Big Lab Researchers Read Almost Zero Papers, Claim Top Conferences Full of Overclaims

kellerjordan0 · x · 2026-08-05

A user revealed on X that most researchers at frontier AI labs now read almost zero academic papers, viewing top conferences like ICLR, ICML, and NeurIPS as primarily full of overclaims and fraud, though a few diamonds remain. The author noted that while this is obvious to insiders, it is a crucial reality check for unconnected students.

Related event: Ex-Big Tech Researcher Slams AI Conferences for Hype(4 posts)→

Original post →

More from Fun

Fun channel →