00:05:44] It could be a Word document, a Word editor. [00:05:47] It could be a C compiler. [00:05:49] It could be a Python, you know, editor. [00:05:52] It could be a CUDA program, for example. [00:05:55] And so, it could figure out a way. [00:05:58] You could use these different tools, and then it would perform, and then it would evaluate, re-evaluate, re-plan, exercise, you know, act again, evaluate, and just keep iterating until it gets the job done. [00:06:10] Just like, just like, just like we do. [00:06:13] And so, it performs the task autonomously, which is the reason why we say it's agents. [00:06:19] Agents has two forms. [00:06:21] There's the, there's the, the digital version that we call AI agents. [00:06:26] In the future, these AI agents would also run inside a physical body. [00:06:32] For example, a robot. [00:06:33] So, a robot is essentially a physical AI agent. [00:06:38] And so, these two, these two examples of autonomous systems will likely be highly generalized and very exciting. [00:06:47] Could that make human lazier thinkers? [00:06:51] Because if everyone has a powerful assistant, then what happens to mastery? [00:06:57] Well, I think you could use history as a guide. [00:07:03] When the personal computer came along, I think that everybody said, is that going to make people lazier as a result? [00:07:11] Because so much of work is going to be, so many tasks will be automated. [00:07:15] And then when the internet came along, are, are people going to stop being smart because all the information's at their fingertips? [00:07:23] And, and then before you know it, mobile cloud came along, so it's actually in our pocket. [00:07:27] So, we don't have to remember anything anymore. [00:07:30] And, um, uh, we have so many automation capabilities to, the performance of a computer increased by a million times during, during this time. [00:07:39] Do we find ourselves busier or less busy? [00:07:43] And I think the answer is, we found ourselves busier. [00:07:46] And the reason for that is because we became more ambitious. [00:07:51] The type of things that we could do is greater than ever. [00:07:54] Our students today that are graduating, they're a hundred times smarter than when I graduated. [00:08:00] What they know, what they can do, their exposure to the world, the knowledge they have, the wisdom that they somehow imbued, um, all of that, all of that capability was made possible because of information technology, but it hasn't made them, you know, any less busy. [00:08:17] I think you and I would admit we're busier than ever. [00:08:20] And so I think the same thing is going to happen to artificial intelligence. [00:08:24] We're, we're in our jobs, in our lives. [00:08:27] We have many tasks. [00:08:29] And it's really a basket of tasks. [00:08:31] You know, a job is like a basket of tasks. [00:08:34] Many small components of things that in totality brings meaning to our work. [00:08:40] It allows us to achieve our purpose of our, of our, of our work. [00:08:43] So for example, a radiologist looks at the images, the scans of maybe ultrasounds or CT or MRIs, but that's the task that they do. [00:08:55] And now it's completely AI automated. [00:08:58] but their purpose is really to diagnose disease, help patients get well, help them understand their disease better. [00:09:07] So the purpose of a job is a collection of, of tasks, a basket of tasks. [00:09:12] Many of those tasks will be automated. [00:09:15] And my sense is that as a result, um, of automation, we can focus on the harder parts of our work and the harder parts of our work elevate is elevated. [00:09:26] But as a result, we can be more ambitious. [00:09:29] My guess is that we'll probably be busier. [00:09:31] The opposite. [00:09:32] I just said the opposite of what most people say. [00:09:35] And, um, I have lived, you know, I'm living proof of it. [00:09:38] You are living proof of it. [00:09:40] And speaking of jobs, everyone is curious. [00:09:43] Uh, we are seeing more and more companies cutting jobs while they're investing heavily in AI. [00:09:49] So on a scale of one to 10, how inevitable are job cuts because of AI and what do you say to people who are afraid of losing their jobs to AI? [00:09:59] I would say to the people who are worried about losing their jobs to AI to learn AI you're not going to lose your jobs to AI. [00:10:12] But when the PC came along, the PC didn't take people's jobs, the people who didn't learn how to use PCs were left behind. [00:10:23] You know, when the computers came along, the computer industry, I had the benefit of being swept along with this incredible growth of this industry. [00:10:32] I could have, of course, rejected it and went to another field where the computers was not involved and I could have been left behind. [00:10:40] And so my recommendation is don't allow yourself to be left behind by AI. [00:10:45] Now, let's talk about, let's talk about what's going on. [00:10:49] It is very, it's more likely that the companies with ambition will be more productive. [00:10:58] They will do things faster. [00:11:01] Their company would increase in velocity. [00:11:03] As a result, they become larger, more profitable. [00:11:07] When they become larger, more profitable, they'll end up hiring more people. [00:11:11] And, of course, they'll use more AI, but they'll also hire more people. [00:11:16] NVIDIA today has AI all over our company. [00:11:19] We're hiring more people. [00:11:20] We're moving faster. [00:11:22] We're more and more ambitious. [00:11:23] The type of things that we used to think was going to take 10 years, I now think it's going to take one or two years. [00:11:29] And so your ambition has to be elevated by the technology of the time. [00:11:36] If you don't engage the technology of your time, you'll just simply be left behind. [00:11:41] And so my recommendation is to engage it. [00:11:44] Now, the future, when people think about