It is troubling that there are some users of LLMs/AI whose mental health may be indicators that they may need supervision when using those tools. LLMs are successful due to how long they can keep a user being engaged with them. If the "reality" that LLMs create replace the reality of the user, that is a recipe for self-harm or for some other costly phenomena. I do not see the CEOs nor designers of LLMs to be overly concerned about the effect that these tools have on some of its users. It's gonna get more and more troubling. Forgive my long rant.
