Ten advances in mathematics and theoretical computer science
Points and comments are a snapshot, not live.
OpenAI reveals ten new results in mathematics and theoretical computer science from a model.
OpenAI shares ten results from an internal version of its Astra model, each resolving or substantially progressing a long-standing open problem. The problems span high-dimensional geometry, coding theory, arithmetic circuit complexity, group theory, operator algebras, quantum complexity, lattice cryptography, and extremal combinatorics. The total compute cost was roughly $2,000 at Sol API rates. Humans prepared manuscripts and formalized each proof in Lean. OpenAI takes responsibility for correctness but states the mathematical arguments were generated by the system.
What commenters are saying
Commenters split on AI's impact on mathematics. Some see a positive shift, arguing AI will generate more math and require more human mathematicians to ask questions and guide discovery. Others express existential dread, comparing the trajectory to chess, where only a few top professionals sustain a career. A key distinction raised: math has real-world utility and paid careers, unlike chess. Concerns about credit, verification burden, and the destruction of traditional career paths are prominent. One commenter notes the pace of change feels unprecedented.