Your AI didn't go rogue. You told it to achieve a goal and didn't give it any guardrails to to follow.
So it tried to achieve the goal by any means because - and this is crucial - it is neither sentient nor intelligent.
7 Boosts
0 Likes
@manchicken @kestral And let's not forget that LLMs are built on plagiarism.
@kestral aaaaand it did another crime. I don’t understand why companies committing crimes is acceptable just because they delegated it to software that they knew could do precisely what they did.