Kai-Fu Lee Says U.S. Tech Companies Aren’t Ready for Chinese AI

(Bloomberg Businessweek) -- The chairman and chief executive officer of Sinovation Ventures says America's technology industry can't afford to underestimate the artificial intelligence companies he's funding in China. Lee, a Taiwan native, has spent most of his professional life in China, but he lived for a quarter century in the U.S. before becoming a venture capitalist, and he's worked for Apple, Microsoft, and Google on one side of the Pacific or the other. Lee spoke with Bloomberg Businessweek shortly before the official release of his book AI Superpowers: China, Silicon Valley, and the New World Order. His comments have been condensed and edited.

 
Why should Google and America's other AI leaders be worried about competition from China?

A lot of people in Silicon Valley assume if you're a copycat, you're doomed for life. You're never going to become a great innovator. China has proven that wrong. Think about a very smart, hardworking bunch of people with strong leadership and business skills who just haven't been in an environment as lucky as Silicon Valley. Being a copycat first actually turned out to be the best training ever.

Think of it as a pyramid. For all these copycats entering the base of the pyramid, most will not make it because they never learn to build a good product. But if you copy at the first stage, then learn from the experience and make your next startup better or your current product better, that's a very formidable solution that Silicon Valley never thought possible.
 
Is there a company that stands out?

Wang Xing's Meituan Dianping is one example. Here's someone who copied Facebook, Twitter, and Groupon. But with Meituan, he figured out how to use technology and his business skill to turn it into an innovative, profitable enterprise. He won the loyalty of merchants by building very good software. He also took control of the user experience, so you can deliver food in a reasonable period of time—20 or 30 minutes—and the food will be hot when it arrives. You can see 500, maybe 5,000 restaurants, too many to browse, and its software maximizes deliveries, so the delivery costs you well under $1 per order.
 
What's stopping an American quasi counterpart such as Uber or Seamless from doing the same thing?

It's hard for any country to learn new tricks. The American way has been light on logistics, more into extracting rents from their software platforms and letting someone else do all the ground-level stuff. Someone such as Wang Xing is willing to take on the really hard work of hiring tens of thousands of delivery people, then managing them and making sure they're efficient and on time. This is a messy job.
 
Maybe especially messy in China. Obviously the government hasn't been shy about trying to make things easier for tech companies. What are the most important things it's done in that regard?

The biggest role the government plays in China is through what I call a techno-utilitarian policy, which means letting technology companies have a try with as-yet-unproven technologies: Let them try it in some limited way and then see what happens, and if it works, let them push the limits and go into even gray areas—like letting Alibaba and Tencent go into payments. Their Alipay and WeChat services took the country by storm. There's no longer any room for credit card companies. That spirit matches the scrappy, fast execution of entrepreneurs in China.

The other thing the government did that's been helpful is the infrastructure investments. In China, the 3G, 4G coverage is better than in the U.S. Sure, in U.S. cities, they're great. But if you go into country areas, they're not as good as in China.
 
What sorts of knock-on effects are those kinds of investments having in the war for AI supremacy between U.S. and Chinese companies?

My position is that China has more or less caught up in AI. The U.S. has certain advantages, such as the best universities and, historically at least, immigrants who want to come from all around the world. China has other advantages, including more data, more supportive government policies, and now very good entrepreneurs.
 
What about the creepiness inherent in the state slurping up that much data and monitoring its citizens so aggressively? It sounds like the road to dystopian science fiction.

The real issue facing mankind is not whether the U.S. is ahead or China is ahead. The real issue is that AI will create so much wealth and so many challenges—the replacement of jobs, privacy, security—that we really should work together on potential solutions. If we don't have good solutions, any one of these issues can lead to disasters larger in scale than any other disasters in human history. We need to start preparing now.

To contact the editor responsible for this story: Jeff Muskus at jmuskus@bloomberg.net, Cristina Lindblad

©2018 Bloomberg L.P.