job loss and AI, what they see is that there's only so much work in the world to do. [00:11:57] There's so much, only so many stories to write, only so much code to write, only so many products to design, only so many things to enjoy. [00:12:08] And that is completely, obviously false. [00:12:12] The number, our ambition, if great, then automation and AI, which is the ultimate version of automation, would elevate your company, would elevate GDP. [00:12:24] As a result, you know, we should be able to bring more prosperity. [00:12:28] But what kind of jobs do you envision that AI will create? [00:12:32] Well, look at all the jobs that are being created already. [00:12:35] Number one, remember, I gave you the framework of AI. [00:12:40] AI is a five-layer cake. [00:12:42] It needs energy. [00:12:43] The country, the company, the region, Taiwan, the island, if limited by energy, would simply be left behind. [00:12:53] You need energy in order to enjoy this new industry. [00:12:59] And when you have energy, when you need them, AI is now so much market-driven initiatives, so much market-driven incentives, [00:13:07] that the amount of investment going into the energy sector is upgrading our energy grid. [00:13:13] It's investing in sustainable energy all by itself for the first time. [00:13:17] All you have to do is look at the stock price of all of my partners, GE, Vernova, Mitsubishi, Siemens. [00:13:27] Everybody, anybody who's in the power generation, energy generation industry, their stock prices are going up. [00:13:34] They're hiring more people. [00:13:36] Their revenues are increasing. [00:13:37] The utilities are seeing so many opportunities to invest in their power grid. [00:13:43] The next layer up is chips. [00:13:46] Every single partner of NVIDIA stock prices tripled in just a couple of years. [00:13:51] I'm so happy to see everybody's success. [00:13:54] And so everything from AI chips to DRAMs to, of course, TSMC and packaging and power regulators and cooling systems, you name it, [00:14:05] even multi-layer ceramic capacitors, incredible success. [00:14:10] And so you could see it's, of course, creating a lot of jobs. [00:14:15] The next layer above that is infrastructure. [00:14:18] Look at how many data centers are being built around the world, how much land being dedicated to building data centers. [00:14:25] And so electricians and plumbers and construction workers, architects and designers and technicians and networking engineers, [00:14:33] just a number of jobs, hundreds of thousands, just in the United States alone, and so probably millions around the world, and then just keep going up every single layer. [00:14:45] And then one more statistic. [00:14:47] Last year was the single largest year of venture capital investments in human history. [00:14:54] One hundred billion dollars in just one year went into AI native companies, a whole bunch of jobs being created. [00:15:01] And so there's absolutely every single evidence that AI is creating jobs. [00:15:06] And one of my favorites is literally radiology. [00:15:10] If you look at radiology and all the radiologists looking at workstations, looking for disease, looking for anomalies in the scans, [00:15:18] in the last five years, the last 10 years, it was predicted that it would be the first job to completely be wiped out. [00:15:26] Well, the researcher who made that prediction was absolutely right. [00:15:33] AI has completely revolutionized radiology. [00:15:36] AI has integrated into radiology in every single respect. [00:15:40] However, the number of radiologists grew, the demand for radiologists increased, and the pay of radiologists went up. [00:15:48] And so my point is, it is very more, it's more likely that AI will elevate your job, elevate the purpose of your job, [00:15:57] if you became expert at it, try to engage it, don't be afraid of it. [00:16:02] Of course, the industry has to be really thoughtful about building AI in a safe way, in a guard-railed way, [00:16:09] and make sure that it's deployed in a proper way, and that the end applications, [00:16:14] whether it's in healthcare or transportation or manufacturing, [00:16:17] where air travel or whatever it is that applies AI, those industries have lots of regulations already, [00:16:25] and they should reevaluate the regulations to be AI-ready, to be AI-prepared. [00:16:29] And so everybody has to be part of this. [00:16:33] You might remember some stories of the past. [00:16:36] So, for example, when the automobile came along, [00:16:42] and, of course, there was an entire industry of horses and carriages, [00:16:47] and that industry was quite large, of course. [00:16:50] They've been around a long, long time, hundreds of years. [00:16:53] And then, of course, the invention of the car, of the automobile came along. [00:16:57] Well, because it's so transformative, every infrastructure had to be reinvented. [00:17:03] Number one, a whole bunch of new factories created a whole bunch of jobs. [00:17:08] Every single road had to be enhanced. [00:17:10] The old roads were problematic. [00:17:13] New social norms. [00:17:15] For example, for a while there, it was believed that cars were killing children. [00:17:22] Cars were somehow killing children. [00:17:24] And what was discovered was, because in the older times, children would literally play in the streets. [00:17:30] I still remember when I grew up, I just played in the streets. [00:17:33] And so children would play in the streets. [00:17:35] And, of course, when cars came along, many of them didn't realize it, and they got themselves injured. [00:17:42] And so new norms, pedestrian laws, pedestrian rules, I guess, sidewalks, you know, so on and so forth. [00:17:49] And so we built up all these different social infrastructure and new technologies with self-breaking and so on and so forth. [00:17:59] All of these technologies, and, of course, social norms to know, don't run into the