OpenAI’s early October release of 722 AI-generated papers—produced by an unreleased frontier model after attempting roughly 4,000 problems—has sharply lifted trader sentiment across the market. Many results include Lean formalizations and target longstanding open questions in algebraic geometry, number theory, and theoretical computer science, building on the company’s September claim of progress on a Navier-Stokes variant. Complementary advances from DeepMind’s AlphaProof Nexus on Erdős problems and Meta AI outputs have reinforced the view that scaled reasoning and agent harnesses can now address research-level mathematics. With only weeks remaining until year-end and verification backlogs at bodies like the Clay Institute still significant, leading contracts such as the Hodge Conjecture trade near 20% implied probability, reflecting both the rapid capability gains and the high bar for confirmed, announced solutions.