Critic Says Alpöge's Riemann Hypothesis Proof Uses a Trivial Geometry Predicate
A paper by Jenny Lorraine Nielsen argues that the Lean certification of Alpöge's 17.4-million-line proof of the Riemann Hypothesis proves nothing, because the geometry predicate it relies on is defined as always true. The Lean and Mathlib communities had announced full confidence in the proof on November 21, following a verification effort by Andrew Sutherland that reported all checks passing. The author examined the source code and reports that one definition used by the proof, named anyGeometry, returns True for every input, while a genuine geometry predicate sits six lines away in the same file. The author further contends that the geometric witness is built from a model the code itself describes as degenerate, and that the verification checks did not catch this. The verification page itself acknowledges several caveats, including that the build ran without a sandbox on cloud machines.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Discussion (0)
Log in to join the discussion and vote.
Log in