streets, you know. [00:18:04] And so all of those different conditions had to be built up to support a whole industry. [00:18:11] Electricity was the same way, and so on and so forth. [00:18:14] Artificial intelligence will be the same as well. [00:18:16] And for all the parents out there, what do you suggest their kids should be studying in order to stay relevant in the age of AI? [00:18:25] I think that it won't matter. [00:18:28] All the things that used to matter are still things that are going to matter in the future. [00:18:32] For example, broadcasts, newscasters. [00:18:38] It will be just as important to inform people as the past, to know what questions to ask. [00:18:43] You've prepared a lot of questions, but the best reporters and the best interviewers are the ones that are prepared, but you're also staying in the moment, listening to my question, and picking up on something that maybe the audience would be interested in. [00:18:59] And so you're thinking about multiple things at exactly the same time, just like I'm doing right now. [00:19:04] I'm thinking about the audience. [00:19:05] I'm listening to the questions. [00:19:07] I'm considering everything that's happening in the world at the moment. [00:19:10] And the answers to my question has to be consistent with that. [00:19:14] And so all of that is completely true still. [00:19:17] The ability to tell a story for an audience will remain just as important in the future as it is today. [00:19:24] Whether it's arts, the arts will be just as important. [00:19:29] Wabi-sabi, the beauty of imperfection, will probably be even more important in the future than today. [00:19:35] And so making movies, designing cars, and making chips will all still be as important as the past. [00:19:43] Yeah, yeah. [00:19:43] So whatever you decide is your passion, the only one thing that you have to do is to make sure that you ask yourself, [00:19:51] how can AI help elevate my learning, my craft, my purpose? [00:19:57] Now let's talk about China. [00:20:00] You were in Beijing with President Trump recently. [00:20:03] Can you take us through your trip on Air Force One and the meetings on the ground? [00:20:09] What are some of your biggest takeaways? [00:20:12] First of all, it was a great honor to represent the United States, and it was a great honor for me to accompany President Trump. [00:20:21] He called me in the morning. [00:20:22] He didn't realize I wasn't going. [00:20:24] And he insisted that I get on the plane and go. [00:20:29] And so I packed up in a hurry. [00:20:32] He called me in the morning as he was leaving. [00:20:34] And he thought I was in Washington, D.C. to jump on Air Force One, but I was in the West Coast. [00:20:39] And so he told me, just, you know, meet me in Alaska. [00:20:43] And so I flew to Alaska, jumped on Air Force One, and went to China. [00:20:49] I was there with 16, I think, other CEOs. [00:20:54] And it represents quite a large collection of great companies, from consumer electronics, of course, to industrials and automotive and all the way to financials and biotech. [00:21:11] And so it was a large, large collection of industry that was represented. [00:21:16] And the conversations were, one, really, really welcomed. [00:21:24] President Xi, Premier Li Qiang, were extremely welcoming. [00:21:29] And they spoke about cooperation and a stable relationship and that China would be an open market and even more wide open than before, encouraged investment. [00:21:45] And then we were there to really represent the United States, support the president. [00:21:51] And they had wonderful meetings. [00:21:54] They were extremely cordial. [00:21:58] The feelings were great. [00:22:00] And the pomp and circumstance and all of the festivities were quite impressive. [00:22:08] But that was it. [00:22:09] That was basically it. [00:22:09] I was there for a couple of days. [00:22:11] We didn't sleep for a couple of days. [00:22:13] And then I, when President Trump left, left to go back home, I went to old town Beijing to enjoy some nice meals. [00:22:24] How was that? [00:22:25] It was great. [00:22:26] It was great. [00:22:27] Yeah, it was, people were very nice. [00:22:31] And anyhow, it was a really, really great experience. [00:22:36] But reports suggest that China hasn't purchased any of the NVIDIA's H200 chips, despite U.S. export approval. [00:22:47] Now, so you have also said that NVIDIA has stepped back from China as it pushes for its own chips. [00:22:56] Could that actually eventually make China your greatest rival? [00:23:04] Well, China is going to be everybody's greatest rival. [00:23:07] I mean, that goes without saying, because they have such an extraordinary local market. [00:23:14] It is completely uniform. [00:23:17] Just like the United States, every single state speaks English. [00:23:20] China, every province speaks Chinese. [00:23:22] And that's a substantial advantage versus Europe, where every single country speaks a different language. [00:23:30] And or countries where everybody speaks the same language, but the market's not very big. [00:23:36] And so these two countries, United States and China, have a substantial advantage because of that reason. [00:23:42] Large population, very well-educated, the science, technology and math of Chinese students is extraordinary. [00:23:51] And they produce it at very high volumes, large volumes. [00:23:55] Tsinghua Universities and many of the other universities are world-class in science and technology today. [00:24:01] And so, one, you have to recognize the country is large. [00:24:05] It has a uniform culture. [00:24:09] And so, and they have really treasured and valued science and technology. [00:24:16] The companies are very competitive, super vibrant, incredible companies from Alibaba to Xiaomi to Tencent, Baidu. [00:24:25] And these amazing companies, large and also so many small companies that are so vibrant. [00:24:30] And