Many people believe in the possibility of artificial intelligence: they think it is possible to produce intelligence just by programming a computer in the right way. Or maybe programming alone is not enough. Maybe the computer’s internal states would also need to relate in the right way to its surroundings, giving it something like perception. And maybe it would need to cause perceptible changes that would manifest its intelligence, giving it something like action. Artificial intelligence might only be possible in a sort of robot. Let us suppose that any such further requirements are satisfied.