From: J. R. Molloy (jr@shasta.com)
Date: Wed Jan 16 2002 - 00:53:58 MST
From: "James Rogers" <jamesr@best.com>
> My point was that all programs expressible on a parallel system are
> expressible on a single processor system.
Since no one has tried "all programs" (all programs haven't been devised yet)
on a single processor system, we cannot yet know if all programs (including
the human brain's parrallel system expressed programs) can be expressed on a
single processor system. In fact, the human brain's programs have not been
fully expressed in *any* system, whether single or parallel processor type.
So, we can't know (yet) if the human brain's programs require a parallel
system.
> Magic doesn't happen when n>>1,
> any more than when n=2 (where "n" is the number of processors) compared to
> when you only have a single processor.
So it would seem. Nevertheless, the proof is in the pudding, and until a
single processing system successfully expresses the entire gamut of the human
brain's programs, there is no empirical evidence to support the contention
that it can do so.
> If you can't do sentience on a serial processor, massive parallelism won't
> help you in a meaningful way, other than potentially giving a bit more
> processing power of a largely uninteresting sort.
In his book, _Afterthought, The Computer Challenge to Human Intelligence_,
James Bailey disagrees with your statement:
"More interesting than random numbers, however, are the computational
potentials of entities like neurons and chromosomes. Clearly, these mechanisms
have accomplished much over time. Large arrays of neurons, for example, are
able to recognize patterns in ways that no geometric or algebraic formulations
can. Unfortunately, it can take them years to learn how. The same is true of
the genetic recombination of chromosomes.
[See also: Scientists build tiny computer from DNA
http://in.news.yahoo.com/011121/107/199x5.html ]
Wondrous results have emerged, but over millions of years. Neither were viable
candidates for computing before computing became electronic. Now, more and
more scientists are heeding the advice of the physicist Stanislaw Ulam: "Ask
not what mathematics can do for biology. Ask what biology can do for
mathematics."
"Add as many mail coaches as you please, you will never get a railroad by so
doing."
--Schumpeter
I think we will get human-level pattern recognition in machine intelligence
via a partnership of sequential human brains and autonomous parallel
electronic circuits, because discursive symbolism alone does not equate to
human-competitive intellectual activity.
--- --- --- --- ---
Useless hypotheses, etc.:
consciousness, phlogiston, philosophy, vitalism, mind, free will, qualia,
analog computing, cultural relativism, GAC, Cyc, Eliza, cryonics, individual
uniqueness, ego, human values, scientific relinquishment, malevolent AI,
non-sensory experience, SETI
We move into a better future in proportion as the scientific method
accurately identifies incorrect thinking.
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:11:42 MST