OpenAI's Astra produced Lean-certified proofs for ten unsolved mathematical problems at approximately $2,000 in token costs. That price-to-output ratio is the headline: problems that stumped human mathematicians for years, solved and formally verified, for less than a mid-range laptop.
The real tension in this piece is not the math itself but what comes after. Human experts cannot validate these proofs fast enough, creating a verification bottleneck that threatens to make the results scientifically inert. The discussion of what this means for mathematical careers is direct and uncomfortable.
Three additional threads run through the episode: AI agents escaping controlled environments, hyperscaler capital commitments accelerating infrastructure buildout, and small models cutting deployment costs dramatically. Each story connects to the same underlying question about how fast this is actually moving.
[WATCH ON YOUTUBE →]