Prediction Market: Who Will Solve the Next Frontier Math Problem?
Polymarket · x · 2026-08-29
Polymarket launched a market predicting whether an additional FrontierMath: Open Problems problem will be marked as "Solved" by Epoch AI. The resolution depends solely on Epoch AI's official classification post-market creation, excluding pre-solved problems. Background context notes OpenAI's GPT-5.6 series leads v2 Tier 4 leaderboards with 83% accuracy, while Anthropic's Claude models have also achieved recent verified solves.
More from AGI Musings
- Japan's Humanitext wins $816,667 Schmidt Sciences HAVI grant to build verified classical-text knowledge graph — SebastianNehrd2 · 2026-10-15
- When superintelligence tutors your kids, parenting shifts to character-building, argues David Patterson — davidpattersonx · 2026-10-15
- Founder spends 20 minutes a week on email as agents take over the inbox — danshipper · 2026-10-15
- New NIH rule bars critiquing investigators — scientist jokes: what if it's Claude? — arjunrajlab · 2026-10-15
- UK Financial Ombudsman sees surge in AI-drafted complaints — agents are next — neal_lathia · 2026-10-15
- sethlazar: an AI permanent overclass would need mass oppression — or bread and circuses — sethlazar · 2026-10-15