Urschel's math proof

News · Heat 1.0 / 10 · 2 shows · 7m airtime

On October 12, John Urschel, a former Baltimore Ravens offensive lineman now at MIT, published a paper proving a 35-year-old hypothesis. Early coverage framed it as a math problem solved by an offensive lineman rather than by AI, contrasting it with Sam Altman's 10,000-agent effort. That framing is now contested: the paper itself cites the help of OpenAI models, leaving open how much of the proof was his alone.

How 2 podcasts covered it

1. TBPN, 2026-10-09 (16:21 to 22:00)
   Profiles the lineman turned mathematician
   Profiles the former Ravens lineman turned MIT mathematician and his random-matrix paper, which cites OpenAI models - one more entry in the deluge of AI-assisted proofs.
   New angle: Urschel's random-matrix paper cites the help of OpenAI models.
2. The Best One Yet, 2026-10-12 (1:09 to 1:35)
   Adds Urschel's proof of a 35-year hypothesis
   Recounts that Sam Altman assigned 10,000 AI agents to a century-old math problem, solved after 86 hours of compute, then pivots to John Urschel, the ex-Ravens lineman turned MIT professor whose new paper proves a 35-year-old hypothesis.
   New angle: John Urschel, a former Ravens lineman turned MIT professor, published a paper proving a 35-year-old hypothesis.
3. The Best One Yet, 2026-10-12 (1:37 to 2:45)
   Recaps the Ravens years and MIT post
   Recaps the Urschel story - the Ravens years, the walk away from the NFL, the MIT professorship - and frames the paper as a win for human math over AI. Stays inside the card's read, with a Newton comparison and a sponsor break.