Tufty Indigo 🪗 on Nostr: Kinda disappointed to see a science communicator uncritically repeating nonsense ...

Kinda disappointed to see a science communicator uncritically repeating nonsense claims like "upholding these physical texts as uncorrupted by shoddy AI writing that’s already polluted so much of the internet (and indeed, newer books)." Outlets like 404 Media were pushing this misconception earlier in the year, that AI-generated text is somehow harmful for training and will lead to a situation where models get worse because they're training on increasingly "sloppy" writing over time. In fact LLMs are trained predominantly on LLM-generated text, and it's one of the reasons performance has improved so fast over the last year.