Russians create artificial human brain
bbeizer at sprintmail.com
Mon Apr 16 13:49:56 EST 2001
"Charles R Martin" <crmartin at indra.com> wrote in message
news:m3wv8kkaic.fsf at localhost.localdomain...
> wroot <wroot at my-deja.com> writes:
> > David Ehrens wrote:
> > It could be a hoax. You won't find anything on Vitaly Valtsev on the
> > Internet
Probably not a hoax in the ordinary sense. Whoever posted it, probably
believes it. During the 50's, 60's, and 70's, we were regularly deluged by
one "breakthrough announcement" after another from the Soviet Union. Some,
if not all of these, were probably written by the KGB and/or the propaganda
ministry. In 1960 or so, I remember an announcement of a "breakthrough" in
ferrite memories. The Russians claimed to have developed a memory with a
capacity of 2 exp(100) bits, which is 1.5 exp(19) gigagbytes. They were
always frothing at the mouth like that and I guess that they still do in
their publicity releases.
Soviet scientists were under terrific pressure to publish something
that beat the West. If you audio transistors are the size of our power
transistors and your power transistors the size of a hockey puck, and if you
have to go to 47st Computer Discounts in NYC and buy Atari games so that you
can cannibalize them for parts to use in your missile guidance system --
then indeed you have to stretch things a bit -- it doesn't matter if the
West believes it or not -- so long as the KGB and the bosses up top do.
Science in totalitarian states is always like that and habits of three
generations die hard.
Excepting many areas of mathematics, control theory, and algorithmic
complexity theory, there is no area of Soviet computer science and
technology that was even remotely on par with the US. Their most advanced
computer was a shameless clone of the IBM 360 series computer. In software,
for the most part, their methods (especially in software engineering,
testing, and quality assurance) were stuck in the early 60's. No concept of
multi-person programming, documentation, configuration control, testing,
etc. I take it back, they were probably world-class leaders in IC and
software reverse engineering.
> > I'm still curious if it's true that human brain neurons are substatially
> > different from the rest of the neurons, as far as modeling is concerned.
People who still pursue AI from the point of view emulating biological
processes are so far behind the curve as to be in another galaxy. Why would
you want to create an "artificial brain" whose components were 90% water
and had to be assembled by unskilled labor?
> I could easily be out of date, but so far as I know, the brain's
> neurons are _not_ significantly different in terms of micro-anatomy
> and the way they generate action potentials and all.
Besides which, the action is not in the neuron, but at the synapses -- and
some the latest thinking is that memory storage may be at the molecular
level -- e.g., on DNA or RNA.
Boris Beizer Ph.D. Seminars and Consulting
1232 Glenbrook Road on Software Testing and
Huntingdon Valley, PA 19006 Quality Assurance
Email direct: bbeizer at sprintmail.com
Email (Forwarded): bbeizer at acm.org, bbeizer at ieee.org
More information about the Neur-sci