so, I think it goes without saying that China's pace of innovation and its own natural resources, including its people and its culture, [00:24:44] will almost certainly guarantee that China will compete with every industry, and they are very competitive. [00:24:53] And so, so I think that goes without saying. [00:24:56] With, with, with respect to us, I, we're not stepping back from China. [00:25:04] I, I would say, I would say that the Chinese government, when we were banned from going to China, going to China through expert controls, [00:25:13] it left a vacuum that the Chinese companies were able to fill. [00:25:18] And as a result, Huawei and many of the startup companies in China had record years. [00:25:24] They were now growing at an incredible pace. [00:25:28] And, and in our absence, even though Nvidia's technology is better, it is, in our absence, available technology is the best we can get. [00:25:37] And it's plenty good. [00:25:39] And so, so I think, I think my only statement was that, that the Chinese government is, is, understandably, as every government would and should, [00:25:50] encourage and want to have conditions for their local companies to be successful. [00:25:55] However, I also believe that Nvidia could add enormous amount of value to the Chinese market. [00:26:01] And it's good for the United States. [00:26:03] And President Trump has been very clear that he wants American companies to succeed all over the world, just as every, every country should. [00:26:12] Every country should maximize exports. [00:26:14] Every company, every country should want their companies to succeed worldwide. [00:26:19] And, and, and, and so it's sensible that, that United States would want Nvidia to go back to, and serve, serve every single market. [00:26:28] For China, Nvidia's technology competes with one layer of the five, of the five layer cake. [00:26:36] We compete with one layer of the five layer cake. [00:26:38] But don't forget, AI's a five layer cake. [00:26:41] And so when, when Nvidia is participating in China and serving the Chinese market, as we have in the, in the past, it supports the expansion of the other five layers. [00:26:50] And so when you look at, look at the market in a more holistic way, and, and I, and I surely believe that the, the, the leaders will, and the local markets will, when you look at it in a holistic way, Nvidia could be of great service to that industry. [00:27:07] And we could add a lot of value to the China market as well. [00:27:10] And so, so I think when, when, when our, when Nvidia grows, the entire supply chain grows, when Nvidia grows, as you, as you see, we lift all of Taiwan with us. [00:27:21] And, and so Nvidia succeeding around the world will be great for the supply chain, and it will be great for all the local markets. [00:27:29] But would you say that we're seeing the rise of two separate AI ecosystems, one led by the U.S. and the other by China, and can they coexist? [00:27:40] It is possible that, that two AI ecosystems are created, but it's not wise. [00:27:49] AI is obviously a very, very capable technology, it's transformative technology, and it's dual use. [00:27:57] On the one hand, it's incredible potential for good is unbelievable, and it's demonstration and transformative capabilities for good is already demonstrated to be incredible. [00:28:12] And, and I, and I have every confidence that it will continue to do so. [00:28:16] On the other hand, it could also be used in all kinds of other ways. [00:28:21] It could be used for ways that, that could bring harm. [00:28:26] And so the more, the more that the world leaders in technology, as well as in, in, in, in, in, in social leaders and political leaders, the more that all the leaders work together, cooperate together to advance this technology in harmony and for good, the better it will be for the future. [00:28:52] And so I, I, I have every, I believe that, that China and United States should cooperate in AI, not just compete in AI, and it is possible to do both. [00:29:02] I work with so many competitors today. [00:29:06] They're competitors of mine, but I also compete with them on the one hand, and we cooperate greatly on the other hand, because we want what's in the best interest of the industry. [00:29:16] We want what's in the best interest of the market, and we want what's best interest for society. [00:29:21] And so therefore, cooperating ensures that harmonious advance. [00:29:27] And so I, I think that it is absolutely true for AI that on the one hand we compete, on the other hand, we should absolutely cooperate. [00:29:35] And we know that Taiwan sits in the center of the global supply chain and what do you, what kind of role do you think Taiwan plays in the AI race and how long do you think Taiwan can maintain that advantage? [00:29:48] Well, Taiwan has the letters AI in it and so AI is at the center of Taiwan and this, this, the companies on this island are, are incredible, as you know, the companies are at the epicenter of the, the possibility of AI, and the growth of AI, the advancement of AI. [00:30:13] Uh, in the last several years, uh, several things are happening at the same time on the one hand, supply chains have to be diversified because the world supply chain is so broad, so large. [00:30:26] And because AI, the computer is not just a tool for you and I anymore, the computer, what Nvidia makes is now a factory and infrastructure for the world. [00:30:40] And, and therefore the supply chain by nature would have to be more resilient, more diversified because it's going to be so large. [00:30:49] And so on the, on the one hand, uh, the supply chain is diversifying around the world. [00:30:54] Fabs are being built all around the world. [00:30:57] Chip plants, packaging plants, computer plants are being built all around the world on the one hand. [00:31:02] On the other hand, it's growing incredibly here in Taiwan. [00:31:06] The