Toward a Science of Consciousness 1998
modlin at concentric.net
modlin at concentric.net
Fri Apr 24 12:22:11 EST 1998
In <6hqde2$pki at ux.cs.niu.edu>, rickert at cs.niu.edu (Neil Rickert) writes:
>modlin at concentric.net writes:
>
>>But if you really mean to say that the architecture used for the
>>computing itself makes a difference to what can be computed, given the
>>necessary input and ignoring performance... then I respectfully suggest
>>that's incorrect.
>
>The important points that you are missing are:
>
> We are not given the necessary output. We have to fetch our
> own input, and make our own decisions as to what input to
> use.
>
> We have to make do with whatever performance we have. It it
> took a year to make the decision whether to eat that morsel
> of food, we should soon starve to death.
I'm not missing those points. I'm explicitly talking about computation,
to which they are irrelevant. They may have a lot to do with whether a
computation is useful or effective, and we need to consider them in
talking about what is needed for consciousness and intelligence... but
they have NOTHING at all to do with whether different architectures can
compute different functions.
Computation is transforming data according to some functional
relationship. What we call a computing architecture is a set of
primitive functions plus some means of combining them to make up other
functions not defined as primitives in the architecture. It turns out
that any of many very simple sets of primitives is enough to allow
combinations implementing any other computable function. We call an
architecture capable of at least such a set of primitives "Turing
complete", and any such machine can compute any function any other such
machine can compute, given enough resources. This is analogous to the
notion of a "Boolean complete" set of primitive boolean operators such
as (AND OR NOT) or (NAND). Given a boolean-complete set of boolean
operators you can generate all possible boolean functions, and given a
Turing-complete set of computing primitives, you can compute all
possible computable functions.
I'm finding it frustrating that you keep posting that you disagree, when
I know that you understand this point because you've made it clearly
yourself, several times. Why disagree when I say the same thing?
Bill Modlin
More information about the Neur-sci
mailing list