Researchers call Codex 'read chat transcripts' rumor baseless and ask it to stop

joshgans · x · 2026-09-08

Afinetheorem, retweeting @sholtodouglas, says the insinuation in the Navier-Stokes letter that OpenAI folks "read chat transcripts to get the idea" makes no sense to anyone who knows how these models and data contracts work, urging people to stop repeating the rumor.

Related event: Researchers Push Back on Claims That OpenAI Codex Trained on User Data(14 posts)→

Original post →

More from Models

Models channel →