I think the main issue is it seems no one even bothers to check if what the LLM just spat out is factually correct. It kind astonishes me how little some people seem to doubt whatever the AI is telling them.