[ExI] Wernicke's aphasia and the CRA.

[Stathis Papaioannou]

2009/12/13 Gordon Swobe <gts_2000 at yahoo.com>:
> --- On Sat, 12/12/09, Stathis Papaioannou <stathisp at gmail.com> wrote:
>
>> The replacement would have to involve artificial neurons
>> that are *functionally equivalent*.
>
> I.e., functionalism.

When I say that the artificial neurons are "functionally equivalent" I
am referring to their externally observable behaviour. Functionalism
is the theory that the mind would follow if the externally observable
behaviour is taken care of, and is what is at issue here.

>> To properly model a brain you may need go down all the way to
>> the molecular level, which would of course require...
>
> Now here you and Searle may almost agree. To create an artificial brain modeled "down all the way to the molecular level", we would need essentially to create a real brain.

It would behave like a biological brain and it would have the
consciousness of a biological brain, but it need not have any
biological components unless it turns out that these components cannot
be modeled on a computer. This is Roger Penrose's position: he claims
that computers will never be able to display human-like intelligence
because the brain utilises non-computable physics. In other words, he
claims that weak AI is impossible. This position is consistent, but
there is no evidence of non-computable physics in the brain or
anywhere else. Searle, on the other hand, claims that weak AI is
possible but strong AI impossible, which is inconsistent. The neural
replacement experiment I described shows why this is so, and you
haven't addressed it.


-- 
Stathis Papaioannou