Post

Turn off the cooling system and let me know when you hear AI screaming.

I disagree completely with this statement. An AI system is made by people (mostly in a cy.) so if the AI does illegal things, these developers have made a system that is illegal. Therefore these developers and those seniors that did not control their staff , shall be held responsible for AI actions

“Too advanced to control” is not a technical finding — it is a governance claim. If a system cannot be refused, it should not be treated as an admissible product; liability and control are the same question asked from two sides.