[ExI] Digital Consciousness

Stathis Papaioannou stathisp at gmail.com
Mon May 6 06:36:00 UTC 2013

On Mon, May 6, 2013 at 3:55 PM, Gordon <gts_2000 at yahoo.com> wrote:
> Stathis Papaioannou <stathisp at gmail.com> wrote:
>> If an artificial NCC can replicate the I/O behaviour of the biological
>> NCC then the person must behave normally, since all the impulses to
>> the muscles will occur normally. Do you see how this must be so? Do
>> you see any problem with the idea that a person behaves normally while
>> a part of his conscious mind (namely, that part due to the NCC
>> replaced with artificial components) is missing?
> I'm not even sure what "replicate the I/O behavior" means when we're
> discussing NCC neurons and qualia. Are qualia inputs? Or are they outputs?
> When someone shines a light in my eye, it seems like an input. On the other
> hand, I seem to be representing that light to myself, in which case the
> quale is more like an output. Perhaps qualia are not inputs or outputs, but
> something else entirely. I don't know. So, if our purpose here is to
> investigate whether digital neurons can replace organic neurons in the NCC
> and create conscious experience, which I think plays a role in driving
> behavior, I would not know how to begin to answer your question as I don't
> know how to interpret your first sentence. Do you?

The inputs and outputs I'm talking about are action potentials which
trigger neurotransmitter release at synapses. The neurons in the NCC
receive inputs from other neurons that connect with them and send
output to other neurons via their axon. In the process, qualia may
somehow be produced, but qualia are neither inputs nor outputs so are
excluded from this part of the analysis. If the artificial NCC neuron
reproduces the outputs given certain inputs, then all the downstream
neurons to which it connects behave normally. This is irrespective of
any qualia it may or may not have, since as you admitted qualia are
not outputs. So the person with the artificial NCC neurons will behave
exactly the same as a normal person - by definition, since the
artificial neurons reproduce the I/O behaviour perfectly. Do you agree
with this?

> Here is my more general issue with functionalism and multiple realizability
> as they relate to strong AI on digital computers:
> I have a ceiling fan in my home. The blades are made of wood. I've seen
> other ceiling fans with metal blades, and still others with stained-glass
> blades. They all function as fans, and so along with functionalists I'm
> happy to call them all fans. I've also seen hammers made of iron and others
> made of steel. Again, both hammers. Fans are realizable in wood, metal and
> glass; hammers are realizable in iron and steel. Only the functions are
> important. The substrates make no difference. All well and good.
> But things get muddled when we begin to talk about the supposed multiple
> realizability of brains in both organic materials and software/hardware
> platforms. A digital computer is not in the same class of things as fans and
> hammers. With fans and hammers, we are looking purely at the direct physical
> effects of one substrate on another. Fans are defined as those physical
> things that directly circulate physical air, hammers are defined as those
> physical things that directly drive physical nails. A computer program, by
> contrast, is defined by functionalists as anything that admits of abstract
> 1's and 0's (or ons and offs, however you want to think of it) and the brain
> (like everything else in the world) does admit to such an interpretation.
> This is why they say brains are multiply realized on computers. But those
> syntactical abstractions are not actually *intrinsic* to the physics of the
> organic brain. Functionalists and computationalists merely assign them to
> the physics.
> Gordon
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat

Stathis Papaioannou

More information about the extropy-chat mailing list