X-Message-Number: 24055
Date: Thu, 06 May 2004 09:03:42 -0700
From: Mike Perry <>
Subject: Computers and brains again

Thomas Donaldson, #24040, writes:

>If we were to try to actually build a structure with (to be simple)
>1 billion nodes, each one with a connection to every one of the others,
>then it should be clear that we'd run into serious practical problems.

This would probably be a stupid way to go about achieving the desired 
connectivity, but I brought it up to show that it did at least offer a 
substantial reduction of your suggested complexity, namely, from 
N-factorial to just under N-squared. (This is really an enormous reduction, 
from worse-than-exponential to low-degree polynomial, even if problems 
still remain.) To go further: yes, I agree with you that massive parallel 
processing would be desirable. The N "neurons" might be independent 
processors in their own right, and we'd then face the problem of how best 
to establish lines of communication between them. Physical, creatable 
connections are certainly one possible answer, but I'm sure there are 
others that would not involve setting up, in effect, a  private telephone 
line between any two parties, as I suggested before. In fact, real 
telephone connections aren't set up this way, and, even though the physical 
details for our hypothetical brain-machine would be quite different, it 
seems to me something short of the physical growth of connections might be 
workable. But whether it would or would not does not by itself change the 
relevance of the theory of computation to what is happening, both in brains 
and in our artificial devices. I will also offer the conjecture that 
artificial brainlike machines of the future may significantly differ from 
today's computers, but will also differ significantly from natural brains, 
because we will have found better ways of accomplishing what they both do.

Mike Perry

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=24055