Post claims GPT-6 rewrote Deligne's bound proof in six pages, but paper says no human checked it
haider1 · x · 2026-11-22
A viral post shares a November 2026-dated math note allegedly written by "GPT-6 Astra Pro", expanding L. Alpöge's footnote sketch into a full sub-six-page proof of Deligne's bound for Frobenius eigenvalues — with the only estimate being Cauchy–Schwarz for a character sum over a finite field, avoiding Lefschetz pencils, monodromy, and Rankin's method.
The tweet claims the proof was checked by two Claude models and publicly acknowledged by a Stanford math professor. But the paper's own abstract states it was checked only by Claude Opus 5.5 and Claude Fable 5.2, ending with "No human has checked it" — directly contradicting the post. The model name "GPT-6 Astra Pro" is also unverifiable, so treat this as an unconfirmed rumor or in-joke.
Related event: GPT-6 Astra Pro Writes Six-Page Proof of Deligne's Bound(5 posts)→
More from Fun
- Meme nails it: technical interviews are Godzilla vs. Kong, the actual job is plush dinosaurs — kalyan_kpl · 2026-11-22
- François Fleuret asks: is the gradient Japanese? It never refuses to give you a direction — francoisfleuret · 2026-11-22
- Writer Uses Claude to Pen a Culture Short Story — and Admits It's Obviously AI-Written — NathanpmYoung · 2026-11-22
- The Great Sidebar Debate: A 3x3 Alignment Chart From Slack to Mario Pipes — thesephist · 2026-11-22
- AI Detector Flags Researcher's 2016 Paper as 79% AI-Generated — AntonioCasilli · 2026-11-22
- ChatGPT Has Fully Mastered the Feedback Sandwich, and Users Noticed — tengyanAI · 2026-11-22