zlacker

There are all sorts of dangerous things where there are restrictions on what the common people can do. Prescription drugs and fully automatic machine guns are two examples. You can't open your own bank either.

For anyone who really believes that AI is dangerous, having some reasonable regulations on it is logical. It's a good start on not being doomed. It goes against everyone's egalitarian/libertarian impulses, though.

The thing is, AI doesn't seem nearly as dangerous as a fully-automatic machine gun. For now. It's just generating text (and video) for fun, right?

>>skybri+(OP)
I move hundreds of thousands of my dollars around between financial institutions just using text.

>>skybri+(OP)
AI and machine guns aren't comparable. Machine guns will never ever decide to autonomously fire.

The shared point of both AI alarmists and advocates is that AI will be highly resistant to being subject to regulation, ultimately. As dictated by the market for it. They won't want to regulate something, assuming they could, for which its free operation underlies everyone's chance of survival against competing systems.

I only find that danger is inherent in the effort of people that casually label things as "dangerous".

I'm still exploring whether its the laziness aspect, itself, of the alarmist vocabulary in the absence of required explanation. Or whether my issue lies with the suspicion of emotional manipulation and an attempt to circumvent having to actually explain one's reasoning, using alarmist language absent required explanation.

Already, AI pessimists are well on their way to losing any window where their arguments will be heard and meaningful. We can tell by their parroting the word "dangerous" as the total substance of their arguments. Which will soon be a laughable defense. They'd better learn more words.