3 ms·
The paradox is analogous to AI poisoning its own training data as more and more AI generated content is released on the Internet. Indeed, I see a hard terminus
by satvikpendem 3mo ago
The paradox is analogous to AI poisoning its own training data as more and more AI generated content is released on the Internet. Indeed, I see a hard terminus for both man and machine at some point.
- dnemmers 3mo agoAbsolutely right. First gen models are trained on 'virgin' non-LLM output. Subsequent models are tainted by ingesting AI replies. Rinse and repeat, and all you end up with is AI copies of AI replies, and the noise will completely overtake the signal.
- literalAardvark 3mo agoI'd have been more worried about that had the models not gotten incredibly good over time. But they did get good and this seems like a non-issue.