On the topic of European artificial intelligence, we met up with AI expert and former exec at Apple, Microsoft and Google Kai-Fu Lee, to get his views on why Europe lags behind China and the US.
Q: While you’re outspoken on the relative potential of the US and China to make huge progress in artificial intelligence, you haven’t publicly shared your views on where Europe stands. What is your perspective on Europe’s potential?
I left out Europe because I didn’t think there was a good chance for it to take even a so-called ‘bronze medal’ in this AI competition. European artificial intelligence is losing the race.
Aside from a strong research tradition, Europe has none of the success factors of the US or China. If you list the top five reasons why Silicon Valley is leading the way, and the top five reasons for China, Europe has one half of one reason why Silicon Valley is ahead.
Plus, Europe has no VC-entrepreneur ecosystem. The entrepreneurs in Europe are nowhere near as innovative as the American ones nor as tenacious as the Chinese ones. Europe’s entrepreneurs lack the experience of dealing with these types of software and artificial intelligence problems.
Europe has had great companies in hardware and telecoms, and a few very good enterprise companies, but has never really built any successful consumer internet companies, social media companies, or huge mobile application companies. So it is missing three links already - and therefore there is no experience in Europe of dealing with large data or large scale AI companies.
The only exception is Google DeepMind - but that is [now] an American company. It was wholly acquired by Alphabet.
Q: What else is European artificial intelligence missing, compared to the US and China?
The US is ahead because it is simultaneously strong in research and implementation, has the world’s best VC-entrepreneur ecosystem and has a huge head start with the likes of Google, Facebook and Amazon.
Already in America there has been four or five decades of interdisciplinary collaboration: semiconductor people helping PC people helping software people helping internet people helping social media people helping mobile people. And now all the resources are helping AI people. This whole linkage and heritage continues in Silicon Valley
China is strong because of the huge market size of a homogenous country where everyone speaks the same language, the same culture, and there is a huge population. The fact that data so easily aggregates creates business models that are profitable because if it works in one city it can work in a hundred big cities; if it works with one demographic, then it applies to all demographics. Also, the market size is an even larger advantage in AI where so much data is needed; the more data, the better it works. In addition, China has developed an entrepreneurial ecosystem with VCs and entrepreneurs really helping each other grow iteratively, much faster than in the US.
In China they had the ‘good fortune’ of basically being behind in so many things so they could leapfrog into a mobile world, go straight to mobile payment skipping credit cards, and now leapfrogging into online retail skipping the shopping malls, and so on. An industry’s legacy is hard to dump because it creates bad habits that are hard to change and baggage that is hard to leave behind.
China has come up with its own business model for building incredible companies that cannot be copied. The Chinese business models are actually more impregnable than the US ones; the tech giants that the Chinese have come up with are much harder to dislodge because of the ‘winner takes all approach’ in China and the pursuit of execution excellence, the work ethic and the speed of execution. All these combine into a whole new way of generating tech company giants.
Finally, China has the government advantage: building infrastructure, providing investments, and setting the countrywide direction.
Q: How do you respond to the suggestion that Europe’s regulatory initiatives - like data protection, action against tech monopolies, and innovation sandboxes - will help Europe's artificial intelligence provisions?
It is an extremely academic argument. For Europe artificial intelligence to truly compete we would have to roll everything back - go to Europe and delete everybody’s phones and internet and start from scratch. But we can’t undo history or people’s habits.
People use Google, Amazon and Facebook in Europe and they like it, even if they dislike elements of it.
It is extremely powerful for Google to know exactly who types what and and says what and searches what. I don’t think it is true that by anonymising that data and creating larger aggregations it would be more powerful. No one can prove this either way, but the fact that Google and Facebook can target you is what allows them to make the most money. So if you remove that ability - how would a new company compete? If you keep Facebook and Google in Europe but then promote a third set of companies with a new approach, they wouldn’t have any of the data.
Google and Facebook essentially have money-printing machines, so I don’t believe new companies under Europe’s regulation will monetise better. I do believe it protects people better but I don’t believe it monetises better. So it feels to me like the only way for Europe to succeed is to kick out Google, Facebook and Amazon.
“It almost feels like the only way for Europe to succeed is to kick out Google, Facebook and Amazon.”
But the idea that Europe can get ahead by limiting the power of these tech giants is naive. It appears unimplementable both because Europe is so far behind and people won’t give up what they love: the benefits of Google. It is a capitalistic world.
Q: How would you advise startups in Europe hoping to forge a path alongside global tech giants from the US and China?
If you want to experiment, find some truly underdeveloped country where there is nothing and try your idea there. Or use the infrastructure that people love but build something totally different. Don’t build another search engine.
Google didn’t beat Microsoft by building a better operating system. But now Microsoft has just replaced the core of its Edge browser with the Google browser Chromium. Microsoft has lost the browser war to Google - this would have been unbelievable two years ago. And it didn’t happen because Google said ‘we will build better technology to beat Microsoft’s dominant browser and its monopoly’. Instead, Google went through a separate process of building a search engine, then monetising the ads, then it caught mobile, it bought Android, then Android became the default standard, then Google figured out how to monetise maps, search and its Android store. Finally, Google’s browser also worked better.
The entire shift of Microsoft’s power to Google’s is very revealing; that is how you beat one giant with another giant. Of course the game is not over and now Microsoft is shifting to the cloud. But this is where I don’t really buy this story that monopolists can never be challenged.
I admire the idealism in Europe and I think the efforts to protect people’s privacy is great. But my advice would be to build a brand new application and focus on developing that. It can be done, you just need to use your idealism to find the next paradigm shift and apply your innovation there.
Q: How does the attitude to data privacy in China affect innovations in the tech sector?
Chinese companies are not a lot different from US companies; they aggregate data, create closed loop systems and use AI to maximise their profit in the same ways as Google, Facebook and Amazon do.
The Chinese government has its own data protection laws, which are not so much protecting individual data privacy as restricting companies selling data to each other. The focus is on the Facebook-Cambridge Analytica case of inappropriate data transfer between companies.
Chinese technical laws tend to be very utilitarian. They tend to let things be implemented and tried until something goes wrong. When the government sees things go wrong, it moves quickly and corrects them.
The government saw a significant amount of fraud based on people buying other people’s data and then sending fake ads and sending fake orders on credit cards. So that was the big problem that it wanted to stop. Incidentally, it also would have stopped the Cambridge Analytica type of case, but that wasn’t the motivating reason.
There haven’t been many cases in China where individual people felt their privacies were inappropriately exploited by the tech giants. All these tech companies - American or Chinese - have to protect their reputation by using the data judiciously so as to not create mass unhappiness, which Facebook seems to have caused now. But Google has done a pretty good job and Chinese companies have done a reasonable job, so there haven’t been so many complaints or major data leaks.
I think the Chinese government will continue to observe consumers and whether they are being hurt or not. If they are not being hurt then I don’t think new laws will come out. In that sense, China and America are more similar to each other than to the Europeans.
There are those that disagree with Kai Fu Lee. Speaking at the European Commission’s policy forum on artificial intelligence, Futurium, this is how experts on European artificial intelligence responded.