China’s mathematicians form an AI alliance as Western debate deepens

Two approaches to AI in mathematics

Twenty-five Fields Medalists, including Terence Tao and Peter Scholze, worry that AI-generated proofs could undermine mathematical understanding. Researchers may come to rely on tools whose reasoning they cannot follow. Some mathematicians have called for a boycott of such proofs; others are struck by the volume of automatically generated submissions.

China’s alliance has a different stated agenda:

The alliance brings together eight institutions and aims to build a network of more than 1,000 researchers. Yau has also warned that China is behind the United States.

What counts as mathematical insight?

Even among specialists who use AI, there is no agreement on whether it can produce genuine mathematical ideas or only search through known methods. It is also unclear whether that distinction matters.

I think the contrast is less settled than the alliance-versus-skepticism framing suggests. China’s plans put verification and transparency at the center, but the announcement does not say how the alliance will judge whether a model has produced an idea rather than recombined familiar methods. That gap matters: automatic checking can test a proof, but it does not by itself settle what the reasoning means.