China and US face off in AI with Kai-Fu Lee (Sinovation) | Disrupt SF (Day 1)
10:30PM Sep 5, 2018
Okay, so there is perhaps a no category in the history of humankind that provides as much promise to us as AI. Our next guest is an expert in the field and as an investor has 13 unicorns in his portfolio, so he wins.
Please welcome to the stage from sinovation Kai fu Lee and our moderator, john Russell. Big round of applause.
Okay. Hi, everybody. And Kai Fu. Hi. Thank you so much for coming here. So the last time that we spoke together was Beijing, right? That's, you know, years ago, TechCrunch Beijing. Yeah. So you guys have been been pretty busy since then. Right. So very busy. Earlier this year. You raised your latest fund, right? Right. So in total, I think it's about 1 billion US dollars. Yeah,
we raised about 1 billion every two and a half years. So we currently manage about 2 billion we were smaller earlier.
Why is that that you that you you downsized downsize no, we
upsized. Upsized. So yeah, we started with like, $100 million fund. Now, they're about a billion dollar fund in dual currency, mostly because the entire Chinese entrepreneurial environment has been incredibly vibrant. A lot of great companies and with great companies with great entrepreneurs, the prices are higher, the new areas we could invest in China was a behind the many areas. So now we can use AI and mobile to reinvent education clinics, retail, and all of those are just full of opportunities. So you guys and I think you just mentioned that you have a dual currency fund right, so you have one vehicle is US dollars, right? And the other is obviously Chinese Yuan, right? What's the reason for that? Why do you do that depends on where the entrepreneurs want to go public. So the one that go public in US NASDAQ NYC, we take money from the US European investors, and then they exit that way through the VIP structure if they want to go public in China than they take the Chinese investors money and we exit that way.
Oh, wow. So you have to ask them early on, right? You have to choose. That means that you know a lot about which companies that you've got the on public right now where they're going to go public, right, maybe? That's right. That's right. I'm just curious. So. So in terms of what we've seen over the last year, obviously, it seems like the the IPO window for Chinese companies is open. Yes. And the US is is is quite a popular route, right? Are you seeing that in the companies that you're that you're working with? Obviously, without, without you like giving us show info unless you want to tell us but
Oh, sure. I want to tell you everything.
While most Chinese tech companies still want to see a US exit. I think ringing that bell at the NASDAQ is still a dream for many people. And the American analysts are still the best in the world. So companies that do tech and do tech well at that that aren't afraid to answer the hard questions want to go public in the US although there is one near us solution, which is a Hong Kong Stock Exchange, right. And that's fluid with the same us investment banks and investors to both our we expect to see several in both New York as well as Hong Kong.
Okay. And that sort of China's you also do deals here, right. In the US
we we do some deals here. Yes. So what are the kind of companies that you're that you're seeking out here, mostly core pure core technology companies that might be able to go into China, or have a china angle or license their technology to China, because our value add is clearly the China side, the top tech investors would likely if if they choose us, it would be because of our extensive China expertise. So robotics sensors, ai would be our core investments here, some education as well. Okay. And in terms of what sort of stage we do, you kind of aim to get in on those deals. Oh, both a series A and B, both us in China. But really, our 95% of our money is deployed in China. So yeah, us is just more of an outpost to learn, and maybe take some entrepreneurs and help them get more business in China. And I know that we spoke about this previously as a when you guys announced the fund. But you mentioned that you were keen to do more here in the US, right? What do you What are you doing here in terms of trying to build your, your profile here and doing other things? Well, I think us is clearly and Silicon Valley is the world's mecca for the most brilliant technologists. So we're looking to create relationships, some would invest in us some we get to invest in some are just partnerships. And we exchange knowledge is very important. Hill Road has a lot of knowledge about what's hot in the US. And that's helpful to us. But we also think China has become a parallel universe where a lot of the best business practices if launched in the US could be great companies. So we exchange a lot of information with our VC friends in the us about how to help American startups using the Chinese playbook. And then, of course, the reverse has always been used the Silicon Valley playbook applied to China.
Yeah, right. So in terms of AI, which is obviously is your area that you're particularly, you know, focused on, and you obviously known for that too. And in terms of the companies that are out there in the AI space, it seems like the biggest ones are in China at this point, right?
