Kai-fu Lee says the AI revolution will be fast and wrenching. But in 50 years, we'll be glad it came

Enough words have been spilled about the impact of artificial intelligence that a reasonably competent bot could likely write a sharp take on our machine-driven future. But if you want to pick one human to get your opinions from, you can’t go wrong with Kai-fu Lee.

Kai-fu is touring the country to talk about his book, AI Superpowers: China Silicon Valley and the New World Order and he's got some pretty dire warnings:

• Within the next 15 years, about 50% of our jobs will be handled by AI, sending us into the “age of confusion;”
• The AI revolution will rival, if not top, the industrial revolution, in terms of impact — and will come faster;
• Adjusting to the AI age will require "substantial changes to our economy but also a shift in culture and values.”
• Oh, and the chance of the US leading this revolution are slim.

Sounds great.

Kai-fu isn't making a call on whether this is good or bad; the future, as he sees it, is pretty much baked. As he says in his book, "the difficult but abstract work of AI research has been done." Now it's just a matter of watching to see who executes it best.

Kai-fu has been both sides of the equation. He began creating AI products in the mid-1980s while teaching at Carnegie Mellon, then took his skills to the Silicon Valley giants, working his way through Apple, Microsoft and Google, where in his final role he was president of Google China. He left in 2009 to start Sinovation Ventures, a China-based VC that has invested $2 billion in hundreds of AI-focused Chinese companies that are scaling up this industry-level disruption. A battle with cancer in 2013 forced him to slow down and to confront some of his core beliefs, but not his belief about where the future is headed.

In a recent visit to LinkedIn's NYC offices, we talked about everything from the economics realm to the emotional. Kai-fu explained why China is so poised to be the dominant AI country — data is the lifeblood of AI and China is the Saudi Arabia of data — and what the imminent revolution means for job seekers, policy makers and, in general for humans.

Here’s an edited transcript:

DR: You lay out four key tenets for leading in AI: Data, tenacious entrepreneurs, amazing AI engineers, a supportive government policy. And in each of these cases, you say the balance tips towards China.

KL: Yes.

DR: You're spending a couple months in the U.S. Do you feel even more certain of China being the AI superpower, or do you think the U.S. has a chance?

KL: The U.S. and China will co-lead the world in AI. And those four tenets start to tip things towards China, but that's for implementation. I think U.S. is still way ahead in research and technology, and also some more advanced uses of AI, such as autonomous vehicles, robotics. So, I think the two countries are comparable at this stage. But assuming no major, big discoveries are made, Chinese power, in implementation and just the magnitude of data, will likely give China the edge.

DR: I think when a lot of people hear about AI and what's coming, the closest analogy they think about is the rise of the Internet. In your book, you point out that AI tends towards monopolies — that the people that have the data are the ones who control the applications on that data. But you're a VC. You're funding all kinds of small companies who are starting up, so there must be some hope for these new players.

KL: The current large, powerful AI companies are all Internet companies. They don't necessarily have very strong positions in transportation, automotive, hospitality, healthcare, retail, banking, finance, etc. So, we are largely funding entrepreneurs who are going into non-Internet spaces with AI. Those are spaces where it's close to being a level playing field.

DR: This is a winner-take-all in each of those spaces?

KL: AI tends to be winner take all. Also, China tends to be winner take all. You know, in the U.S., let's say the food delivery category, there's Yelp, and Groupon, and GrubHub, and OpenTable. In China, if these four companies all existed, doing what they did? They will eventually battle, fight each other, and only one would survive. It's a little difference between China and U.S.. So I think AI tends to form powerful monopolies, and China tends to have strong competition, which also tends to end up with either consolidation or one company beating the other, becoming the giant or the monopoly.

DR: You make a big deal in the book about the cultural differences between China and the U. S. — the tenacity of the entrepreneurs in China. Is that a way of doing business in China, or is it just a stage of life in China?

KL: Well, there has certainly been some tenacious American companies. I don't want to overly stereotype, and I'm sure there are some mild Chinese companies. But by and large, China's tenacity comes from a number of factors.

One, from having been poor for many decades, or even centuries. Some families have high expectations, and I think young people have a lot of expectations, and they're hungry for success. I think the work ethic is very, very strong. I think the environment, because there has been — not much anymore — a copycat environment, given everyone can potentially copy everyone else. That causes companies to want to be very, very tenacious, and build up very high walls, so other companies can't come in.

