; brightskeptic @ Nov 30 2007, 12:17

ColonelZen @ Nov 30 2007, 01:51 Yes an AI(?) which sucessfully and completely emulates a parrot is conscous ... at the rather unimpressive level of a parrot.

But do you see a difference between a computer program that emulates a parrot and an AI that has parrot consciousness?

No I don't. Other than the obvious that the programmers will have extraordinarily more control over a program, and ability to change it, than they do over a parrot.

brightskeptic @ Nov 30 2007, 12:17
I don't think anyone here has a problem with creating an AI - whether silicon, metal and plastic or biological greymatter. No-one sees anything magical about consciousness, despite each side trying to accuse the other side of this. I think we all find consciousness special - certainly our own consciousness.

The problem is that I don't see what you mean by *special* if you don't mean something almost similar to "magic".

My consciousness is "special" only in the way that my truck or cellphone is "special". It's mine and I'm more familiar and comfortable with it than others can be. The only real "special" about my "consciousness" is that no one else (at least given current technology, and its epistemologically arguable if ever) can experience it exactly the way I do. But the same can be said of my "driving", or "sleeping" or any other gerund form of verb denoting an action or mental experience applicable to me.

brightskeptic @ Nov 30 2007, 12:17
I don't know if your definition of consciousness is right (Searle has one that fits better with my experience / intuition, although I don't know if that one is right either). What I am pretty sure of is that your definition can't be used to distinguish between simulated consciousness and real consciousness, so on a practical level it is not sufficient.

My whole point since I began participating on this thread is that there is no difference between "simulated" or AI consciousness and "real" or grey goo consciousness of similar functionality.

The first point is that "consciousness" is an abstract idea. I'm not talking about consciousness itself, but the idea of consciousness. It is purely and singly a "thing" which happens in the mind, an artifact. To communicate we need to be able specify that idea well enough to know that we're talking about the same thing. It is unknown and possibly unknowable exactly how similar or different our individual "consciousnesses" are, so to be sure we're talking about the same thing, we need a definition which makes no reference to our subjective knowledge. At this level, the level of labeling the idea, "consciousness" is no more magical or distinctive than "addition".

Once we can agree on an objective, observable framework for discussing consciousness (and again, please take my definition as a "first swing" by someone generally ignorant - I'd love to hear on objective functional definition by someone who might know what they're talking about) and gain some familiarity and experience using it, THEN we might profitably begin mapping our objective, observable behaviors to our subjective mental states and maybe, just maybe begin learning more about how we tick on the inside. But wallowing around discussing consciousness without having a baseline grounded in repeatable, observable real world phenomena gets us nowhere - reread the larger mass of this topic in forum for examples in wealth.

Now (once we agree) we have a definition of an immaterial process (of course it transpires within matter, but it's a closed box and not germane to the consequential behavior resulting from that process) there is no longer a difference between "simulation" and "real". The behavior being described by our definition doesn't change - it is still that behavior. The abstraction of the idea and given the label "consciousness" is an abstraction and description of that behavior, not the processes bringing it about.

Again, there is no difference or distinction between real or "simulated", goo or silicon "consciousness". Because "consciousness" is a label, an abstract idea, purely and solely a set of observed descriptive characteristics (also ideas) with well defined relations between them ... else we don't know what we're talking about.

Now certainly the internal "experience" generating those externally observable behaviors will be vastly different between silicon and goo, just as I'm sure they're vastly different between me and my cat. Yes I think it highly likely that other humans and I share a lot of similarities in that internal experience, probably far more that between my cat and myself. And you and I probably have far more similar internal experiences than we will with any AI. But on the other hand I *know* I think *far* differently about most things than most people around me. So I don't really know how similar your and my internal experiences are.

But none of the similarities or differences obviates the need for an objective and purely functional definition. And that definition, being itself an abstraction, applies to the objective behaviors, if it is truly objective and functional, need not reference the means by which the behavior is generated.

Ergo there is no "difference" between "simulated" and "real" consciousness unless you vacuously chose to define it so. The difference between "real" and "simulated" consciousness, if/when/once we get to a sufficiently clear and objective definition is as pointless as the difference between "real" and "simulated" addition.