Kai-Fu Lee on the Great Game to Dominate Data

Published Oct 11, 2018, 8:00 AM

Don't fret too much about tariffs and trade deficits. The real competition between the U.S. and China will be in artificial intelligence and data, says Kai-Fu Lee, author of "AI Superpowers: China, Silicon Valley and the New World Order." Lee explores who wins in this struggle for influence, and how it affects workers.

For all the focus on deficits and tariffs, you might think economic ties between the US and China are just about how much stuff is put on a ship, how much it costs, where it's made, and where it's going. You would be mistaken. The future belongs to data, not merchandise, and the world economy is a duopoly. America and China own it. Welcome to Benjamin, a show about the global econome. I'm Daniel Moss, columnist at Bloomberg Opinion in New York. A few people are as familiar with both the viewers and China as Kai Fu Lei. Born in Taiwan, Doctor Lee has spent his career straddling Silicon Valley the tech precincts of Beijing. He used to run Google China and now leads sign Ovation Ventures, a firm that focuses on developing the next big thing. Along the way, he had a personal crisis that further sharpened his insights. Dr Lee is the author of Ai Superpowers, China, Silicon Valley, and the New World Order. It's probably the business book of the moment. Dr leeh thanks for joining us, and thanks for not putting the word disruption in the title. Thank you. It's great to talk to you. I've heard you describe China as the new Saudi Arabia and data as the new oil. What did you mean by that and how does that capture the essence of this moment? Sure? The way are the official intelligence or AI works today is that within this single application, when an AI algorithm is trained down a huge amount of data with corresponding labels, it can then make decisions in that domain at a superhuman accuracy. So the more the data, the better. So the system is not programmed by humans to do make decisions, but it learns from data. And the amount of data that's needed is is huge. So today China has the most data in the world and that gives a certain advantage to China. Now, technology, business, and economic literature has been gushing about artificial intelligence for a while. What's changed. I think the biggest change was that ten years ago a new technology called deep learning was invented and it was applied in many many domains, and starting about six years ago, we began to see its efficacy in face wreck cognition, later speech recognition, later the game of Go, and almost all subsequent technologies were built on this deep learning or associated technologies. So that was the single big breakthrough that became clear to the researchers about six years ago, but to the rest of the world about two and a half years ago when Alpha go emerged, and this technology coincided with the rapidly increasing amount of data because deep learning is a very deep network with potentially billions of parameters to train, so it was hungry for data. So the accessibility and growth of Internet data alongside with that the development of deep learning together created never before seeing accuracy in all kinds of tasks that deep learning plus large data tackles. Can you put this in the context of economic history for the past couple of centuries? Where does this rank in the period that began with say, the Industrial Revolution in the United Kingdom? Okay, I think AI should definitely rank among the Industrial Revolution, in particular with the steam engine, electricity, and perhaps the internet computer revolution. These probably are the top three or four in my personal opinion. It will end up proven to be number one, but of course that remains to be seen. Does the industrial and economic map of the world now resemble something from the nine century but instead of European powers gobbling up territory in Africa and Asia, China in the United States carving up the technology world. What does that map look like today? I think if we take a snapshot right now, US is still has the upper hand with American technologies penetrating developed countries and some developing countries, and China is largely still within China, with a little bit of penetration into other countries. But we're seeing a rapidly increasing pace that Chinese technology companies, and that's primarily mobile apps and AI companies going from China to Southeast Asia, Middle East, and Africa. They're just at the beginning, so it's too early to tell the final results. But given that the American giants have largely ignored those regions. Um as you add India also, since that's generally separate from the Southeast Asia, US is generally not paid too much attention to those four regions. I think it gives China an opportunity to go into these four regions. And in addition, the Chinese demographics may match those regions better than the US, so the products may be more attractive. And coincidentally, it also matches the Belt Road initiative from China. So I think the developed countries will end up continuing to be a stronghold of American dominance in technology. But I think most of the rest of our world would potentially have a high degree of Chinese penetration, probably not as strong as US in the developed countries, but let's say a more than fifty percent penetration or dependency on Chinese technologies. Over let's say the next five to ten years. So broadly