Why we’re falling out of love with our AI confidants
https://psyche.co/ideas/why-were-falling-out-of-love-with-our-ai-confidantsOpen linkView original on lemmy.worldhttps://psyche.co/ideas/why-were-falling-out-of-love-with-our-ai-confidantsOpen linkView original on lemmy.world
So even the astonishingly stupid people who were using AI for therapy are starting to realize it doesn't work.
I think the more emotionally desperate people are the ones who are most likely to depend on a LLM. It is built to be sycophantic, so the emotionally vulnerable easily become entrapped.
That's why it validates everything you say to it, even for mundane office tasks. Even vulnerable people who aren't using it for therapy end up feeling better when they use it, so they use it more.