X-Message-Number: 25988
Date: Fri, 8 Apr 2005 11:00:26 -0400
From: Thomas Donaldson <>
Subject: CryoNet #25982 - #25984

To Yvan and Daniel Crevier, re parallel computers:

Yes, Yvan, I am perfectly well aware of FPGA based neural networks
with over 1000 neurons on a single chip. Etc and Etc. If you can
make one with several million so-called neurons on the chip you
might get a bit closer. (To call the units on such a chip "neurons"
is absurd to begin with).

Models of brains can be very useful, but as I ended my message,
they should not be confused with real brains. In fact, your tiny
FPGA chips just might imitate a single neuron, which gets you
into "time-sharing" (though parallel computers DO NOT TIME SHARE
at all: all their processors are working at once. That is the 
point of parallelism).

It's not that I insist that a computer work like a brain. I do
insist that one capable of BEING a brain have some features 
allowing it to do what brains do. Your proposals fail on that
point, and fail utterly. 

PERIASTRON has many articles which discuss how BRAINS work, as 
much as we know of that at present. If you wish to subscribe,
you'll be welcome. 

             Best wishes and long long life to all,

                 Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=25988