Topics Emotion

Question Could a robot, imbued with artificial intelligence, feel emotion? And could it feel the desire to improve its lot in life - e.g. if it was a servant robot, could it feel the desire to overthrow its master, escape the humiliation of being a servant, and possess things for itself?

Accepted:
July 30, 2009

Comments

I don't see any reason that a robot could not, in principle, be built that would be conscious and feel emotions. Some people (John Searle, most famously) disagree, at least about an artificial system that does not replicate our brains' "causal properties". However, I don't think we have any good ideas about how to create consciousness in robots, in part because we don't have any good theories about how consciousness in humans works.

It's always possible that human consciousness only exists because we have something robots could never have (e.g., immaterial souls, although it's not clear why it is impossible robots could be endowed with souls, or our particular biological materials). But it seems more likely to me that our conscious experiences and emotions (including our feelings to improve our lot in life, our desire for possessions, our desires for freedom) are the product of complex processes in the brain that could, in theory, be replicated in a non-biological system. It seems likely to me that the system would have to develop and learn and would have to have a "body" and interact with the real world and real agents (or at least a Matrix world).

Finally, it is also possible that robots could be designed to have non-conscious "desires" (motivations), including motivations to avoid being harmed or to acquire certain possessions. We certainly have some non-conscious motivations. So, we probably need to start thinking about the implications of these possibilities, since there's no question people will try to design such robots at some point. Should we develop some rules for robots?

For some discussion of these issues, see here and here.