I remember seeing a talk a while back where the presenter was attempting to define the utility function that characterizes intelligent agents. His thesis was th

I remember seeing a talk a while back where the presenter was attempting to define the utility function that characterizes intelligent agents. His thesis was that an intelligent agent would work to maximize, up to some event horizon, its number of possible courses of action. This reminded me of that talk - money, power, etc... are all ways to increase one's freedom of choice and thus autonomy.