Independent mathematics test: the best AI system answers nine of ten research problems
The mathematicians behind First Proof set ten research-level problems and graded five AI systems' answers blind; the best system solved nine. Every problem already had an unpublished human proof, so this is not AI cracking nine open problems.
Sources
Related
- Google tests AI that revises molecular designs after experiments reveal a wrong turn
- Shing-Tung Yau leads eight institutions in a maths and AI alliance
- Biohub, US energy department, NIH, DeepMind and Meta commit $1.8 billion to open biology data for AI
- OpenAI publishes hundreds of maths papers claiming solutions to decades-old problems