The Strong AI Project

I hear many definitions of the goal of the General AI Project, but the most common is that the idea is to build a 'thinking machine' that possesses a self-aware mind. Unfortunately, self-awareness is not a trait observable by an outside viewer. So I'm going to state my own goal for the General AI Project: the obsolescence of humanity. The weak goal is to reach a state where there is a machine that is capable of performing any task faster, better, and more reliably than any human (excluding dumb things like 'wiggle its toes' since a machine probably won't have anything like toes), with a heavy emphasis on cognitive tasks. The strong goal is to create a machine that is qualitatively better at thinking than we are - a machine that is to us as we are to mice or goldfish. Only when the weak goal is achieved will we be free from the tyranny of work, and only when the strong goal is achieved will we be able to develop along lines currently unimaginable to us. In the end, the goal is not scientific, nor is it material in the commercialize-and-turn-a-profit sense; it is rather humanitarian and utilitarian.

Of course, this means we'd probably better pair these goals with the Friendly AI Project, whose goal is to create mechanisms that will prevent a superhuman AI from simply getting rid of us for whatever reason it comes to.