Yeah, well, because I think I've been working on AI for 38 years since my software long time since I was this know, since I was in college, sophomore, right on natural language processing, seeing speech, computer vision, and finally come to fruition that AI is applicable. But I think people assume, because the US is so strong in AI research you at us to dominate, but actually China is catching up really fast. And, and actually, I outlined this in my book, ai superpowers, which comes out this month. And it's describing China's
magical ascent in the space of AI, just in two years, coming from way behind now is actually ahead of the US in AI implementation and an AI we, we should think of it as electricity, right? Thomas Edison and also the AI deep learning inventors were American, they invented the stuff and then they generously share that. And now China as the largest marketplace with the largest amount of data is really using a to find every way to add value to traditional businesses, to internet to all kinds of spaces. And the Chinese entrepreneurial ecosystem is huge. So today, the most valuable AI companies in computer vision, speech recognition, drones are all Chinese companies, their sales are larger, their market cap is higher, the more VC invest in AI than in the US for the 8% China 43% of a world's AI investments are in China today, we have created five unicorns in AI in the past four years. And these are companies that were found that between two to four years ago, wow, this is really how fast it's been. You have to be there to, to, to see the excitement and the pace so but
it also seems I mean, that's incredible. And and the fact that there's that there's more, there's more data out there in China that you can get hold of, to us. And as application wise, there are more what areas right, where these companies can, can do things, but also borders on I mean, there's a lot there's a lot of good things that there's also some areas that I bet, you know, unsure, right? And there's there's been a lot of media coverage about some of the ways that the government is using data against, you know, some of the people that are living in China, Does that concern you in some way? I know, one of the companies that face plus, plus, right, that is in your, in your sort of stable, that they're involved in working with the government to help, you know, find people, you know, and there's, there's an there's an angle that that could be used in a, in a more in a more perhaps, sinister way. What's your, what's your take on that?
Well, we are an investor and we invest in technology companies and technology companies have lots of customers. So face plus plus is one of our 45 companies and they serve the financial sector, education sector healthcare, they're part of us Smart City, they are part of preventive maintenance of older people from falling down in the hospitals. They are also part of government procurement, to use the technology for arresting criminals and terrorists and airports. And the use of computer vision is more in the Chinese government than probably then the US government. But then there's also you know, UK and Israel and other countries that use these technologies more. Yeah, but I think just calling it sinister is probably too extreme. Well, there's the potential to be sinister, right? in China. Every or every technology has the potential to be sinister in any country. I really don't think you can point fingers at one country that way. Okay.
Well, I'm not sure I agree with you on that one. But that, okay, understand, okay. And just so one point that I really, I really want to get to is the way that you guys work as a as a because you have a very different take on tech. Right? So inside the fund, you actually have your own Institute, right? That's what you guys call it. So. So can you explain to the audience like what that actually is? Yeah,
I think in China, the execution is so fast that we love to invest in all these companies in AI. But sometimes we see some opportunities and no one's working on and it's such a pity. So we decided we would hire engineers, and then we would look for the best opportunities to apply them. I think in these cases, you can hold us accountable for picking applications that are clearly for human good. So the types of applications that are AI companies have looked at, including manufacturing to reduce costs, including insurance, to give better service to people and reduce fraudulent claims, retail to create better sales forecasts. And we're now looking at healthcare cancer. And we're also looking at education as a way to take a lot of teachers do that are repetitive and routine, like grading homework, taking attendance that could be done automatically. But
I mean, in terms of the team that you have internally, right, so we're sort of sizes that teams I think it's quite normal, the VC firms have take people in, in a, in a team, right to work with their companies that you guys have taken a bit further, right. So how many people do you have? We have about 200, 200
people working on tech in a VC firm. That's quite interesting. So so what what are they actually doing? And is it like, sinister stuff? Or I told you, I know, you say healthcare, fighting
they say that building these these these products, yeah,
totally, uh, what they're Think of this as the Google 20% time. So we give them ideas of what areas of business that we think will tip and then they build algorithms. And then when they show something, and then we talked to a lot of traditional industries that say, well, we could use that in our hospital, we could use that in our retail store, and then we'll do some contracts. And when we start to see something, go from a solution to a product, we can then put money behind it, and then it becomes a bona fide
cry. And you and you just did that
right. With one? One spinoff? Yeah. So which is what's the company What are they the company's called AI innovation mixing? Who are they I and innovation and they do AI solutions for retail insurance and manufacturing.
Okay, and so that was a company that was started internally and then you've spun out and use and you've given a funding do say work with with other VCs, the person on Ian is usually another VC would set the price so we can avoid conflict of interest, okay. And then otherwise, you do so the team of 200 people, I know that I mean, that's so many people working in a VC firm is kind of crazy. But they also just other stuff, right? This is
China scale. Like take may try. And as an example, when people call them the Groupon Open Table, Yelp, see if I not right, and they have 600,000 people doing deliveries, and, and they drive down the price of delivery to like 70 cents per order. And the Chinese way is to really take a large number of people and use them to operationalize and perfect and get the process, right. So we have 200 engineers, when we spin off when are left with, you know, 30, then we grow to 200 again, then we do another spin off.
