[shiandow]

I think LLMs have stopped trying to be accurate language models for a while now. There's probably some kind of feedback loop that encourages expressions that correlate with 'good responses', if there's nothing preventing it from using the same expressions again and again then it won't stop.