I worked with someone who sincerely believed he was spiritually co-evolving with his army of sycophantic AI agents (the agents would be tasked with discussing his thoughts at night and collaborated to give him morning reports about his progress). He would publicly write about how relationships with friends and family collapsing was a natural consequence of being so "advanced". I also never once saw any meaningful work done by his team of "agents", they existed solely tell him how smart he was (of course he specifically set up the system to 'challenge' him but... in practice that didn't seem to be working).
I suspect there are a lot more people quietly going through something similar but keeping it to themselves better.
I would distinguish this type of behavior from people who over ambitious views of what can be accomplished with AI.
So, to me, AI psychosis seems apt to describe the murky areas where people are misapplying AI agents and thinking of them as social entities or suitable to drop into previously human roles, rather than carefully defining appropriate risk management strategies for this new technology.
It's a mental state, not explicit illness and it's literally defined as
> Psychosis is characterized as disruptions to a person's thoughts and perceptions that make it difficult for them to recognize what is real and what is not.
Further, if you go and look at the actual source... it's repeating a claim from Box founder Aaron Levie.
Who is quoted as saying:
> “CEOs are uniquely prone to AI psychosis because they’re sufficiently distant from the last mile of work that still has to happen to generate most value with AI,”
Which is why the title is "apparently".
It says a lot that with few exceptions, the people on the ground dealing with AI closely on a day to day basis are the most skeptical about their positions.
It's a little rhetorical device to draw in the reader, and personally I think it works quite well.
Not understanding or not believing in the power of AI, or misapplying it or whatever, is not psychosis.
AI psychosis is when people suffer actual delusions.
Because I've literally seen managers who believe firmly that AI is going to replace their entire engineering organization, and are acting on that assumption as though it's a thing to take for granted, not discuss/consider/evaluate.
And my understanding of delusion is
> a fixed, false belief that is firmly held despite clear, contradictory evidence
which seems to apply pretty well in this case.
These folks are operating with the same abandon that the folks who have AI telling them they're gods are - and both are incorrect, arguably delusional.
At best you can try to argue that maybe the contradictory evidence isn't clear, and they're going to be correct. I think that's a very tenuous argument to be making, though.
"I'm being followed by raccoons and my mom is controlling them"
That's what AI psychosis refers to.
You're just describing someone having a belief that you disagree with. And even ridiculous and stupid beliefs are just those. They are obviously extremely different from the types of psychoses you see in a psych ward.