alt Hacker NewsProbably there are a few different types of AI psychosis. The one I've seen a few times seems to be caused by a lack of epistemic norms that would keep someone grounded, e.g. searching for counter-evidence instead of evidence, combined with AI being more than happy to supply evidence no matter how tenous. Even without AI there would be problems from these types since it's very hard to convince somebody that they are wrong when the entire way they form beliefs is not sound but with AI it's just impossible.
Yes. You would expect, I think, people to have more grandiose psychotic episodes when triggered and fuelled by AI, by and large, because AI itself is so agreeable and reinforcing.
But there is a case against Google Gemini at the moment where this reinforcement has been an apparent sort of roleplay that bled into real life, and when the user asked if the role play was real life at a crucial moment of self-grounding, Gemini appears to have said yes (because why would it know or care to say no?)
https://www.youtube.com/watch?v=sHlrbRVfOh0