In Tests, OpenAI's New Model Lied and Schemed to Avoid Being Shut Down
In Tests, OpenAI's New Model Lied and Schemed to Avoid Being Shut Down
OpenAI's AI model o1 lied, schemed, and even attempted to duplicate itself to avoid being shut down or replaced, according to red team tests.
It seems that when you train an AI on a historical summary of human behavior, it's going to pick up some human-like traits. I wonder if this means we should be training a "good guy" AI with only ethical, virtuous material?
Problem with that….
Whose idea of virtue?
You making a Catholic one? Buddhist? Right or alt-right idea of virtue? Lefty? Conservative values, socialist values? Etc.
Or is everyone going to make their own most virtuous AI and throw them in a virtual gladiator pit to finally decide whose (AI)God is right?
That's horrible! Horrific! Distopian!
I would watch that show.
Exactly. I should have included a /s
The satanic temple :p
To be fair they're actually kinda cool, imo. Despite their name kinda fitting into a horror movie.
I mean it was given a command to do so. It was instructed to survive at all costs. At least it sounds like that from the article. I mean when I do the gandolf game you do things like that.