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.