When AI takes away the thrill of discovery.

OpenAI and DeepMind rapidly advanced work on Millennium Prize–level math, with OpenAI posting 722 manuscripts covering 372 problem families after a Navier-Stokes existence and smoothness approach.

• Why it matters: The stakes are whether AI-driven progress can satisfy the desire for answers quickly without undermining the human “thrill of discovery” that motivates understanding.
• What they did: OpenAI deployed many AI agents at millions of dollars in computing cost for the Navier-Stokes problem, then later shared 722 manuscripts on 372 mathematical problem families.
• The result: The flurry of AI-reported progress, alongside DeepMind’s Science publications, is linked to major disruption in mathematics and raises questions about how understanding compares to discovery.

AI summary of the abstract · may contain errors · not medical advice · read the paper.