alt Hacker NewsI'm in several facebook groups for cars and bikes, and holy smokes, someone asks a question and usually the top response is someone literally pasting in a screenshot of a google AI generated response....which is wrong like, solid 50% of the time. And then you point it out and people get aggro on you. I hate it.
Replies
This phenomenon is so widespread and I can't see it as anything other than serious AI psychosis.
I think people want to feel useful and helpful in general, and now a lot of people are getting this feeling by using AI to answer questions. And it's like you said they get so mad when you point out that their AI screenshot is wrong or, even more obviously, that the person who asked could also just read the google gemini blurb. They get mad because you're invalidating their feeling of being helpful.
Last year someone went missing on the water in my area. There was a lot of talk on facebook during the search, and a coordinated facebook group. And every post would have people posting their chats with ChatGPT, or saying they used AI to analyze the weather and currents on the day he went missing to figure out he probably landed on this or that shore (he didn't). These posts are equivalent to the mediums and psychics who also come in with these unhelpful advice posts.
I think fundamentally it's the same as how AI is giving some people an extremely strong feeling of being an expert, and how their psyche seems to lash out to protect that feeling when it's threatened.
Another facebook anecdote, I recently saw someone post in a local genealogy and history group some pure AI slop infographics. When people called him out he said that it's AI but he carefully vets all the information and that they need to stop being luddites or they'll be stuck in the past blah blah devolving into calling them liberals and the like. But his infographics were all incredibly obviously wrong, to anyone who had ever seen a map it's obvious that the coastlines were incredibly wrong, and the city weren't in the correct locations at all. So what can you even do with someone like that? They're just going to retreat into the safety of the chatbot to continue their "research".
I have literally seen multiple screenshots now of people pasting Google Gemini answers in an attempt to tell other people whether something they're doing is legal or not in a particular state, as it relates to NFA item firearms in the USA... And they seem to take it as the gospel truth. People are literally risking federal prison over information that Gemini pulled out of the ass-end of some internet forum of multiple cletus-intellect level individuals talking about their homemade 3D printed silencers. And then Gemini further hallucinates and fabricates new ways to parse that information different than how the original person wrote it. It's astonishing.