China fights US for supremacy in artificial intelligence

At the end of March, I attended for the ninth time the China Development Forum (CDF). The visit stimulated my recent observations on China's economy and politics. But what made the CDF more valuable was that I attended a meeting with Kai-Fu Lee, former president of Google China and now a leading venture capitalist in the Chinese technology field.

Lee gave me a copy of his new book, AI Superpowers: China, Silicon Valley and the New World Order (The AI ​​Superpowers: China, Silicon Valley and the New World Order). According to him, for the first time since the industrial revolution, China will be at the forefront of a huge economic transformation: the revolution of artificial intelligence (AI).

He begins his book by talking about the "Sputnik Moment" in China, in which DeepMind of Google's AlphaGo program beat Ke Jie, the most experienced player of the Chinese Go game. This demonstrated the ability of modern AI. But, implicitly, the book of. Lee anticipates another similar moment in which the United States will know that you are no longer a leader in the global application of AI. The original moment of Sputnik took place when the Soviet Union put into orbit the first satellite in 1957. This led to the space race of the 1960s, to which the United States subscribed. won correctly. What will the current "race" lead to?

Lee does not guarantee that China will lead the fundamental innovation in this area. But this may not be important because the great intellectual progress has already taken place. What matters most is implementation, not innovation. And, according to him, in this area, China has many advantages.

First, the work of leading researchers in artificial intelligence is available online. After all, the Internet is an engine of choice for spreading intellectual advances, including those related to AI.

Secondly, China's hyper-competitive and entrepreneurial economy lives according to the famous slogan of Facebook founder Mark Zuckerberg, "Let's act quickly and break things". Lee describes a world of relentless commercial activity and incessant imitations, which already allowed Chinese companies to defeat the major Western competitors in their local market. The incessant "trial and error" of the Chinese economic model are, according to him, perfectly adequate to deploy the fruits of AI throughout the economy. This could, for example, work much better for the introduction of autonomous vehicles than the focus on Western security. The multitude of people in China can be ineffective, but it is effective. That's what matters.

Third, China's densely populated urban areas have created a strong demand for delivery services and other similar services. "US startups prefer to stick to what they know: create clean digital platforms that facilitate the exchange of information," Lee said. But Chinese companies are involved in the real world. They integrate the worlds online and offline.

Fourth, China's backlog has allowed companies to overcome existing services. This is why China has been able to switch to universal digital payment systems, while Western companies continue to use outdated technology.

Fifth, China has a scale. The country has more Internet users than the United States. and Europe together. If the data is actually the fuel of the AI ​​revolution, China simply has more than anyone else.

Sixth, China has a solidary government. Lee quoted Prime Minister Li Keqiang's 2014 address at the World Economic Forum's "Summer Davos," calling for "a tremendous entrepreneurial spirit and innovation." In his report "Deciphering AI's Dream of China," Jeffrey Ding, of Oxford University, pointed to the State Council's national strategy for the development of the country. 39; IA. The Chinese government has ambitious goals and is willing to take risks to achieve them. One of the things that China can do more easily than any other place is to build complementary infrastructure.

Finally, Lee wrote, Chinese people are much more relaxed about privacy than Westerners. It can be argued that Chinese leaders see no justification for privacy (with the exception of their own).

So where is this supposed "race" between the United States today? and China? Lee mentioned four facets of AI. Internet AI: the AI ​​that follows what you do on the Internet; Business AI: AI that allows businesses to better leverage their data; the AI ​​of perception: the IA who contemplates the world around it; and autonomous AI: the AI ​​that interacts with us in the real world. Currently, he thinks that China is on a par with the United States. in the first, extremely in the second, a little later in the third and, again, extremely in the fourth. But Lee thinks that in five years, China could be a little ahead in the first, less late in the second, well ahead in the third and even in the last. There are, in his opinion, other competitors.

Ding badyzes the drivers in a different way. It differentiates between hardware, data, research and the commercial ecosystem. China is far behind the United States in terms of semiconductor production, at the forefront of the number of potential users and has about half the number of experts in artificial intelligence and a little less than the half of the companies in artificial intelligence. In total, China's potential is about half that of the United States. However, Ding badyzes AI in general, while Lee focuses on commercial applications.

The historical experience suggests that rents created by an advantage of a significant technology are valuable, even though they are often not permanent. Therefore, which country will be leading the application of AI is really important. But the economic and social impact of AI is a major problem and concerns each country.

As Lee points out, AI's advancements offer advantages. This applies not only to personal convenience, but also to the improvement of medical diagnoses, the adaptation of education to each student, the management of transportation systems and the Energy, fairness of the courts, etc. successively

However, the AI ​​also threatens to cause serious disruption, especially in the labor markets. Many of the tasks (or tasks) that AI could do today are performed by relatively educated people. It seems reasonable to fear that AI accelerates the emptying of the middle part of the distribution of profits, perhaps even the upper middle bracket, while increasing concentrations of private wealth and power at the top.

However, the most important consequence is perhaps the intensity of the influence and vigilance made possible by mobile devices and sensors monitored by the AI. George Orwell's big brother (or many big commercial brothers) could watch us all the time. Such a perfect follow-up might be of interest to the Chinese state. For me it is horrible and I hope it is horrible for other billions of people.

Lee insists that artificial intelligence is not the same as artificial intelligence: the true superbrain is still very far away. Nevertheless, the challenges created by this AI are huge. We will not stop it. But it is likely that we conclude that we have created a monster.