Mathematicians Call Astra 'Big News' but Warn Nobody Has Had Time to Referee the Proofs
Thomas Bloom of the University of Manchester called the Astra results 'big news' and more significant than May's counterexample to the unit distance conjecture, while cautioning that AI is not replacing mathematicians — Astra draws on over a century of human-generated theory. Coverage consistently notes external mathematicians have not had time to work through arguments at the depth these conjectures normally attract, and that reviewers will want to know how much human problem-shaping preceded each run. Terence Tao, who signed June 2026's Leiden Declaration on AI and Mathematics, has framed current models as tireless literature-scanning assistants rather than sources of deep original ideas — the Lean certificates address correctness but not that provenance question.
↳ Follow the thread