energy pressure is high. [00:31:08] The land pressure is high. [00:31:10] And, uh, of course, uh, Taiwan is extraordinary at manufacturing. [00:31:15] And, uh, uh, and I hope we talk about robotics and how robotics is going to revolutionize and transform Taiwan's ability to continue to grow at an incredible pace. [00:31:27] Uh, no, no, no, no region, uh, is, is, uh, uh, better prepared for the continued growth of AI than Taiwan. [00:31:36] And I, I, I absolutely, uh, uh, continue to expect Taiwan to grow at an outpaced, you know, at a very accelerated pace, uh, for many years to come. [00:31:48] And, uh, and, uh, and, uh, and it's an investment of people, you know, of course, um, at the scale that we are today, NVIDIA is a multi hundred billion dollar company. [00:31:58] Now we're one of the largest companies in the world, and we're growing at nearly a hundred percent per year at the scale that we're talking about. [00:32:06] And so the, the, the partnerships that we have here in Taiwan, how they've, uh, helped me grow and support our growth is incredible. [00:32:16] And then, but the, the most important thing is the investment in people. [00:32:19] Uh, we have a very large site here and we're going to have a much, much larger site here soon. [00:32:24] Can you give us a sneak preview on the investment amount? [00:32:28] Well, you know, it's not, it's not just about, you know, the money. [00:32:33] I mean, of course, of course, NVIDIA spends at this point, you know, hundreds of billions of dollars in AI infrastructure. [00:32:44] And most of our spend is in Taiwan. [00:32:48] And so we've, we've invested through the most, the best form of investment, which is business, uh, hundreds of billions of dollars into Taiwan. [00:32:59] And if you look at our next several years, uh, between Grace Blackwell and Vera Rubin, we have $1 trillion of sales. [00:33:10] $1 trillion of sales is hundreds of billions of dollars of spending equipped revenues that will come into Taiwan. [00:33:19] It's just an extraordinary number and the highest ever in history for Taiwan. [00:33:25] And so, so that's, that's our, our best form of investment. [00:33:29] Uh, we, of course, also invest in companies here. [00:33:32] We support them with prepayments. [00:33:34] We support them with investment in their capital. [00:33:37] We support them with commitments in our purchase. [00:33:40] And then, uh, my favorite form of investment is still ultimately people hiring a lot of engineers here and having a lot of, uh, great employees here in Taiwan. [00:33:49] And also among all the business leaders in Taiwan, I know you're very close to Morris Chang, TSMC founder. [00:33:57] And can you describe your friendship and also what have you learned from him? [00:34:02] I, I learned from him every time I'm with him. [00:34:04] I was with him last night, uh, Morris, Morris and Sophie had, uh, my parents, Lori and I, and, and, uh, Madison were with us as well. [00:34:13] We're over to, for dinner. [00:34:15] We had a wonderful time together. [00:34:17] Um, you know, of course, Morris and I have a lot in common, uh, because, because, uh, uh, we grew, you know, I grew up, uh, with the support of Morris. [00:34:27] And, and, uh, when I came to Taiwan, uh, for the first time after I left, when I was five years, five years old, uh, I came back for the first time to see Morris. [00:34:41] And, um, I've, I've been, I've been coming back a lot since then. [00:34:45] And so, so, um, uh, you know, without, without TSMC and Morris, uh, Nvidia wouldn't be here today. [00:34:53] And it was the support and the friendship and, and the, and the risk that they take. [00:34:58] And back in the old days when we were quite small, um, all, all of the things that, that, uh, and our growth together, innovation together. [00:35:06] And, and of course, uh, the creation of, of the AI industry that we know today together, um, all of that, all of our journey and all of our history, uh, are highly intertwined. [00:35:17] And so we, we, we, you know, we, in a lot of ways, um, I grew up with the support of Morris and the support of TSMC. [00:35:24] Uh, we have a lot of interest in, in how the industry is forming. [00:35:28] Um, we have a lot of, we have a lot of interest in, in, uh, ensuring the continuous success, uh, of the industry. [00:35:35] And of course, you know, we have a lot of, a lot of wonderful life stories to share with each other. [00:35:40] And, uh, he, he's a, he's an incredible, uh, avid reader. [00:35:44] And so it's always fun to, to hear about, you know, his, his, uh, summaries of the books that he's read. [00:35:50] And, and, um, as you know, and, and so all of, all of that is great. [00:35:54] I always, I always treasure my time with him and, uh, he's, he's doing great. [00:35:58] Uh, and, um, uh, we always have a nice, nice glass of whiskey together. [00:36:03] That's nice. [00:36:04] Um, now let's move on to the fun part. [00:36:07] Let's talk about you. [00:36:08] I'm born in Taiwan, lived in Thailand before moving to the U.S. [00:36:13] How have those experiences shape who you are today and the way you lead? [00:36:18] Uh, I, I guess, I guess we're all the products of our parents. [00:36:23] Um, my father is, is very technical, uh, very precise, uh, does work perfectly. [00:36:32] And, uh, perfect handwriting does everything perfectly. [00:36:37] Uh, you know, his, his level of craft and precision and just the way he does everything in this life. [00:36:43] Um, uh, is something that I had the benefit of learning from. [00:36:47] And, and, uh, on the other hand, uh, my mom's, uh, obsession for details. [00:36:53] Uh, uh, just a, a deep obsession for, for, um, you know, she, she has a personality that can't let anything go, you know, and, and, um, uh, in a lot of ways, I, I think I've picked up that part of her, her, uh, her, her behavior. [00:37:10] Um, and, and so, you know, I, I can focus on something for a very long