[ExI] The symbol grounding problem in strong AI

Gordon Swobe gts_2000 at yahoo.com
Mon Jan 4 20:09:50 UTC 2010


--- On Mon, 1/4/10, Stathis Papaioannou <stathisp at gmail.com> wrote:

>> I don't see how that follows, nor do I posit any guess
>> as to their actual experiences (especially for Cram, who may
>> have none by the time the doctors finish with him).
> 
> You alter Cram's consciousness, but it has no effect on his
> behaviour. 

Yes, the initial operation most almost certainly affect his behavior before he leaves the hospital, causing him do and say strange things. His surgeon corrects those symptoms and side-effects with more programming/replacements with p-neurons until he can call his patient cured. But unbeknownst to the surgeon, who according to your experimental set up has no understanding of philosophy, the cured patient has no subjective experience. 

When I say I reject epiphenonemalism, I mean that I reject it as an explanation of normal human consciousness. Because I think consciousness plays a role in human behavior, I think Sam will fail his TT unless his m-neurons give him real consciousness. And unlike Cram's doctors, Sam's have no way to correct any side-effects if the m-neurons don't work as advertised.

 
>> We needn't create artificial neurons to study the NCC.
>> We need to identify possible target areas and then to test
>> our theories with technology that switches it off and on in
>> a live patient. 

> But how would we ever distinguish the NCC from something
> else that just had an effect on general neural function? 
> If hypoxia causes loss of consciousness, that doesn't mean that 
> the NCC is oxygen.

We know ahead of time that the presence of oxygen will play a critical role. 

Let us say we think neurons in brain region A play the key role in consciousness. If we do not shut off the supply of oxygen but instead shut off the supply of XYZ to region A, and the patient loses consciousness, we then have reason to say that oxygen, XYZ and the neurons in region A play important roles in consciousness.  We then test many similar hypotheses with many similar experiments until we have a complete working hypothesis to explain the NCC.

At the end of our research project we should have a reasonable theory that explains why George Foreman fell to the mat and could not get up after Muhammad Ali clobbered him in the 8th round in the Rumble in the Jungle. That happened over 30 years ago, and still nobody knows. 

-gts




      



More information about the extropy-chat mailing list