[Reader-list] alternative Software

[Are Flagan]

On 11/30/02 13:46, "Supreet Sethi" <supreet at sdf.lonestar.org> wrote:

> Probably what I am understanding of your article or essay is wrong so I
> need clarification on certain words and constructs
> 1) What does Alternative mean here

I guess that's what I am partly trying to figure out. If we are so keen on
open source and the development of independent standards, surely these would
by definition be approaching this "alternative," at least in their intent.
The essay asks a tentative "what if..."


> 2) How does it offer a critique of universal principles laid down by
> hardware

I would say that it only points to a critique of universality (and hints at
its effects).

> 3) Is this essay making a claim that the mathematical model on which
> computeres and many electronic devices is wrong. Finite autmata as a model
> has been used over and over again and is considered a thoroughly
> understood subject

It is not really making a binary value judgment explicitly (it is not a
computer, in other words). One could perhaps also say, with reference to how
software/hardware (finite automata) is used and distributed, and what
economies it participates in, that it is equally misunderstood.

                   
> Is there any sort of study on which problems/solution scenerios are beyond
> computational margins.
> Could you point me to the texts which cover this topic well.
>

The straight answer would be that anything that is not either true or false
and can be processed according to the operators AND, OR, NOT would easily
fall beyond the present computational margins. The links in Mark Crosby's
post contain, at first glance, some interesting overviews of various
questions posed... 
  
> It would be even better if you could point to a live problem/solution
> sceneario where hardware/software constraints could be made glaringly
> visible.
>

This is the crux and contention: the conjuncture of
hardware/software/logic/math and language/culture/society effectively seek
to preclude such visibility. It seamlessly integrates its program into a
machine operation and thereby mechanizes and naturalizes its appearance.
>  
> As you have already stated in your easy about neural network, I would like
> to push the discussion around that topic. Neural network approach to
> computing does not base its treatment of data not on the basis of preset
> algorithm but rely on a learning of certain types of data and computing on
> basis of shift of synaptic weights.
>  
> Minsky wrote a doctorate thesis at princeton which sort of dealt with
> topic of neural networks.
> In 1967 Minsky wrote a book called Computation: Finite and Infinite
> Machines which put "neural networks" in context of automata theory and the
> theory of computation
>  
> Would like hear more on this

I see it as a super vast field with many angles and possibilities. If
software/hardware is indeed an incarnation, an embodiment, of mind, is it
possible for it to think for itself? Immediately we return to the core split
in the philosophy of mind (reflected in the Cartesian subject): are these
thoughts really mine or do they belong to discursive differences, outside
me? Can a neural net be other than the learning mechanisms imposed upon it,
the instructions already given? Ada Lovelace remarked about the very first
"computer" that it could do nothing but what we told it. Turing seemed to
believe that his machines could inhabit their own intelligence, a character
specific to its algebraic make up. Our understanding and application of
"software" may ultimately be contingent upon exactly how such epic
conundrums are resolved and/or resisted.

-af