The OpenAI proof disproving Paul Erdős's 1946 planar unit distance conjecture now travels with a companion paper carrying nine mathematicians' names. The list, as published in the commentary PDF on OpenAI's site, includes Noga Alon, Thomas Bloom, Timothy Gowers, Daniel Litt, Will Sawin, Ananth Shankar, Jacob Tsimerman, Yufei Wang, and Melanie Wood. [1]

The paper's Monday brief on Thomas Bloom's verification and the Annals recommendation registered the first external endorsement. The second-day artifact is now the named cohort. Sawin, in Princeton commentary, characterized the construction's gain as roughly 1 percent more unit-distance pairs per doubling of point count — small in margin, decisive in implication, because Erdős's conjecture demanded near-linear growth. [2] Litt told Scientific American that the result is "the unique interesting result produced autonomously by AI so far." [3]

Bloom's X post called it "without doubt, the most impressive achievement of AI in mathematics so far." OpenAI's announcement post traveled the platform alongside Greg Brockman's framing — the first time, he wrote, an AI system has autonomously solved a prominent open problem central to a field of mathematics.

The path to Annals submission is the next test. A journal of that altitude moves on its own clock; verification by a recommendation letter is not yet acceptance. What the paper has, six days after announcement, is a 125-page argument and nine signatures attesting that the construction holds.

-- ANNA WEBER, Berlin