[ExI] Is the brain a digital computer?

[Stathis Papaioannou]

On 1 March 2010 11:42, Gordon Swobe <gts_2000 at yahoo.com> wrote:
> --- On Sat, 2/27/10, Stathis Papaioannou <stathisp at gmail.com> wrote:
>
>> It's true that there is no requirement for an AI to follow
>> brain architecture, but I am considering the special case where
>> it does.
>
> After mulling over your thought experiment for oh so many weeks, it seems to me apparent that weak AI *must* deviate from organic structure. So I don't know what you mean by "special case", unless you've just begged the question of strong AI.

What I mean by an AI following brain architecture is that it is made
up of components that exactly duplicate the I/O behaviour of the
corresponding biological component. There is no requirement that the
artificial component itself follow the structure of the biological one
internally, although as a practical matter it might be easier if it
does. I have mainly been talking about these artificial components
being neurons but they could just as easily be subneuronal or
multineuronal structures. For example, we could remove a sphere of
brain tissue one centimetre in diameter and replace it with a sphere
of similar size comprised of machinery which interacts with the cut
surface exactly as the replaced tissue would have. Within the implant
there would be a computer running a program based on information from
scanning or dissecting the original tissue, controlling sensors and
effectors at the surface. Essentially, the implant is a little
philosophical zombie designed to fool the surrounding brain into
believing that it is normal brain tissue, in the same way as a fully
grown philosophical zombie is designed to fool humans into thinking
that it is a conscious human.

You agree that the fully grown zombie is possible but you have
expressed doubts about smaller and simpler versions of it. If you have
presented an argument to support this view I have missed it.


-- 
Stathis Papaioannou