The LLM of AI chatbots are trained on what humans know. They don't seem any more dangerous to me than exceptionally well-informed humans. Of course, I could be wrong. Keith