Mathematicians claim proof of 167-year-old Riemann hypothesis using AI, human methods
New Delhi, Nov 21 (IANS) Two mathematicians have made competing claims to have solved the 167-year-old Riemann hypothesis, one of mathematics' most famous unsolved problems, setting up a rare contest between an AI-assisted approach and a proof developed through human mathematical reasoning.
Levent Alpöge, a mathematician at artificial intelligence (AI) company Anthropic, earlier this week said that he had discovered a proof of the Riemann hypothesis with assistance from Anthropic's Claude AI model.
While Alpöge's work has cleared several important computer-based verification checks, Masullo’s proof has attracted queries from mathematicians examining the draft.
However, the problems reported by several mathematicians were also identified with them using assistance from different AI models.
Alpöge released a four-page note and a 14-part archive containing nearly 17.4 million lines of Lean code, a mathematical language used to formalise and verify proofs. He said the work also establishes the Generalised Riemann Hypothesis for Dirichlet L-functions.
The Riemann hypothesis was proposed by German mathematician Bernhard Riemann in 1859, as a mathematical theory on the pattern in which prime numbers appear.
Proving the hypothesis is among seven major mathematical challenges identified by the Clay Mathematics Institute in 2000. However, a proposed proof must undergo at least two years of scrutiny after appearing in a qualifying peer-reviewed journal.
—IANS
aar/ag