Re: [-empyre-] free will and determinism

["tobias c. van Veen" <tobias@techno.ca>]

> If I understand the objections some have stated on the list to the notion
> that human thought processes could be algorithmic in their nature, without
> exception, it seems to be that there's a feeling this denies free will.


I think Steven Shaviro in his new book, _Connected_ (Minnesota UP 2003),
would give all the brain-algorithmic processes a run for the money. A
questioning of Algorithmic Truth does not necessarily have anything to do
with 'free will' (that being a rather weird concept today, in some ways)--it
has more to do with the deep-seated intuition that structures of time/space
are far more complex than equivalents, and that there is a certain prophetic
.. closure .. to claiming that The Answer Is This (brain=algorithm, bling!).
Quote:


- On Ray Kurzweil's belief that the brain can be copied/downloaded into a
computer:

"Kurzweil seems to believe that we can do this without worrying about the
underlying hardware of the brain... But contra Kurzweil... one could not
"literally copy it [the brain]" in the first place unless one paid attention
to the "elaborate structure" underlying it all. Kurzweil does entertain the
idea that a downloaded mind will need some sort of new body, if only because
a "disembodied mind will quickly get depressed." But he fails to grasp the
reciprocal correlation between the mind and the body, or software and
hardware, or the virtual and the actual." (114)

Yes, that nagging, leaky body. 'Tis a problem. Where be the body in math?

- On brains = Turing Machines:

"[Shaviro explains that multitasking is actually 'rapid-enough' linear
switching that creates the simulation of multiple processing]. That is to
say, a computer's operations are *actually* linear and sequential, but
*virtually* multiple and synchronous. Now, this is precisely the opposite of
what happens in human brains. Our underlying neural activity seems to be
massively parallel. But nearly all of this activity is unconscious. Human
consciousness, on the other hand, is experienced serially. Language, too, is
a linear phenomenon, though its underlying structure is synchronic. Nicholas
Humphrey suggests, therefore, that consciousness is something like a serial
interface to the underlying parallel processing that is fundamentally
constitutes mental activity. Consciousness is not thought itself, but an
easily accessible simulation of thought. The brain's activities are
*actually* multiple and synchronous, but *virtually*, they are linear and
sequential. Or to put the same point in slightly different terms: the brain
is not a Turing Machine. It does not work by means of algorithmic
calculations. Rather, it operates according to entirely different
principles, ones that we do not currently understand. Nonetheless, the
calculations of digital computing--and consciousness--can often arrive at
the same conclusions as the unconscious brain; that is to say, many brain
processes can at least be simulated by Turing Machines. Even brain
parallelism can be simulated to some extent through neural nets or multiple
processors. The questions still remains, though, as to whether or not all
the brain's operations can be simulated by Turing Machines such as digital
computers and conscious minds. This is the real issue at stake in current
arguments about the powers and limitations of artificial intelligence,
although it is often overlooked by advocates on both sides of the debate.
Partisans of 'strong AI', such as Daniel Dennett, seem to assume, without
evidence, not only that algorithmic computation can simulate brain activity,
but that the brain actually operates this way. Opponents of 'strong AI',
such as Roger Penrose and John Searle, rightly reject this assumption, but
then they make the equally dubious claim that consciousness is a uniquely
privileged aspect of mental activity that cannot be produced by
computational means. Against both camps, I want to suggest two things.
First, that the question of whether the brain actually is a Turing Machine
and the question of whether the brain can be simulated by a Turing Machine
are entirely separate issues. Second, if there does turn out to be some sort
of brain activity that cannot be simulated by a computer, then this activity
will turn out to be unconscious as well, and indeed radically inaccessible
to consciousness."


Thus, the point here is not 'free will' .. its our own unknowable: that an
algorithm as model cannot model that which we do not already know,
ourselves, as our own brainbodies: radically inaccessibility, an archive of
ourselves that cannot be copied due to systems that exceed 'static copy
archive', a photoflash that could be cut to capture 'essence of human
thinkthought' would not 'copy' PROCESS -- the point being TIME -- the
difference of time in (attempting to) place thought exceeds our
understanding of the algorithm, exceeds the algorithm.

In any case, I only wanted to draw attention to this text for its wonderful
enumeration of a number of arguments, all in Shaviro's knowledgeable & witty
style, written in these 'bytes'--contemporary aphorisms--that interweave a
position that I think offers more than enough significant challenges to the
formulation that algorithms = brain (and goes on to develop this through
sci-fi, music, strong arguments against libertarianism, and so forth) as if
a sudden development in technology and mathematics became the end of
brainperson history (another end of history), as if this moment in
technology became equivalent to brain (minus-body) in an incredible negation
of race, colour, body, and gender (a wonderful [white] myth of math erasing
all difference), as if we've discovered the brainbaby & the bathwaterbrain
together, and yes, they are indeed an algorithm, as if 1. we are
representational structures [that the history of shifting histories of
representation be negated as representation today IS representation ALWAYS
due to Truth of Maths] and 2. we represent algorithmically [as Math Today is
True Math Unshifting Unchanging -- even the shifting changing algorithm math
as unshifting truth of shifting]. Everything on the contrary --the very
mechanisms of 'uploading/downloading' -- structures of representation --
which are brought to unfolding madness when brought into contact with the
algorithm, and vice-versa, meaning that algorithms and representation form
only one particular model of a temporality which is a modelling too, a
spatialization in fact (as Bergson noted: the reduction of multiple time to
spatial understanding), of some kind of present which we are far from
forming an 'understanding'--whatever that would mean, here .. -- the present
being a fold: and our ability to simulate this fold may have become
accelerated, but the fold IS not algorithm.


 - tobias