[ExI] common sense

[Stathis Papaioannou]

2010/3/4 Will Steinberg <steinberg.will at gmail.com>:
> On Thu, Mar 4, 2010 at 4:28 AM, Gordon Swobe <gts_2000 at yahoo.com> wrote:
>>
>> --- On Wed, 3/3/10, Stefano Vaj <stefano.vaj at gmail.com> wrote:
>>  If input equals "Given that your performance seems equal to that of a
>> human, do you have a mind?" then output "Yes indeed, I have a mind just like
>> Stefano has a mind."
>
>
> Well then Gordon, I will give credence to your views if you can prove to me
> that you are not such a program.
> Nobody is calling you crazy for thinking this program will not produce a
> mind, except, I guess, your Straw Man.  What you are saying is not being
> argued, and is in fact obvious.
> The fact that you can program a fake mind does not in anyway preclude the
> fact that a real one could be programmed.  I do see how you think the brain
> may have special components which may prevent today's computers from
> emulating it.  This is probably true.  I do not see how you believe that the
> brain will always have some inextricable property allowing it consciousness.
>  The very nature of our universe tells us we can mimic anything in that
> manner.  Maybe a mechanism in the brain prevents any large-scale objects
> from forming a mind because it only takes place at microscopic levels.  In
> this case, we can make our brain out of tiny beer cans.  It's a bit hasty of
> people to say that ANY medium could mimic neurons, because we do not fully
> comprehend the phenomena involved.  However, SOME medium absolutely could.

A digital computer with enough memory can emulate (perfectly simulate)
anything that can be described algorithmically; that is, described in
a finite number of instructions. This is the Church-Turing thesis, and
generally held to be true. What is less certain is whether everything
in the universe can be described algorithmically. Everything we know
of so far can, but that does not mean everything can. Maybe the as yet
undiscovered theory of quantum gravity is non-algorithmic, and maybe
our brains utilise quantum gravity effects in order to do what they
do. It's far-fetched - there is in fact no evidence for it - but it's
not impossible. If it's the case, it means that the brain is "not
computable". A computer program might be able to display some of the
behaviour that a human displays, but there will come a point where it
fails: it may be unable to match a human in using natural language,
for example, and therefore would reveal itself as the computer in a
Turing test.

Gordon does not have a problem with the emulation, but he has a
problem with the associated consciousness. It isn't immediately
obvious, but there are several lines of reasoning which lead to the
conclusion that an emulation *would* necessarily have the associated
consciousness of the brain being emulated. These arguments are
unnecessary, but still valid, if one assumes from the start that
consciousness is just an epiphenomenon supervening on intelligence.


-- 
Stathis Papaioannou