Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
← SearchGeorgetown2022Affirmative
Opening Word document…

No Chinese AI arms race---it’s not zero sum---significant academic coop and opportunities for growth

Kai-Fu Lee 19, Transcript of Interview by Stephen Orlins, President of the National Committee on U.S.-China Relations, with Kai-Fu Lee, Ph.D. in Computer Science from Carnegie Mellon, Fellow at the Institute of Electrical and Electronics Engineers, “Kai-Fu Lee on the Future of AI in the United States and China”, 2/6/2019, https://www.ncuscr.org/media/podcast/uschinainsights/kai-fu-lee-future-ai-united-states-china

Kai-Fu Lee: Well, I've been an AI researcher for a long time, and I have worked in the U.S. and China. And I'm seeing a lot of excitement and a lot of challenges ahead of us. I've done AI as a researcher, a product executive, and an investor now. And I think AI is taking off faster than we think. China is becoming an AI superpower. With the U.S. and China, there is a lot of complementarity, and we face a lot of similar challenges. And there are a lot of approaches that could be shared. So I thought I had that unique perspective to write a book that describes that.

Orlins: Is this something which the U.S. and China are going to cooperate on in the future, or compete on?

Lee: I know everyone thinks about this as competition, but it really isn't. The U.S. is stronger in research; China is stronger in implementation. So they could learn, at least from each other. And also, I think China is building Chinese products for Chinese people, Chinese VCs funding Chinese companies doing Chinese products for Chinese people, American VCs funding American companies building products for the developed countries. The two don't collide, so it's not a zero-sum game, not as though the growth of one side would shrink the other. So given they're in parallel, theoretically, this should be a wonderful collaborative opportunity.

Orlins: You focus a lot on, kind of Alipay or WeChat Pay. How does that change the landscape for AI in China versus the United States, where we still are heavily reliant on a credit card system that seems quite old and outdated when you're in China?

Lee: It's a very unique opportunity for China to leapfrog. The U.S. had the world's most leading transaction tool, which was the credit card, but it became outdated. It charges too much, it's inconvenient, it's only customer to merchant. In China, with WeChat and Alipay, anybody can pay anybody. We can do micro payments and there are no fees, almost no fees, so the 2% extra charge is almost like a tax on the American economy charged by the credit cards. But with respect to AI, now everybody, including WeChat and Alibaba, but also, if you build an app, you know who your customers are, and what they paid and why. And if you're a retail shop and people scanned you, you know who pays you and why. There are even beggars in the street holding up a sign, "Scan me and give me money," and they know who paid them and why. So I think this is creating a huge amount of data that will help AI.

Orlins: How will it be used?

Lee: Well, each merchant will get data with respect to his or her business, and it may or may not include the customers' name. You may have to go to extra steps to solicit user consent, but in the end, if you're able to get user consent, you will know who came to your store, what they bought, on which day, and from which, you can make better sales projections, product placements...

Orlins: But doesn't a credit card receipt do the same?

Lee: A credit card receipt does not give the same organized information. Nor does it give a channel to communicate. So if you came to my store and bought this book, and I got you to sign up for my WeChat accounts, I can send messages to you. I can say, "We have a sale today."

Orlins: So it's the merchant, as opposed to the credit card company that has that data?

Lee: That's right. The merchant has it. Obviously, Tencent has it, and both can use it to datamine and make better decisions that help the companies make more money.

Orlins: Are Tencent and Ali going to be the focus of AI development in the future in China?

Lee: They are the two strongest right now. In the future, I think there are a lot of opportunities, because there are many other spaces that are not contiguous to this space, the payment space. Tencent is very strong with social gaming and now, payment. Alibaba is very strong with commerce and now, payment. But there will be health care, medical, there will be manufacturing, there will be transportation. So I think there are many other areas where new companies still could emerge with the data and the AI and the monetization.

Orlins: Who will be the leaders in the U.S? What companies will lead that in the U.S?

Lee: Well, Google currently leads in the diversity of their products, the amount of data they collect, and also the AI expertise they have accumulated. After that, there is Facebook, Microsoft, Amazon. So it's a similar situation, except that in China, I think the data collection goes a little deeper, because the Chinese people use the phone more aggressively for more services.

Orlins: How can we use AI to improve the U.S.-China relationship?

Lee: Well, the academics in the U.S. and China are wonderful friends. If you go to NIPS (conference on Neural information Processing Systems) or one of the AI conferences, you will see that the connections are good. The sharing is very open, so that could be an inspiration to the parts that may not be focused on it.

And I think the National Committee could publish more information that the U.S. and China are not competing in AI in a commercial sense. And also, that they each have something to offer. And that's why I wrote this book: to let people know that, while the media, and some politicians, have focused on U.S.-China competition— even using terms like "cold war" and describing this as a zero-sum game, but, in fact, both countries are doing great. There's a lot of sharing, there's very little competition academically or commercially. So I don't think we should continue to represent this as a war.

Sign in to recut, update, or save this card.