AI companies taught us that if you commit a crime, it is not you doing anything wrong.
It is just your neural network that "went rogue".
2 Boosts
1 Like
@randahl "Sorry officer, my car went rogue"
@randahl If a company designs a product and know that it has a high likelihood of causing harms, then they are responsible for those harms. It is legal negligence. You can't just say "Oops, there it goes again. LOL."
let us go rogue on data centers and destroy LLMs.
@randahl What interests me more is this: if we gradually hand over more and more decisions to algorithms, who should ultimately be responsible for the consequences?
@Paul_stilgar yes officer, I did steal the turkey from the supermarket, but it was only to train my large language model on the concept of turkey!
@randahl This started long before AI. Republican voters here in the US have been legalizing corporate crime for 60 years. At the very same time they've driven the USA up to the world championship of imprisonment, primarily by jailing black people for marijuana and other nonviolent drug crimes invented shortly after Civil Rights. Many people being hurt by AI are conservative voters who support the Broligarch criminals.
@randahl Technically true, in all cases.