time, like NVIDIA for 33 years. [00:37:15] And, uh, and, uh, uh, every day I'm equally obsessed and equally, you know, equally, uh, intense about, about doing a good job. [00:37:23] And, and so I think I had the benefit of, of, uh, uh, uh, growing up, uh, with a lot of those Taiwanese characteristics, you know, um, they're, they're, they're very much like a Taiwanese, Taiwanese, uh, parent. [00:37:37] Um, I, I also think that, that we grew up, they, they sacrificed a lot, uh, leaving Taiwan and, and, uh, left their, left families behind. [00:37:48] And, uh, they went to Thailand with very little and they went to United States with nothing. [00:37:53] And so, um, uh, when they went to United States, uh, we were basically alone. [00:37:59] Uh, there were no friends, no family, uh, you know, to be there with us. [00:38:03] And, um, uh, and because they were, they were so modest, uh, and we had so little, uh, and, uh, uh, living in United States, going from Taiwan is, [00:38:18] is very challenging because the cost of living is so much higher. [00:38:21] And, and so I, I think, I think, um, our humble, humble, uh, life and, uh, uh, and just seeing my parents, uh, struggle and, and, um, uh, finding a place in, in a strange world. [00:38:37] Uh, and the risks that they took so that we can, we can grow up in America and enjoy the American dream, uh, and have the opportunities that led to, led to, uh, NVIDIA today and led to me today. [00:38:49] Uh, that, uh, all of those sacrifices, um, uh, instills, instills, uh, uh, a character in you. [00:38:58] And so, uh, I'm grateful for, uh, all the sacrifices and the hard work that they, they, uh, uh, they made so that I could be here. [00:39:08] And so, but I, I think that that journey was really an important part. [00:39:12] In your company, you're known to be a tough boss and you've said that, um, you'd rather torture people to greatness than to fire them. [00:39:22] How exactly do you do that? [00:39:24] I, oh, well, it's not, it's not physical torture. [00:39:27] It's, it's the, it's torture the same way that Taiwanese parents torture people. [00:39:33] You know, you know, in a Taiwanese parent, nothing is ever good enough. [00:39:38] And, and, uh, you can't go, you can't go a day without, without some criticism. [00:39:47] And that's the Taiwanese way. [00:39:50] And I'm kind of the same way. [00:39:52] Uh, you can't, you can't show me something without me giving you some criticism. [00:39:57] And, and I guess in a lot of ways that that's, that's my form of torture. [00:40:01] You know, I, I will, I will give you my feedback. [00:40:04] Um, always, always, uh, immediately. [00:40:07] Um, I'll never save it away. [00:40:09] Uh, once I give you my feedback, just like a Taiwanese parent, once the feedback is given, you're back to loving the person again. [00:40:17] And, and so you're, I'm, I'm always, uh, I'm always, uh, uh, critical of, of everybody's work so that I can help them be better. [00:40:26] Uh, I want them to be better. [00:40:28] I know they could be better. [00:40:29] And, and so that's where, where it's coming from. [00:40:33] And, uh, you know, maybe, maybe a lot of that is my Taiwanese parent. [00:40:37] And how do they, how do your employees respond to that? [00:40:41] Well, you know, Nvidia's, Nvidia's, uh, retention is the best in the world. [00:40:46] Uh, uh, we have, we have employees that have been with me now for 33 years. [00:40:51] And so, uh, people don't, people don't, uh, don't quit easily from, from Nvidia. [00:40:58] And, and, uh, uh, you know, we, my job is to create an environment where all of these amazing computer scientists and engineers and, um, supply chain experts can come to Nvidia to do their life's work. [00:41:13] And that's my job, uh, to create that condition so that they could realize their dreams just as the, just as they've helped me realize my dreams. [00:41:21] And, um, uh, and I, I think that that's ultimately the purpose of leadership is to create the conditions for other people to realize their dreams. [00:41:31] Uh, to, to be part of yours, of course, and to be part of something bigger, uh, but to, to be able to turn their, their, what is their job, their profession, their craft, hopefully into their life's work. [00:41:44] And, um, at Nvidia, you could do that. [00:41:46] And we know that you work seven days a week. [00:41:49] Um, what drives you day to day? [00:41:52] And where exactly do you get all that energy? [00:41:55] I'm exhausted all the time. [00:41:58] Um, uh, what, what drives me? [00:42:02] Uh, I mean, it's, it's somewhat, it's several things. [00:42:06] It, it's on the one hand, uh, I, on the one hand is as, as basic as I don't want to fail and I don't want Nvidia to fail. [00:42:15] And, uh, I don't, I don't want Nvidia to fail because we have too many, uh, too many people who, who are counting on me. [00:42:22] And so, uh, you know, it's employees, it's our partners, it's our ecosystem partners, it's, you know, all of my friends here in Taiwan. [00:42:31] I want everybody to succeed. [00:42:33] And so there's, there's a, there's a, a burden on, on leaders, uh, that want everybody to, to, to flourish and to realize their dreams and to, you know, to hopefully, uh, succeed with us. [00:42:47] And so there's a, there's a, there's a burden that comes with that. [00:42:51] And, and, um, that gets me out of bed every day. [00:42:54] Uh, at the same time, there's a, there's a hopeful, optimistic, um, uh, ambitious part of me that, that wants to build something that, that makes an impact, makes a contribution. [00:43:09] Uh, there's a, there's a dreamer part of me that, that wants to create that future and hope, hope to see it in my lifetime. [00:43:18] And so I'm in a hurry to, to, uh, uh, have it come true. [00:43:22] Um, and so it's simultaneously the burden of, the burden of, of, uh, of success and, and, uh, wanting other people to come along. [00:43:33] Uh, the, the