That sounds both kinda fascinating and extremely disturbing. Especially since LLMs, in my experience, have a lot of trouble rephrasing or summarizing things I say in a way I would agree with. Maybe it's just the way I communicate, but I don't trust AI to speak for me at all.