AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Can You Design the Ultimate Objective of the Machine?
Can you design the ultimate objective of a machine in such a way that it never can work against the welfare of humanity? We do this absolutely all the time, designing objective functions to entice humans to do the right thing. This is not something that's new in aa. We've been doing this for thousands of yearsit seems you need something more than the laws,. And you need some notion that humans are above the law and can tell the ai when it's exploiting loophole.