Comment by cedws

What's the endgame of this "AI models are trained on copyrighted data" stuff? I don't see how LLMs can work going forward if every copyright owner needs to be paid or asked for permission. Do they just want LLM development to stop?