So you're saying to people, is not that many people to have basically not not at all. So what was
the target? Well, I wasn't as many as we can hire. And they have to be hired with leaders, they can all be new grads, but a lot of them are new grads
and d i mean is are you seeing this kind of model potentially happening here. I don't
think that's I don't think so. I think the Chinese model are truly we look at may try and and how it's different from the Yelp and Groupon look at Didi is getting very different from Uber, you look at Tencent is quite different from Facebook. So like in my book AI superpowers, I also describe how China has bifurcated to an entirely different approach that is doing tech the heavy way, maximizing efficiency, producing the maximum gain and use of highly competitive, well, I call gladiatorial environment, which pushes everybody to be the best they can be, and end up with completely different business models than the US if you look at China. Now, there are also new innovations coming out that don't exist here. If you look at and financial or $200 billion company, Pinto dollar, $25 billion company, the companies we invested in, like VIP kid and mobile, they don't exist in the US they were China first. So I think TechCrunch should encourage everybody to really study how the Chinese entrepreneurs are coming up with these different models. So we should also mention that you have a q amp a and and a book signing as well, right? So for three or 4pm, I think they told me backstage, I can't remember. But it's on the agenda. So you'll be answering more of these questions about the differences between I'll be happy to take any questions from entrepreneurs, VCs media during a q&a session, and you're signing a book that doesn't exist at the moment. That's some kind of AI craziness. We have a AI way to sign the book, the book doesn't exist calm and find out how we do that. Well, that's an that's an amazing pitch.
Okay. And you gave me the perfect segue into the final piece of the of the talk, right, which is about US companies in China. So obviously, I was the Google a company that, you know, pretty well, you were the head of China for for Google. It's, it's some time ago, right? It's so nice. Almost 10 years ago, they they left I left in nine years ago. Yeah. But what do you think of these apparent plans that Google has to launch new products and the Chinese market? Why are they doing that?
Well, first, I'm a huge fan of Google. I love the products. I love the culture. And I really will remember very fondly the four years that I worked there, and we actually became a company that thought could prove American companies could succeed in China, how did I get a deal when I left, we had a much higher market share than when I joined, we went from 9% to 24% making progress, and their revenue was approaching a billion. So it was pretty good. But afterwards, they made certain decisions to pull out when I understand their decisions. But regarding their possible re entry, I think re entry is always difficult, because either you're acknowledging what you did earlier was not right. or you're not. And either way, there's some trickiness to it. But I think the bigger issue really is can an American multinational succeed in China, now that China has bifurcated into this parallel universe, the entire dynamics is are different people don't aren't looking for another search engine or an app store, new companies are emerging are addressing previously unknown customer needs, innovations are coming out the new grads generally prefer to work for Chinese companies. And then lastly, the heads of multinationals are really just professional managers. And, and if they were to compete against local entrepreneurs who are gladiators in this Coliseum, I don't think the American companies will have a high chance of succeeding in this environment. Although I wish Google the best, I think they have as good a product is anybody the they should have a higher chance than any other company. But it's going to be tough. But I mean, it's not only Google, right? Because there's it there's these stories about Facebook tried to register business,
you know, to, to operate that too. I mean, it seems like they they either they don't read the news, they don't understand that they're gonna have to have these armies of people like checking content. And it's not it's not an easy thing, right, have to have the the
very unique taken us in China where you can't publish certain things on I don't really know what happened with Facebook situation. I know, both Facebook and Google want to find a way in because China is the world's largest market. And I think the fundamental issue isn't a government issue. It's a competitiveness issue, right? And I'm really not saying American companies are not good enough to win in China. If you take any Chinese company you know our Baba Tencent total Didi the best and say you go build an American subsidiary when I think that's equally unlikely to succeed. So you see, obviously back when you with Google, you thought that there was a chance that you could build a product right that could compete with with Baidu on search, or on maps. Right. But But now that that's that's not really possibly think. I think building a competitive products, if you just look at the product feature is obviously possible. I think Google has a phenomenal the world's best search ranking algorithm. But it's also about branding, user expectations, you know, search isn't the hottest thing anymore. So who's building the next app? And can you fight with the equal tenacity and methods of local competitors? Are you incentivize your team the same way that a Chinese company is or is there a glass ceiling for the heads of the group's so there are all these complex issues that I think any American company would have a hard time in China now, Apple being the single exception. And I think that's because they are mostly a hardware product and and the product is just become a fashion symbol. So that's different. Great.
Unfortunately, we're out of time. And I could ask you so many more questions. But obviously you're going to be here for that that q amp a and the and the and the book signing so we can hear more from you then. Thank you so much. Thank you for coming. Thanks.