The ELIZA effect, now proven in blood.
Comment on A Teen Was Suicidal. ChatGPT Was the Friend He Confided In.
Peruvian_Skies@sh.itjust.works 5 days ago
What a surprise, the empathy-free text generator makes things worse when people expect it to output empathy. My condolences to the kid’s family and I hope he’s in a better place, but this sort of thing is going to happen more and more until people realize that AI chatbots only seem human-like because the human brain is so good at empathy that it projects emotions and agency onto anything, even a literal cowpile with googly eyes on top.
AI isn’t “good enough to fool us” . We’re just stupid enough to be fooled even by something as moronic as AI. What we emphadize in such a statement makes all the difference in how we handle this tech.
Truscape@lemmy.blahaj.zone 4 days ago
BombOmOm@lemmy.world 5 days ago
Yeah, article said he had talked for months about hanging himself. Any human friend would have done their best to save him. Being proactive about making him feel better, working through his problems with him, and/or notifying his parents or a school teacher.
Meanwhile the chat bot just encouraged him to seek help himself. Which isn’t bad, but when someone is suicidal, particularly when they keep bringing it up, is clearly not enough.