Damien Broderick, <email@example.com>, writes:
> My 2 A-cents, from THE LKAST MORTAL GENERATION, chap. 4:
How much is that in real money?
> One can imagine a world of zombies, [philosopher David Chalmers] asserts,
> just like ours but *lacking* consciousness. The brains of these zombies
> mimic ours precisely, but there's no light on inside. They act and speak
> and laugh and `love', but are mere automatons. Since this nightmare is
> logically possible, Chalmers says, consciousness must be something over and
> above mere neural structure in action.
Eliezer's unconscious aliens may escape zombiehood. They do not precisely mimic our own minds - they have no qualia but they don't claim to have them either, unlike zombies. If you ask them about the strange mystery of consciousness, that peculiar quality of direct perception of reality and self, they stare at you blankly.
It's not completely clear to me that this is a realistic model, even assuming that consciousness and qualia are noncomputational. As Wei suggested, maybe even unconscious aliens could have mental structures which caused them to act as though they had qualia. Maybe for example self-representations are so different from representations of other objects that they provoke a similar sense of mystery. This seemed to be the idea that Hofstadter was pursuing, that self representation would turn out to be the key to consciousness.
> What is it, then? Information, he concludes. Not an answer to give the
> faithful any comfort, admittedly. Chalmers believes a complex artificial
> intelligence system would be conscious. So he is not proposing to
> reinstate an immaterial soul. He puts it neatly: `Experience is
> information from the inside; physics is information from the outside.'
> Inside what, though? Inside the mind, with the qualia. But that leaves us
> where we came in.
I need to make time to read Chalmers' book. I saw him speak last year and it seemed that he was adopting a dualist stance, that there had to be more to consciousness than computation. It was a pretty technical argument, though.
> Besides, I would argue that we cannot truly imagine a
> zombie world, any more than we can truly imagine a world exactly like ours,
> full of jittering molecules but without heat. Stewart and Cohen provide an
> amusing analogy to support this suspicion. Imagine a *zombike*, they
> suggest, `which is *exactly like a bicycle in every way* except that it
> does not move when the pedals are pushed. Oh, mystic miracle of ineffable
> immateriality, the source of motion in a bicycle is not anything
> physical!'29 The zombie analogy, at root, is no more persuasive nor even
This was the deep technical argument where I was unable to follow Chalmers. He agrees that the zombie world is impossible, but is it logically impossible, physically impossible, or metaphysically impossible (whatever that means!). He did a detailed analysis of classes of impossibility and concluded that zombies are logically possible but (I think) metaphysically impossible. From this he argues that computationalism can't be the explanation of consciousness.
He seemed to lean towards what he called pan-proto-psychism, the notion that the world is full of specks of proto-consciousness. This is an actual physical phenomenon which, when put together in the right way, produces consciousness. Perhaps this is consistent with Eliezer's noncomputational views.