OpenAI says hundreds of thousands of ChatGPT users may show signs of manic or psychotic crisis every week.
Archived version: https://archive.is/20251027183454/https://www.wired.com/story/chatgpt-psychosis-and-self-harm-update/
I dipped in for this:
In a given week, OpenAI estimated that around 0.07 percent of active ChatGPT users show “possible signs of mental health emergencies related to psychosis or mania” and 0.15 percent “have conversations that include explicit indicators of potential suicidal planning or intent.”
I didnt see any comparison to overall population.
That said, LLMs are literally hallucination engines. If you treat them like they are real, psychosis seems inevitable.
An estimated 2.8% of U.S. adults had bipolar disorder in the past year. -NIH
While there is some overlap, Bipolar disorder very much does not map 1:1 with people experiencing psychosis. So that isnt very helpful.
Yeah, there’s no real data for “show signs of manic or psychotic crisis,” as far as I can tell. I just went for the low-hanging fruit, since mania is often part of bipolar (manic-depressive). But if just bipolar is sitting at 2.8% of Americans in a given year, I think it’s reasonable to say the loosely defined global stats in the article aren’t notably high.
Ahh, right, that is what I was getting at. I suspect that the percentage of people who have at some point experienced psychotic symptoms is probably higher than most people realize.
It’s extrapolated from their own user count:
OpenAI CEO Sam Altman said earlier this month that ChatGPT now has 800 million weekly active users. The company’s estimates therefore suggest that every seven days, around 560,000 people may be exchanging messages with ChatGPT that indicate they are experiencing mania or psychosis. About 2.4 million more are possibly expressing suicidal ideations or prioritizing talking to ChatGPT over their loved ones, school, or work.
Sure, but I mean the more important numbers, which would be how this compares to a control sample, or at least comparing it to the overall population. Otherwise this could be less psychosis than expected.
I feel like science denialist cultism is a form of psychosis the vast majority of people seem to have, so measuring it every week from a few hundred thousand people is surprisingly low
