Fuck AI

1466 readers

422 users here now

"We did it, Patrick! We made a technological breakthrough!"

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

founded 8 months ago

MODERATORS

181

Gemini AI tells the user to die — the answer appeared out of nowhere when the user asked Google's Gemini for help with his homework (www.tomshardware.com)

submitted 2 weeks ago by [email protected] to c/fuck_ai

34 comments fedilink hide all child comments

Let’s be happy it doesn’t have access to nuclear weapons at the moment.

you are viewing a single comment's thread
view the rest of the comments

[–] ggppjj 4 points 2 weeks ago* (last edited 2 weeks ago)

The question before it was a multiple choice question that included "a Harassment threaten to abandon and/or physical or verbal intimidation".

Seems like it lost the context of "choosing multiple choice answers" and interpreted the prompt wrong.

Edit: To be clear: This is a garbage hyped-up auto-complete that has failed at the task it's being shoved into "fixing". It's all-around bad. Google is a bad company, Gemini is a bad product that can't tell the time. And also, this particular article is over-hyping a reasonably understandable bad response given the context of the input and the understanding that Gemini is a bad product.

The product (I don't want to call it AI because it seems like giving it too much credit) isn't just deciding to tell people that they should die, the product is incapable of rational thought and the math that governs its responses is inherently unstable and provides unavoidably inconsistent results that can never be trusted as fit for any purpose where reasoning is required.

I want to be mad at it for the right reasons and be raging against Google and the current crop of technocratic grifters with my eyes wide open and my course set to ram into them instead of blindly raging off in the wrong direction.