[ExI] Digital Consciousness

[Stathis Papaioannou]

On Mon, May 6, 2013 at 5:41 PM, Gordon <gts_2000 at yahoo.com <javascript:;>>
wrote:
> Stathis Papaioannou <stathisp at gmail.com <javascript:;>> wrote:
>
>>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.
>
> Just third-person descriptions in the language of physics.
>
>> In the process, qualia may somehow be produced, but qualia are neither
>> inputs nor outputs so are
>> excluded from this part of the analysis.
>
> Yes, qualia "may somehow be produced." Actually, I think are produced.
lol.
> How does that happen?
>
> Consider for a moment that the world might not be fully understandable in
> the third-person language of physics. Perhaps there is something that we
> might call first-person ontology. As science-minded people, we want to
> describe everything in the world in the objective language of physics. But
> perhaps the world is both subjective and objective.
>
> Toothaches are not the same thing as the physics that describe them. They
> really hurt!

A motor neuron controlling your vocal cords will only fire if the neurons
connected to it fire, and those neurons will in turn only fire if the
upstream neurons fire. There is to be sure a very complex network of
neurons between the sensory organ and the motor neuron, but each
component of the network follows a relatively simple set of rules. The
neurons of the NCC also follow these rules, and it is the timing and
amplitude of their output (action potential propagating down the axon)
which determines if the downstream neurons fire. So if these NCC neurons
are replaced with artificial neurons that replicate their pattern of firing
in response to upstream stimuli, the downstream neurons must respond in the
same way and the subject must behave in the same way as with the original
brain. Is there any part if this that you don't agree with?

>> 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.
>
> No, I wrote that qualia might be inputs or outputs, or something else
> entirely.
>
> I notice that you did not respond to my criticism of functionalism and
> multiple realizability, which I consider very important. You quoted it
below
> but did not answer.
>
>>> 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.

My conviction that functionalism is correct does not come from
considerations like yours. I agree that it is not immediately obvious that
a computer could be conscious.  But if a device (computer or
otherwise) could reproduce the observable behaviour of a neuron without the
consciousness that would lead to absurdity, as I have explained repeatedly,
and this leads to the conclusion that it is impossible to make such a
device.


--
Stathis Papaioannou


-- 
Stathis Papaioannou
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20130507/b314423a/attachment.html>