What annoys me is the way yhe media talk about stuff like this. "Chatgpt said". "Grok apologized". No it didn't, you stupid fucks. It is not a person. It is not sentient. It cannot apologize. It cannot say anything at all. Its 'apology' is the result of someone instructing a computer to output something resembling an apology. There's as much meaning to it as printing out a text document.Seriously, countries need to draft laws to shut a load of this shit down. And bluntly, it needs to be targetted at the companies.
Apparently, robotaxis are operating in a legal grey area too. They cause a traffic violation, who pays the fine? The state can send a charge off to a company, but it appears that legally there's no-one to hold accountable so the companies won't have to pay them.
I understand that the conversation about AI faces some difficulties. We don't really have the language for something that 'talks' without also being sentient. But all this anthropomorphizing of AI is achieving l, is letting the tech bros actually running the shots off the hook.
So yes, fully agreed, it's not AI that needs legislated.