speaking, the US would have Western Europe, Canada, and Japan, and China would have the rest, probably not quite I think um. I think the US would have nearly total penetration of most of, if not all, of Europe, Canada, Australia, and Japan, and possibly some of South America and a little bit at the rest of the world. And China would have all of China and more than fifty of India, Southeast Asia, Middle East, and Africa. The trade disputes between the United States and China, which dominate much of the public discourse in America these days, that principally about merchandise trade, about deficits, about the cost of production. That seems to be a completely different conversation from that of the world that you're inhabiting. Well to be fair. There's also intellectual property, which I think is crosses both. I do agree that the trade disputes are include the intellectual property don't affect um our area, which is AI mobile software. But some of those issues certainly do affect semiconductor PC UM mobile phones and so on, which are very much related to the area as we invest in now. You were born in Taiwan, you came to the United States as a small boy to Tennessee. You've worked in the Valley, You've worked for an American company in China, you have your own venture capital firm. As a student of the United States and the economic and technological currents, what from your perspective, is going on in this country right now? How would you define this moment? Do you mean the technology development or the trade disputes? If you've wrapped up the overall vibe in America right now compared with previous visits, from the time you were growing up, from the time you were a student at university, here, what's changed in this country and why? I think in Silicon Valley not much has changed. There has been and still is a very strong self confidence that it is the center of the world. The rest of America is beginning to realize that there is an alternate technological force emerging that's called China. And I think the Silicon Valley centric view is potentially dangerous because it is a great view. It is a great way to develop products, but China has proven that it is not the only way. So it would be wise to consider at least the two alternate ways that great innovative products can be developed. And in terms of overall sentiment, I think I sense more divisiveness in America. I think before there was a much greater unity, and I think there is um more disagreements today, and I think there are it feels like they're more uncertainty, but the economy is good and the people feel confident about the American technologies. Yeah, that's basically what I see. How profoundly did your brush with cancer effect your perspective? Very very profoundly. In my first fifty plus years of my life, I've I've really was a workaholic, and I felt work was the center of my life. And family was UM a set of people I depended on, but I gave them time as I could afford from my work. UM gave enough times so that they tolerate my lack of attention. But as I found out that I had cancer, I realized that all the accomplishment really meant nothing, and that, like most people facing the possibility of death, I realized that what was important in life was first and foremost to love and give love back to the people who love me, and secondly, to follow the things and do things that I am passionate about. And that's very consistent across thousands of people facing death, and I think there's a lot of wisdom in people facing death. So fortunately, I'm now in remission and when I now come back to work, I no longer put work as my only priority, and when my kids come home, I actually just don't work, and it's a reversal of what I used to do. I still spend a lot of hours working less than before, but it's a matter of priorities. It's a matter of finally realizing what was important. And it also made me realize that workaholism is something that haunts both the Americans and the Chinese, and that there really is more to life than work. So in some sense, uh, as we enter the era of AI, I think AI is here to relieve us from a lot of redundant or repetitive or routine work that we're doing, giving us a lot more time back to us so that we can give time to the people who love us, We can do the things that we are passionate about, and and we can have time to think about the real meaning of life, which is definitely not work. In keeping with that, what does this new technological world order main for the future of work? Because AI is capable of doing single, distinct tasks much better than people, that means we don't have to be threatened that AI will become robot overlords ruling over us because there are tools that we control. On the other hand, if they can do single tasks better than us, there are many people who do jobs that are composed of a number of single tasks. That means those jobs will be displaced by AI, either one on one or through industry disruption. So in terms of the future of work, I'm quite concerned that a certain percentage of jobs that are not requiring the most creativity and not requiring the greatest human touch and interaction, those kinds of routine jobs will be replaced by AI, and that the time has come for us to plan to real to reskill the people in those jobs, and while we still have time to create jobs that can be both satisfying and reasonably play paying, so that people can move on to the next steps in