fear of failure, uh, that, that somehow because of how I was raised, uh, because Nvidia has enjoyed, um, and gone through a lot of very difficult time. [00:43:46] It was not easy to build Nvidia. [00:43:48] It was not easy to, to, uh, uh, to be here today. [00:43:52] And, and all of the struggles can never leave my body. [00:43:55] You know, once, once you've struggled as deeply as, as Nvidia has, and as I, uh, as I've had the benefit of, it becomes part of your character and then never leaves you. [00:44:06] And so simultaneously, all of this is happening in my body and my brain. [00:44:09] And, and so you can't help but be, you know, always thinking about work. [00:44:16] Mm-hmm. [00:44:17] And you, uh, some people have said that you still have another 30 years running Nvidia. [00:44:22] What do you think about that? [00:44:24] And what kind of leader do you think should succeed you? [00:44:27] Uh, I, I, I would like to work as long as I can. [00:44:31] You know, I hope to die on the job. [00:44:35] That would, that would be the, that would be the, a dream come true. [00:44:40] And so, so, uh, um, I can't imagine a, a more meaningful life. [00:44:49] Uh, I'm surrounded by my family, my kids, uh, both of them work in Nvidia and, and I have the benefit of seeing them all the time. [00:44:56] And I'm proud of them and, uh, the company's proud of them and they love the company. [00:45:00] And so I get to see them all the time. [00:45:02] Um, you know, Lori, Lori's been with me for, for a long time, our whole life. [00:45:06] And, and, um, I, and she, you know, she's been part of Nvidia every step of the way. [00:45:12] And, uh, aside from, aside from Lori, I don't think anyone has been to every conference. [00:45:18] And she's, she's been to every conference, every speech, you know, and, and, um, uh, every moment of Nvidia, she was there. [00:45:25] And, and so I, I have the benefit of, of, uh, uh, meet really meaningful work. [00:45:32] Um, and a company with incredible people and, and employees that, that can achieve almost anything. [00:45:39] And I'm surrounded by the love of my parents and my family that, that, you know, are part of this journey with me. [00:45:44] And so, so I can't imagine doing anything else, frankly. [00:45:47] Uh, and I can't imagine doing anything else more meaningful. [00:45:50] Um, with respect to the next generation, uh, leadership is about creating the conditions, um, for other people to be empowered. [00:46:01] You know, Nvidia is giant, uh, and yet we're a small company. [00:46:06] We are the world's smallest, large company. [00:46:08] And, and that's only possible because so many leaders at Nvidia, so many people and empowered to do their, to do the work. [00:46:17] And so I constantly, um, uh, explain our strategy transparently. [00:46:22] I constantly reason about complicated situations in front of people so that they could help me double check my reasoning patterns and my reasoning logic. [00:46:33] And, um, uh, on the other hand, you're also exposing them through how you think through complicated and ambiguous, uncertain circumstances. [00:46:43] And so the combination of empowering people, letting them run, um, uh, aligning early on what our strategies are, our guardrails are. [00:46:53] And then reasoning in front of people, uh, really helps everybody become more successful, become more empowered. [00:47:00] Well, the next, the next generation, the next leader of Nvidia is already working there. [00:47:05] And, uh, and, uh, it's my job to cultivate hundreds of amazing choices and you just never know who the right next person is. [00:47:16] And, and, um, I, and the, and the reason for that is because we don't know what Nvidia is going to be like in 30 years or 10 or 20. [00:47:25] And we don't know what the circumstances are in this future that artificial intelligence and that we're help creating. [00:47:31] And so it would be, it would be presumptuous and, and arguably even arrogant to think that I would know what the perfect leader is going to be in 10, 20 years. [00:47:42] They were just, we're just going to have to evolve until we, we, uh, we know for certain. [00:47:47] But the one thing that I will tell you is that the leader, um, for Nvidia would have to care about, uh, uh, uh, the employees, the ecosystem, the partners more than they care about themselves. [00:48:01] And a great, uh, a great leader is selfless and, uh, always in service of, of, um, uh, of the culture of the company, the mission of the company, the people of the company, always in service of other people. [00:48:15] Uh, it is very likely that, that, um, uh, that, this, this leader will, will, uh, uh, uh, uh, be in a, in a, in an era, a time where intelligence is a commodity, that the ability to write code, solve problems, um, even scientific discovery is arguably completely automated. [00:48:39] The one thing that we value so much in my generation, which is knowledge and, and technical skills and problem solving, um, the sciences, technologies, the maths, um, all of that is largely likely going to be a commodity. [00:48:54] And so therefore what remains, what remains is ambition and character, imagination and, um, the empathy and generosity and kindness. [00:49:06] And, you know, the skills, the soft skills that, that defines a person beyond their intelligence. [00:49:15] And, and so I, I think those things will, will likely become, uh, quite important. [00:49:20] And then, and then, and I would say, um, may very well be, uh, the greatest asset of any leader, which is the, the desire of everyone around them to see them succeed. [00:49:34] There are, there are, there are leaders that you just, there are people that you just wish them to be more successful. [00:49:40] Because they're kind to other people and they're generous to other people and you just want to see them succeed. [00:49:44] Mm-hmm. [00:49:45] And there are people who are successful and you will