Will an AI System Solve One of the Remaining Millennium Prize Problems by June 2025?
37
Ṁ5320
2025
8%
chance

The Millennium Prize Problems are a set of seven of the most notorious unsolved problems in mathematics that were stated by the Clay Mathematics Institute in 2000. A correct solution to any of the problems results in a $1 million prize awarded by the institute.


As of now, only one of these problems, the Poincaré Conjecture, has been solved by Grigori Perelman in 2003. The remaining six problems: Birch and Swinnerton-Dyer Conjecture, Hodge Conjecture, Navier–Stokes Existence and Smoothness, Riemann Hypothesis, Yang-Mills Existence and Mass Gap, and the Diophantine Equation, still remain unsolved.


This question asks whether an AI system will be credit to have solved one of these remaining six Millennium Prize Problems by the end of June 2025.

Resolution Criteria:
The question resolves positively if, by the end of June 2025, at least 2 reputable papers (e.g., on arXiv or in peer-reviewed journals) show an AI system solved one of the remaining six Millennium Prize Problems. The AI should not just verify a human-proposed solution but must be credited as being the primary solver. It's acceptable if the AI was initially trained or directed by humans. The question resolution isn't dependent on the Clay Mathematics Institute's official recognition but on the academic acceptance of the AI's solution. If the solution is disputed or inconclusive, resolution may be postponed until a consensus is reached.

Get Ṁ1,000 play money
Sort by:

Only 2 years left until June 2025 (to the month!). I'd put this at a higher probability if the deadline was 2030 or later, but given how slow progress on named mathematical conjectures tends to be I put this question at a very low subjective probability (something like ~5%?).