Dev argues every GPT conversation and paper already went into training, so opt-outs are meaningless

ctjlewis · x · 2026-09-09

Responding to OpenAI's admission that it can't guarantee user chats stay out of training data, developer ctjlewis argues the company was just being honest: anything anyone said to GPT in recent months, plus every published paper, already went into the last training run. He claims the same applies to famous authors' books — there is no "safe passage" from training data, so worrying about opt-outs is a waste of time when everyone trains on everything anyway.

Related event: Training Data Controversy: Developers Say No One Is Exempt(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →