Fudan University professor sees bright AI future for China
Doctoral research is 'top class,' but fostering creativity remains a challenge
SHANGHAI -- China is turning out a growing number of world-class artificial intelligence researchers, and rapidly catching up with the U.S. in terms of the number of patent applications.
Its work in the field is led by universities such as Peking and Tsinghua universities, along with information technology companies such as Baidu. In a recent interview with the Nikkei, professor Wei Hui of Fudan University explained how China has been able to produce this talent.
Q: Is China going to be a prime producer of AI researchers in the future?
A: AlphaGo, AI-based go [a Chinese strategy game] software developed by a Google cousin, drew people's attention to deep learning. Young talent inspired by that software will gather [in China].
Across China, there are many universities and research projects. At international symposiums, Chinese educational institutions are beginning to make their names with their research results.
China is already past the stage of following previous studies done by overseas researchers. Graduate school students [in China] must submit dissertations that surpass global academic standards in order to receive their degrees. Research in the U.S. is no doubt top class but at the doctoral level, China is no less so.
Q: Will patent applications by universities grow significantly?
A: About a decade ago, researchers and funding were scarce in China. But the tide has turned, thanks to the AI boom. In addition to the government doling out subsidies, China's leading IT companies and startups have been making investment to get access to technologies from universities. There is a virtuous circle of increased funds attracting talent, with universities serving as the place for [developing] technology
China's advantages are its huge population and the ease of gathering big data, which is a basis for AI research. For example, if you think about hospitals and patients' medical records, you can see how much potential there is for AI research.
Q: Is China's education system also helping?
A: In [China's] compulsory education system in primary and secondary schools, the emphasis is on such subjects as math, physics and chemistry. Students are trained to think logically from an early age. Parents understand the importance of scientific learning, and they send their children to cram schools. Programming courses are becoming more widespread, boosting the technological level of computer science.
Q: Doesn't cramming negatively affect children?
A: People used to be critical of focusing on specific areas to develop talent. But things are starting to change. Today's students are taught to think on their own to find answers. The idea of learning things one cannot learn from textbooks is spreading.
At [Fudan] University, I ask my students to have a broad view of the world. It is also important to acquire a deep understanding of the human brain through neurology? and psychology. I do not want my students to stay in their own fields. Students tend to be averse to such an approach because it is less likely to yield results in the short term. But over the long term, it will produce talent who can get ahead of the curve.
Q: What does China need to do to lead in AI research?
A: The AI boom began in the U.S. Chinese research, which tended to be about catching up with studies done in the U.S., is certainly beginning to focus more on quality, such as whether a study has been cited by other researchers or adopted by industry.
Whether China can nurture originality and creativity is the key. Although the number of AI researchers has been increasing fast, few have been trained for these qualities. I want to train researchers who will ditch their past success in pursuit of state-of-the-art technology.
Interviewed by Nikkei staff writer Yasuhiro Morizono