I think that AI would be programmed to think in terms of an independent, intelligent agent that competes with other entities (Toshiba vs. Sony vs. Toyota perhaps). Tic-tac-toe logic dictates that randomness and sacrifice is necessary to win in the longrun, and those emotional aspects contribute enough randomness to actions to make the outcome non-predictable yet effective in a self-learning environment.

Oh, yes — if an artificial entity such as corporations needed to create legal protections to safeguard profits, I don’t see why moneymaking machines wouldn’t get similar artificial yet legal, profit maximizing protections.

The idea that AI’s evolution will evolve to the point where “sentient” machines would have the same needs and desires as human beings is fanciful.

Humans needs are also externally manipulated; humans have market driven needs, so I think corporations and profit will surely have an easier time of this imprinting human-like behavior on robots.

Those are animal emotions which are simply more complex in humans because our brains are more complex (not to mention, we can understand human emotions better, being human). Cats and dogs exhibit jealousy. Almost all animals have fear responses; many have behaviors associated with play that can be attributed to joy.

Our own emotions aren’t a result of a specific need, they’re the byproduct of a delicate balance of chemical and neural interactions. Like all the frilly stuff on a Mandelbrot diagram, emotions aren’t the raison d’etre of the complexity of the human mind. they’re a result of the complexity of the human mind.

Unlike biological minds, AI isn’t going to spring from the same tree of evolution, but any AI of a complexity sufficient to be considered sentient (by which I mean self-aware like a small animal) is going to be complex enough to develop quirks that will be the robot equivalent of emotions. They’ll probably be mappable to recognizable human emotions.

I mean, that is, unless you believe that there’s some spark of soul inherent in the chemicals that make up the human body that’s missing from the chemicals used to make circuit boards and memory chips.

Intelligent Machines will have none of our evolutionary baggage, precisely because they won’t be an accidental side-effect of primate ambition. (Well, OK, they will, but not by way of “reproduction after the same kind”. ;-) ) Some much more serious issues will be authority (who gets to give orders to the machines), inflexibility, and accidental obsessions/superstitions. (That last seems to be a issue of associative memory in general, so I think it’ll be a problem for early AIs.)