The new article “Mathematics of the Future”, written by Pavel Etingof and Slava Gerovitch from the Department of Mathematics of MIT, discusses how rapid advances in artificial intelligence are transforming mathematical research, raising critical questions about the role, training, evaluation, and social value of human mathematicians.
“In 2026 frontier LLMs gained the ability to do mathematical research, and this ability is rapidly growing. Long-standing open problems began to fall to AI-powered attacks one after another. LLMs can now generate a decent (and sometimes excellent) mathematical result within an hour, with minimal or no intervention by a human mathematician. While it is not yet clear how far this trend will go, there are good reasons to believe that it will continue, with AI uncovering a large amount of highly interesting mathematics in the immediate future. For the growth of mathematical knowledge, this is extremely exciting. Yet this comes at a steep price. These developments raise fundamental questions about authorship, credit, evaluation, and the role of mathematicians. It is already clear that in the years and even months to come, our community will have to undergo profound change. Even if AI progress stopped today, mathematics could no longer be practiced the old way, and we would still need to adapt our teaching, research, and communication.”
Join the discussion
Sign in with GitHub to leave a comment.