Skeptic quips the Riemann proof is "probably in the training data somewhere"

tlakomy · x · 2026-11-21

Responding to the hype around a rumored AI breakthrough on the Riemann Hypothesis (pointing at Anthropic), tlakomy deadpans: "I feel like Riemann Hypothesis proof is probably in the training data somewhere you guys."

The quip hits the core controversy in evaluating AI math results: is the model genuinely reasoning, or retrieving and recombining mathematical knowledge already present in its training corpus — the most common skepticism aimed at AI math breakthrough claims.

Related event: Anthropic Mathematician Claims Claude-Helped Proof of Riemann Hypothesis(126 posts)→

Original post →

More from Fun

Fun channel →