
Humans outperform AI in rigorous mathematical research testing
The First Proof project tested four AI systems against original, unpublished research-level math problems, finding that while AI shows promise, it still lags behind top human researchers.
虎嗅 · 104d ago



