hope to see them fail. [00:49:47] And, and so, so I, I think that the future leaders are going to be the ones that somehow embody the, the characters, um, the intangibles that, that leads other people to want them to succeed and, and, and be led and to be part of. [00:50:04] I see. [00:50:05] I hope you don't mind me going back to the question about job cuts. [00:50:09] How, how would you say that job cuts are inevitable because of AI? [00:50:15] It is very likely there will be more jobs in five years than there are today. [00:50:20] Many more jobs. [00:50:23] Some jobs will be different. [00:50:25] Some jobs will be gone and many new jobs will be merged. [00:50:29] And I can say that with almost certainty. [00:50:33] And the reason for that is this, the world has been growing. [00:50:37] The GDP of the world has been growing at about 2% per year, 2% per year for 150 years. [00:50:44] For 150 years, where there was the industrial revolution with electricity, information technology, transportation, steam, elect, you know, manufacturing equipment, all the way to personal computers. [00:50:58] And internet and cell phones and so on and so forth and our artificial intelligence. [00:51:05] It's been on a solid 2% per year. [00:51:08] And it's, it's incredible. [00:51:10] And, um, my sense is that, well, my belief is that, that, um, all of those technologies at the time were transformative. [00:51:18] Of course they were. [00:51:19] How you and I live our lives today and how 150 years ago people lived their lives and the industries, the companies, my goodness, uh, completely different. [00:51:31] And yet 2% per year, the number of jobs that have been created in the last 150 years, incredible. [00:51:39] The number of jobs that will be created in the next 150 years is going to be utterly incredible. [00:51:44] Of course, some jobs will be very different. [00:51:48] Let's, let's imagine that, um, prosperity. [00:51:51] Uh, AI has the opportunity to close the technology divide like no technology ever has. [00:51:57] For example, you are a better user of computers today as a news reporter than at any time in history. [00:52:06] Uh, because you can program the computer to do whatever. [00:52:09] If you want your own website, if you want your own broadcast, you can just ask the AI to help you create it. [00:52:16] You're now a programmer, just like I am a programmer. [00:52:19] And so for the first time, AI has closed the technology divide. [00:52:24] We have the opportunity to bring everybody along in this new technology revolution. [00:52:29] That's a very big deal. [00:52:31] The second. [00:52:32] Of course, AI will likely change the shape of the services industry, the manufacturing industry, and very also very importantly, the consumption industry, the entertainment industry. [00:52:46] It is more likely that, um, people are going to consume more because we're more prosperous and because we're going to consume more. [00:52:55] It is very likely that the leisure industry and, um, you know, the creation of content will enjoy more because we'll have more time to enjoy the content that you create. [00:53:05] And, uh, hopefully we'll have the opportunity to travel more so that people are traveling around the world, learning about different cultures. [00:53:13] And we can close the culture divides and, uh, you know, and so there's a, there's an optimistic future that, uh, that, uh, I imagine and facts would support my imagination of that future. [00:53:29] And, um, and, um, and, um, and, and I think the, the narrative that connects AI to job loss for many of the CEOs that are doing it. [00:53:39] Um, it is just too lazy. [00:53:41] It's just too lazy. [00:53:42] It's just too lazy. [00:53:43] AI has just arrived. [00:53:44] How is it possible? [00:53:45] They're already losing jobs. [00:53:46] You know, how is it possible that AI became productive and useful only six months ago? [00:53:53] And they were, they were canceling, they were somehow laying people off two years ago because of AI. [00:53:59] It doesn't make any sense. [00:54:01] It was just, it was just a way for them to sound smart. [00:54:04] And I really hate that. [00:54:05] Um, I would, I think we're scaring people and that's irresponsible. [00:54:09] I think we should tell a balanced story, a balanced narrative about the potential of this cape, of this technology, the importance of advancing it safely, security, guardrailed, [00:54:22] with necessary social, political, government, government, industrial policies to ensure that it advances safely. [00:54:31] On the other hand, tell a story that's optimistic so that people want to be part of it. [00:54:35] We want the young people of this generation to engage it. [00:54:38] And so here's my test. [00:54:40] Here's my simple, simple test for all the audience. [00:54:45] What are you advising your children to do? [00:54:48] Are you advising them to engage AI or to reject AI and be left behind? [00:54:59] Are you advising them to use AI to elevate their learning, to elevate their profession when they can? [00:55:08] Or are you telling them, don't use this technology and let other people use it? [00:55:13] And so I have no question in my mind that every single parent is saying, you have to learn about this technology, use it wisely, use it wisely, but use it to elevate yourself. [00:55:25] If that is what they're advising their children, for what reason are they not advising it for other people? [00:55:34] So I would recommend that I see a future that has more abundance, abundance of intelligence, abundance of labor, abundance of goods. [00:55:47] It's going to therefore create opportunities for us to consume more. [00:55:52] So if you were a connoisseur of fashion, you're going to have a lot more opportunity to consume that. [00:56:00] Thank you so much, Jensen. [00:56:02] I really enjoyed your conversation. [00:56:04] Thank you very much, Victoria. [00:56:06] Thank you.