http://www.the-brights.net/forums/forum/index.php?s=&showtopic=7643&view=findpost&p=124335
IR brightskeptic ...
My definition creates a low bar to "consciousness" ... but imputes at least three quantitative (though somewhat loosely) dimensions. I think there is very much a scale of consciousness.
I'm not well familiar with the behaviors of parrots or lizards but IMO *any* learning capacity implies some low level consciousness. What it (implies we should) looks for as (at least somewhat) objectively distinct behaviors we can reliably correlate to a mental state (not too difficult as any need or desire can be qualified as such) and then look for variants of that behavior based on environmental circumstances which do not immediately result in attainment of that state (satisfaction of that desire) but change the odds.
Thus my cat "knows" (predicts) that meowing at people improves her odds of being fed.
My brother's dog "knows" (predicts) that bringing me it's ball improves his chances of being taken out for play.
My brother's dog, a husky/shepard mix is very instructive. His behaviors appear to include both exuberance and shame in appropriate conditions. I don't *think* I'm anthropomorphizing too much in saying so. If these are correct assessments of his mental state it implies he has some sense of "self" as well. And if you think back to my peroration on the purposes of "I" in mental life, it's not at all surprising that an animal descended from hunting pack animals with a social heirarchy would have a sense of self.
Yes an AI(?) which sucessfully and completely emulates a parrot is conscous ... at the rather unimpressive level of a parrot.
The lesson here, is that the idea of consciousness is too broad and before my definition (or hopefully someone can find a better definition by somebody who's skilled and being paid for spending real time on thinking about such things) too vague to be of much value in qualifying what steps need to occur to actually build an AI.
The second lesson but primary for this thread in the nature of a rebuttal to Searle's proposition, is that I don't think there's anything special or "magic" about consciousness. It is simply a named property observed in goo and gristle thinking machines beyond a minimal competency, and it can be characterized, described, and when required implemented in silicon, metal and plastic
-- TWZ
|