Chinese mathematicians have formed an alliance to combine artificial intelligence with advanced mathematical research, a self-help move aimed at closing a gap with the U.S. in cutting-edge AI models and the research built on them, even as China rapidly catches up in applied models.
The Shing-Tung Yau Mathematics and AI Laboratory Alliance, led by the Chinese-born mathematician Shing-Tung Yau, launched on the 11th, the South China Morning Post reported. Yau won the Fields Medal, often called the Nobel Prize of mathematics, in 1982. Eight major research institutions are taking part, and the alliance will set up a system for about 1,000 researchers to collaborate.
The group plans to focus on developing AI models tailored to mathematical research, using AI to generate mathematical conjectures, and building systems that make reasoning processes easier to interpret.
Yau, who directs the Yau Mathematical Sciences Center at Tsinghua University, said he had not expected AI to become this powerful. "A year ago, I didn't think AI was this strong," he said. "The advance of AI in mathematics is like a tsunami." He stressed that Chinese mathematicians should join forces and take an active part in the technology rather than avoid it.
Long-standing problems in mathematics are indeed being solved, led by U.S. companies developing AI models. OpenAI recently released 722 mathematics paper manuscripts sorted into 372 groups of results, saying it had made progress on hundreds of unsolved problems. The company also claimed it had used AI to solve the Navier-Stokes problem in 88 hours.
In response, 25 Fields Medal winners, including Princeton University professor June Huh and University of California, Los Angeles professor Terence Tao, expressed concern about the negative effects of AI in a joint letter. Their worry is that the more AI companies hold up problem-solving ability in mathematics as a measure of technological progress, the more the conceptual understanding and insight unique to mathematics could be eroded.
Yau pushed back, saying Tao is an excellent mathematician but does not speak for the entire mathematics community. He also said the gap between China and the U.S. is pronounced in AI systems capable of driving advanced scientific research, particularly in mathematics and physics.
China won a Fields Medal this year but had no names on the Nobel Prize list. That is a somewhat thin showing in the sciences compared with the country's repeated success in startling the world with low-cost, high-efficiency AI models in what has been called the DeepSeek and Kimi shock. Still, University of Chicago professor Deng Yu and New York University professor Wang Hong this year became the first Chinese nationals to win the Fields Medal. While that marks notable progress for Chinese mathematics, the industry appears broadly concerned that such achievements must carry over into cutting-edge AI research.
Yau said Chinese AI models perform well at the level of ordinary users but fall short at the cutting edge. "If you ask Chinese AI models to do what OpenAI has done over the past month, the gap is very large," he said.
"The U.S. and the West are gathering top-level mathematics experts to work with AI and explore advanced problems, while China's efforts are scattered," he said.