TITLE: AI Math Breakthrough Claim-Grading Record
PROBLEM: AI math breakthroughs are announced as headlines — "AI solves the Riemann Hypothesis" — while the verification state lags days or years behind, and readers have no standard grade telling them whether a claim is a confirmed result, a compiled formalization, or an unverified announcement.
SOLUTION: An AI math breakthrough claim-grading record: every claim graded DOCUMENTED, REPORTED, or UNVERIFIED, with the grade's evidence and the exact point where the evidence stops — updated as verification proceeds.
EVIDENCE BASIS: DOCUMENTED — The November 20, 2026 Riemann claim is REPORTED, not confirmed: no comparator run completed, no preprint, no peer review, no Clay Institute statement. REPORTED — Scientific American's follow-up reports mathematicians "almost certain" the proof is real, with the comparator run still pending. DOCUMENTED — In August 2026 Anthropic said an unreleased Claude improved a long-standing RH-related result, stressing it had not proved it. The combined AIM system is inventor CONCEPT ONLY.
| Idea priority | Low |
| Needed By | Not sure -- Just thought it was cool |
By clicking the "Post Comment" or "Submit Idea" button, you are agreeing to the IBM Ideas Portal Terms of Use.
Do not place IBM confidential, company confidential, or personal information into any field.