Newsgroups: comp.robotics
Path: brunix!cat.cis.Brown.EDU!noc.near.net!paperboy.wellfleet.com!news-feed-1.peachnet.edu!news.duke.edu!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!swidir.switch.ch!newsfeed.ACO.net!Austria.EU.net!EU.net!uunet!hobbes!earth.armory.com!rstevew
From: rstevew@armory.com (Richard Steven Walz)
Subject: Re: 100 Billion Neurons Nonsense
Organization: The Armory
Date: Tue, 19 Jul 1994 18:11:08 GMT
Message-ID: <Ct796M.rr@armory.com>
References: <38qJPc1w165w@sfrsa.com>
Sender: news@armory.com (Usenet News)
Nntp-Posting-Host: deeptht.armory.com
Lines: 38

In article <38qJPc1w165w@sfrsa.com>, bsmall <bsmall@sfrsa.com> wrote:
>I've been reading over Steve Walz's long stream-of-conscience post 
>and have these comments. First he mentions multiprocessed funtions 
>which I hope doesn't mean that a sequential machine can't do 
>thinking. I really believe that it is a moot point to worry about 
>how the thinking is to be done (parallel processors and all that).
> 
>Next I'd have to agree with him that 95% of our brain is 
>evolutionary junk and although we may need this evolutionary 
>process to create thinking, we may be able to program the 
>machines after they evolved into leaner machines. That's one
>nice thing about working with silicon instead of neural tissue.
> 
>I do agree with him that "awareness" (yikes philosophy shows
>its ugly head again) may be attainable with really simple 
>algorithms. It certainly won't be reached by piling on thousands
>of lines of C or Lisp code, trying to solve every problem in 
>the world with mathematics or logic.
> 
>Brad Smallridge
>bsmall@sfrsa.com
------------------------
Thanks for the condensed version. Your "short-story" was better than my
"novel"!:-)  I *AM* an awkward writer, but I keep going in all directions
at once about this stuff. I would swear that I have awakened a couple times
with the key to an awareness core algorithm that might be seminal, and
there is simply too much to say about it, and then suddenly it fades back
into dream time. But it leaves the "feeling", and that's what I tried to
express. 

As to sequential versus parallel, to me they are just the same thing, but a
number of asynchronous functions running on parallel processors, in
otherwords, subsumption architecture using major processors, may be the way
to get the algorithm implemented so it can talk fast enough to ask that it
not be turned off!!! Simulating awareness on one machine might mean an
"interstellar" wait for a reply!
-Steve Walz

