Mathematician Terence Tao has expressed the view that high-quality unsolved problems in mathematics are being "mined" by AI in a non-renewable manner, potentially making such problems scarce in the future.
Terence Tao Warns of Risk That AI May Consume Unsolved Mathematical Problems
This article is a translation. Read the Japanese original
According to Tao, while advancements in technology and infrastructure lower the barriers to mathematical research and push the field forward, they also have a side effect of flattening the "difficulty landscape" of the discipline. He specifically pointed out that the use of AI could impair the ability to identify promising questions within certain domains.
He also mentioned the tendency of AI companies not to disclose the processes leading to a solution or the negative results encountered. Tao warned that if AI rapidly solves existing unsolved problems, it could lead to a phenomenon of "scooping," where problems are "eaten up" before researchers can derive meaningful insights. He stated that this could hinder the tradition of open science and the healthy development of the academic community.
Sources
- Tao: Open math problems being non-renewably mined by AI （Hacker News Frontpage, 2026-09-08）