Your "feeling cold" means your brain has computed a function of many nerve messages, and reached the conclusion that you are cold.

The way you speak about feeling is common in dualistic viewpoints, when a "soul" is proposed to answer why/how you think/remember/create/feel.

On the other hand, are you proposing that a robot couldn't have feelings? memories? suspicions? (original, creative) ideas? How could you tell?

I have no first-hand knowledge of anyone's feelings other than my own. I only extrapolate to other, seemingly-the-same-type-as-me beings that they feel. But how can I really know? On the other hand, if metal and plastic are arranged in such a way that it reacts, or even computes, in regard to surrounding conditions, how can I say that I've added awareness? The simple thermostat I've mentioned reacts to the cold by turning on the furnace. How would extra complexity add to its "feeling" of cold? We can see all sorts of feedback loops and imagine others, but that doesn't add up to a feeling.

I am inclined to agree with Charlie.

Complexity of the data is no reason for feeling the data. Someone that has never felt snow can be given all the data in the world, and never feel anything until they actually "feel" snow.

I don't think androids dream of electric sheep or at all, but I do think that real sheep "feel". I can plainly see where feelings would be a great help to survival instincts. For example, sticking your hand on a hot eye of the stove. If you had no feeling, you might look upon the action with an indifference. It's feeling that gives the body an instant jerk to remove the hand. Obviously robots are generally not built with any survival instincts.

Maybe a better question would be, does a robot ever question where it came from? Does a robot ever consider that when it mechanically fails, that it will live on in some form? Does a robot ever wonder if there are robots on other planets? Does a robot ever wonder when the sun will cease to shine? Is a robot curious about anything at all ever? How about a sheep or a cat? Do sheep dream of electric robots?

How do you know that *you* don't have harm-avoiding-mechanisms such as a robot could have?

And, before answering, in fact you do have such mechanisms. Most pain reactions are processed at the spine, and do *not* reach the brain... you notice what you did to avoid pain *after* having done it.

One reason is what need would there be for feelings in a robot. Robots are programmed basically for performing tasks. Most of the times feelings are not enhanced task performers, but rather are inhibiting task performing. Since robots are are fabricated by man, what purpose would they ever have for programming feeling into a machine?

"Since robots are are fabricated by man, what purpose would they ever have for programming feeling into a machine?" There might be people who would want an android to have feelings. Such as a couple who lost a child and couldn't bare to be without him. Haven't you ever seen AI?

I'd rather bet that computers do not feel either excited or prideful or anything else... but then again, how could I know?

But there's no way of defining a feeling except internally to oneself.

There's no logical refutation to solipsism, which posits that no other real persons exist. It is only really a matter of faith that an individuals posit that other individuals are feeling entities like themselves (assuming you all exist out there).

I firmly believe a program could calculate the movements at every moment in history of every particle in the entire universe and not feel a thing ever by having this higher level of brain/programming capacity.

I don't think feelings are the by-product of higher level thinking. I would in fact, lean more towards higher level thinking being a by-product of feelings and emotions. (I presume we are talking about internal feelings in this notion, and not a feeling such as a prick of the finger.) A huge part of the memories that are stored in most of our brains are stored on the basis of a direct emotion/feeling.

Hugo, in the altzheimers' case, correlation is often not causation. Suppose that humans were the strongest animals on earth. One might postulate that feelings are the by-product of strength. I think we can safely rule that postulate as incorrect. As for the "feeling" word, I just presume that we have expanded the sense of touch, to emotional level. I'm going to look pretty silly if I am wrong.

Danny, I did see that movie. It was one of the worst movies I have ever made it completely through! :) I guess there might be some people now or in the future that would desire programming feelings into a robot, so I concede on that point.

Charlie, it is true that people often find a loss of words and can't describe their feelings. An emotional state is often expressed more in body language than in words, and it usually highly contagious. Sometimes that can easily be shared as somewhat of a common ground. Most would probably agree that the emotions and feelings shared and observed were very close, but it is impossible to say for certain that they were the same.

So in that sense, I suppose a robot might spread an emotion by imitating it, but I still don't think it would feel it. Following solipsism, alas I cannot prove it! :)

[Sorry for the long post: it's what happens when you get to a conversation late]

Obviously all this discussion about "feelings" is the same as the classical arguments about consciousness.

One of the important questions in philosophy of mind is the possibility of "zombies". Is it possible for a person to exist who exhibits all the outward qualities of a normal person (yelps in pain when hit, complains when hungry, has long philosophical discussions about his "feelings") without having the internal experience of "consciousness"?

In some ways there do exist such people. "Blindsight" is a very interesting phenomenon as the result of brain damage where people adamantly insist that they have no conscious experience of seeing, yet can react to catch a ball when thrown at them, or can correctly state whether a line shown in front of them is horizontal or vertical. Interestingly they usually claim that they are "guessing", yet they are usually accurate about 95-100% of the time.

A second example is hemi-neglect, where people aren't conscious of the world in one half of their visual field. However, if you show them two houses, one on top of the other, identical except that the left half of one is on fire, and ask them to chose the one they'd prefer to live in, they always point to the one that's not on fire. When asked why, they usually say that the one on fire "feels unsafe", even though they claim not to be able to see the fire. It seems that the brain is interpreting the images correctly, they are being sent to higher decision-making areas of the brain (pre-frontal cortex, probably), yet are somehow "skipping" the area that makes the person say "Ah-ha! That building is on fire!"

So it seems that people _can_ react intelligently in some limited domains without all the conscious experience being present. Yet we _do_ have a conscious experience. Why?

My take would be that consciousness is an evolved phenomenon, necessary for higher processing of the environment and of one's internal states. I think that "lower" animals have less consciousness -- consciousness isn't all-or-nothing but rather is graded. For instance, if an ant smells smoke and runs away (not sure if they do this or not), does he have a conscious experience of smelling smoke? I'd say not really - instead his conscious experience is similar to that of the person with hemi-neglect: a "feeling" that they should go in the other direction. I'd say that those areas in the brain that were activated when the blind-sighted person caught the ball or when the hemi-neglect person chose he safe building were parts that were evolved earlier in our history, and we share them with many animals.

But so if consciousness is an evolved phenomenon, that implies two things: first, creatures who were more conscious than others benefited from an inherent fitness increase that allowed them to do better within their environments than their competitors; second: consciousness ought to have some kind of physical correlate in the brain. The last point is the trickiest, however, since there is clearly no area in the brain that we can point to and say "Ah-ha! That explains consciousness!"