Xu Feiyu Photo: Courtesy of Lenovo

Editor's Note:

There seems to have been a change in how people are talking about Chinese science over the past 70 years. Slowly but steadily, artificial intelligence (AI) is changing everything about our lives. What are the strengths and weaknesses of China's AI ecosystem? Will the emerging AI industry be immune to geopolitical games? To find answers to these questions, Global Times reporter Hu Weijia (GT) talked to Xu Feiyu (Xu), vice president of Lenovo Group and head of AI Lab, in an exclusive interview in San Francisco.

GT: You are one of several Western-trained, top Chinese scientists who have returned to China to work in the industry or academia. What is the reason behind your decision to return to China?

Xu: I studied for a master's and PhD in Germany, and worked at the German Research Center for Artificial Intelligence (DFKI) for nearly 20 years. Although I had an excellent position as principal researcher of DFKI, I had always wanted to contribute to China's AI development. 

In 2016, several Chinese high-tech enterprises asked me if I could help them by leading their team in AI. Despite a few other invitations from top enterprises in the Western world, I realized the time had come to return to China.

GT: What are the differences between the AI environments in China and Germany?

Xu: DFKI is a non-profit research organization. Most research at DFKI can only use public data, while enterprise data cannot be easily obtained. However, enterprise data is crucial for the commercialization of AI technology. Some German scientists at DFKI envy me because now I enjoy free access to all historical data from Lenovo.

If I take a taxi in Berlin, I will probably forget to pay. Why? People in China so accustomed to mobile payments, with no cash. This can be seen as the epitome of today's situation in China. The whole country is open to embracing new technology. It's a really good thing for scientists. In contrast, the West always takes a wait-and-see attitude when it comes to adopting new concepts.

China has a much more urgent need to develop new technology than Western society because the country has a large population and relatively few resources. 

My average commuting time from home to the office is about one hour. If we make progress on smart cities with AI technology, commuting time can be reduced to ensure a better quality of life.

GT: How do you evaluate the development of AI technologies in China, the US and Europe in past decades? 

Xu: China is one of the best-performing players in the world in AI commercialization, venture capital and government-led promotion. The country is likely to take the lead in the global race for the commercialization of AI technology. However, China is a latecomer among big powers in terms of AI technologies, so more efforts are still needed in basic research, frontier research and original innovation. 

Europe is good at basic research for AI technology, but lags behind the US and China in terms of commercialization. 

The US is successful in basic research, talent development and commercialization. Continuous investment in research and development has made the US a global leader in AI technology.

GT: Do you keep an open mind toward technology cooperation with US entities? Can China-US cooperation push forward technological development in the US?

Xu: We are a global company based in China. We are inclusive and open to technical cooperation. We have worked with a team from Duke University and won a global competition in voice activity detection. 

We have opened an internship program to the students of Stanford University. We are seeing more and more students taking internships at our Lenovo Research labs - not only Chinese from America, but also non-Chinese students, because the Chinese market is becoming increasingly attractive due to China's economic growth.

We benefit from open-source projects. In return, we would like to contribute to them. But to be a responsible contributor, we need to clean the codes and make explicit notes to be in line with international standards. That would mean a lot of extra work, but we will overcome this issue because we want to contribute to global innovation and technology.

GT: Currently, there are concerns that China and the US will likely decouple their economies from each other. Is there any possibility that the technological worlds in China and the US will be completely split?

Xu: A senior Chinese official recently said that no country can develop alone in the era of globalization, and no cutting-edge technology can be developed without mutually beneficial cooperation. I agree with this. Scientific work extends beyond national boundaries.

I fully support data sharing and exchange. Take the example of a rare disease; it is possible that there are only 100 cases in China, 20 cases in Germany, and 10 cases in the US. Data sharing is a precondition for smart medical care which adopts AI technologies. 

The world must establish an information sharing mechanism among different countries. Smart medical care is not just a challenge for China or the EU, but also a matter for the UN. If there is a global organization that can promote better sharing of data for wider research and application, such efforts will be beneficial to every country and all human beings.

GT: So, in your opinion, technology exchange among different countries should not be reduced, but needs to be increased? 

Xu: Increased. Yes, of course. Actually, I also believe international exchange will be boosted in coming decades.

GT: What is China's overall advantage in the AI field?

Xu: China may have an advantage in smart manufacturing. The country has a good industrial base. After decades of development, China is transforming from the "world factory" into a new smart manufacturing powerhouse, as AI technologies serve as a driving force for development. 

As China's population ages, pressure is building for policy-makers to force an upgrade of the industry toward higher-value products. It is a sensible choice for the country to develop smart manufacturing and AI technologies.

If China combines AI technologies with manufacturing under the concept of Industry 4.0, it may stand shoulder to shoulder with Germany in the next few years in smart manufacturing.

At present, almost all companies globally are talking about smart transformation. Many US companies want to do business in China, while many Chinese firms have turned their eyes toward overseas markets. We have to offer AI solutions with a global vision. We can promote the smart transformation of foreign and domestic firms, and help them to internationalize effectively.