Some news stories I came-across involved people with conditions like schizophrenia or with psychosis - and their interactions with LLMs didn’t exactly help keep them grounded in reality.

…but that is distinct from the people who noncritically appraise ChatGPT’s stochastic-parrot wisdom.

…and both situations are problems and I’ve no idea how the LLM vendors - or the public at-large - will address them.