Popularly the idea seems to be: our AIs are getting more and more capable, and someday they will achieve AGI, which is basically a digital version of a human consciousness and can do everything a human brain can do. From there, now that it can code its…

Leaving aside all the other ethical questions of AI, I'm curious about the pros and cons of LLM use by people with mental health challenges. In some ways it can be a free form of therapy and provide useful advice to people who can't acces…