I dunno, I think there might be a confluence in suggestible types with a feedback loop created by these products. Yes, there are always people who are challenged, but interfacing them with LLMs seems to be especially dangerous.
pmc.ncbi.nlm.nih.gov/articles/PMC10686326/
futurism.com/openai-investor-chatgpt-mental-healt…
psychologytoday.com/…/can-ai-chatbots-worsen-psyc…
rollingstone.com/…/ai-spiritual-delusions-destroy…
hanrahan@slrpnk.net 8 months ago
Sure but normally they’re confined to 4chan
ArgumentativeMonotheist@lemmy.world 8 months ago
Lol, they’re just poorly socialized and on the spectrum.