their lives rather than just to see AI take over one type of job after another, Doctor Lee. You're probably familiar with some of the criticism that's made of Northern California and in some ways the state in general, that the scene is to uniform, there's too much group think, it's a one party state, and it's lost its age. What's your perspective on that about California, about Silicon Valley California in general, versus what you're seeing when you hang out in the tech precincts of Beijing. Well, I am very conflicted because at the same time, I have tremendous respect of the vision that's come out of Silicon Valley, that has led the world for the last thirty years, and that still has many of the world's most amazing companies and startups. But I I'm also at the same time concerned that the success has created some degree of self entitlement and maybe even a little bit of hubris, the feeling that the world revolves around Silicon Valley. They belief that um only the companies there matter, the belief that only Silicon Valley innovates, and only the method of Silicon Valley that matters when it comes to innovation. And I think that is myopic. And if if we, if Americans all really believe that, then it would be missing all the exciting things happening in China. Chinese model of innovation is different. It's heavier, it's building impregnable business models that actually are more built to last, but maybe less visionary and technically exciting. I think as citizens of this world, we should all be hum ball and study different types of innovation and be open to different models. And if Silicon Valley continues to only look within and not accept external innovation, then it will be missing half of the teaching materials for the young people in Silicon Valley. That I can assure you that the young people in China are studying humbly both the successes in the US, the Googles and facebooks and the successes from China the Ali Baba's and made Twins and ten cents, and that gives a more robust set of instructions, if you will, for the Chinese entrepreneurs that there's no reason American entrepreneurs should um keep the blinders on and throw away half of the instructions that could be so helpful to their growth as entrepreneurs. What role should the state play in this? I was fortunate enough to be at an event last month with Audrey Tongue, Taiwan's Digital minister, whom you probably know. Uh. That was a very special experience. It should China and the United States each have a digital affairs minister like Audrey Tongue. I'm not sure. I think UM, whatever role is called chief technical officer, chief Digital officer, chief AI officer, that role has to be empowered in order to be impactful for any country. I think what's perhaps more important is to let private enterprises do what they do best. That is, let private capital invest in private companies to build companies, and to the extent that they're successful in getting users, in generating revenues and profits, they will get more money and then potentially be listed publicly under to success. But the role of governments is really to provide the types of infrastructure that private companies cannot do, and to have really as open as possible technical policies that let new technologies have a chance to perhaps grow faster and potentially even disrupt traditional UM technologies. I think it is that's openness and infrastructure building UH that will ensure the country moves forward. China does a reasonably good job on both UH. Infrastructure building would be examples like building a new highway for autonomous driving or even a new city the size of Chicago for autonomous driving. And as far as open policies would be the example of letting the Chinese software companies develop payment so much so that these up mobile payments have squeezed out cash as well as credit cards from almost into obsolescence. So these are the kinds of examples that has pushed China ahead to the extent possible. I think other countries can try to build similar infrastructures and have such open mindedness about technology, otherwise there would be risks of falling behind. I attended your Asia Society event in New York October one, and I have to admit to being surprised to see you in an extremely smart, well tailed three piece suit with an immaculate tie. I don't think it was many people's vision of what a tech guru would look like. Where had you just been I knew I was going to the Asia Society. I was going to greet many people who wear suits and tie. Um, but to be fair, I do wear a certain tie quite often and that's my prefer attire. Dr Lee, thank you so much for joining us, and congratulations on the book. Okay, thanks so much. Benchmark will be back next week. Until then, you can find us on the Bloomberg terminal, Bloomberg dot com, our Bloomberg app, as well as podcast destinations such as Apple Podcasts, Spotify, or wherever you listen. We'd love it if he took the time to rate and review the show so more people can find us. You can follow me on Twitter at moss underscore E. Benchmark is produced by tofa Foreheirs. The head of Bloomberg Podcasts is Francesco Leaving. Thanks for listening. See you next time. It was not the Boy

In 1 playlist(s)

  1. Trumponomics

    429 clip(s)

Trumponomics

Tariffs, crypto, deregulation, tax cuts, protectionism, are just some of the things back on the tabl 
Social links
Follow podcast
Recent clips
Browse 429 clip(s)