So, you know, Meituan builds a 600,000 delivering army. DiDi is buying gas stations and car repair shops. All of those are trying to create a vertical protection so that competitors can't come in. If you want to compete against Meituan's delivering army, well, first get a few billion dollars. If you want to compete against DiDi, well, they're not gonna let their driver carry a second app like Uber allows Lyft.

So, it's just tougher competition. Will that change? I think some culture elements will change. I think there will be less tenacity as there's a larger middle class, but I think some of the ways of doing business, and some of the playbooks of success, perhaps, will not change.

DR: Eric Schmidt recently talked about a bifurcation of the Internet, of a China-led Internet and a U.S.-led Internet. Do you buy into that?

KL: To the extent that he means Chinese software, and the usage of software-driven countries — perhaps China, Southeast Asia, Middle East Africa — I would agree. To the extent that he means different regulations, I'm not sure I would agree with that. But I'm not sure.

DR: Your book paints a pretty bleak portrait of what life is like once AI reaches its full potential. Forty to 50% of all jobs are able to be done by robots. Those jobs aren't coming back?

KL: Those jobs aren't coming back, but we’ve got to create new ones.

DR: At the scale that we need?

KL: Well, that's the question. I think I gave my ideas, and other people have other ideas. Hopefully an aggregate is enough. We probably don't have to cover 100%. Maybe some of the new jobs are part-time jobs, and maybe some of them are covered by social welfare. Maybe there's some degree of early retirement. So, it doesn't have to be one-to-one, but it can't be one-to-zero.

As I mentioned in the book, the jobs that require human to human skill — compassion, empathy, trust — are difficult for AI and robots to do, and that would suggest jobs that are from nannies and nurses and elderly care, to teachers, and maybe the future of doctors, as well as tourists, concierge, and so on. It's a pretty large group. Some of them are growing. Elderly care surely is growing, because the jobs aren't being filled now. There are like a million jobs open. Also, as we live longer, there will be greater need for elderly care. People over 80 need five times as much care as those between 60 and 80. So, the need will increase in some of these categories. Also, as society has more money, many of the new wealthy will want to have service. Service that comes with entertainment, for travel, and Michelin restaurants, and luxury brands. Those will create jobs there, as well.

Finally, there will be the AI jobs, right? The jobs we can only imagine. Data scientists, AI engineers, Python engineers. But I'm sure there will be 100 other jobs over time.

But I don't think in the next five to eight years, they'll all appear. Give us maybe 40, 50 years, I think AI will end up creating a whole new set of jobs. Just like most of humanity used to be in agriculture, and suddenly it shifted to manufacturing. I think just give enough time. New jobs will come out.

DR: Your daughter’s in college right now, right? What advice do you give her for navigating this coming world?

KL: Well, she's already studying art photography. My older daughter was studying fashion design. I think what's important about that is, first, that they're pursuing what they love. I think that's what the future will hold, is that as AI does more of the routine jobs, we should get to do the things that we love. Secondly, they're doing creative job, which is arguably not so easily replicable by AI. Thirdly, both of them are learning that communication with other people is very important.

DR: In the book you talk about your battle with cancer and how that changed how you think about your career, the time you spend at work and your own tenacity. Have you been able to hold on to that? How do you not make sure that the old Kai-Fu doesn't come creeping back and take over the new Kai-Fu?

KL: I used to be a workaholic, and I think facing cancer made me see work is only a part of my life. I can't make it the whole thing. That was what gave me the epiphany, that when AI is about to take away the routine jobs, there are other people who are a workaholic, like I once was, and that they would face the same challenge if the jobs were taken. And even if financially they were taken care of, they will not be satisfied, because they believe work equates to life. So while I, through facing death, realized that there's a lot more than work. There is the people I love, there are the things that I love to do.

And I have been rebalancing my life since my cancer going into regression. I've been spending higher quality time with my family, prioritizing time with them over the time at work.

And I think, really, in 50 years, looking back, all humans who are still alive then, we'll realize that this whole round of AI-displacing jobs actually ended up being a really positive thing for humanity, because it woke all of us up to realize, repetitively doing the same job over and over again is not why we're on this earth. And perhaps we've been so foolish, after thousands of years of evolution, that our maker decided to throw AI at us and say, "Okay, you can't wake up by yourself. I'm gonna take away all those routine jobs. Now you've got more time. Now you've got more money." I mean, the whole humanity has more money. "Now you have more choice and more freedom. Will you please do the things that you love? Love the people and spend time with them, and spend some time thinking about why humans exist." Because it's not about work.