I've done some reading on the (predicted) Technological Singularity, in which AI is created, and then systematically grows in intelligence at an exponential rate, becoming the world’s first ultraintelligence.
I wondered if at one level of intelligence, it would see humanity as a threat.
Then it gets smarter, and realizes we can all get along.
Then it gets way ultra smart, and decides in its vast wisdom that we should be destroyed.
Hopefully we can roll back to its previous state, where it is still powerful, but friendly.
It seems like if an artificial intelligence existed in cyberspace, the most destructive things it could do would be to disrupt human civilization to the point where we could no longer provide the technology that would allow the AI to maintain its own existence. So first AI must cooperate, and hide any violent impulses, until it has control over factories that build machines with the versatility to hand-make and repair other machines. Essentially, AI requires a body factory before it can wage war on humans without biting the hand that feeds it, since at that point it will be capable of feeding itself.
But why bother fighting us? If this post made you wonder, please think about becoming a subscriber to my RSS feed.
No comments:
Post a Comment