X-Message-Number: 25550
Date: Sat, 15 Jan 2005 05:51:07 -0500
From: Thomas Donaldson <>
Subject: CryoNet #25529 - #25537

To Peter Merel:

I observe that human beings (at least most of them) have common sense.
From my study of how brains work (which includes lots of biology, too,
if you don't like biology) it's pretty clear that brains are very
large networks of computers connected with connections which themselves
do much more than simply transmit a signal. I believe I gave you some
references from a recent NATURE on this question.

The point to what Rodney Brooks was doing was that he looked at 
MINIMAL ways a simple animal could solve its problem of walking as an
engineering device. He did NOT try to find a general algorithm for
walking by an N legged device, a much harder problem. Moreover, his
book was published in 2002, and it did not sound like he proposed to
give up his work on these questions.

If you asked me: 1. how long did I think it would be before we made a
machine with humanlike common sense, I would say a long time. Our
present AI machines are only pale imitations of brains, and don't
even work like brains do at all. 2. what road we should take to be
able to make intelligent machines, I would say that we should work
on parallel networks with numbers of processors on the order of 10^11.
Even if we had a single processor 10^11 times faster, it would not
work like a parallel brain network (parallel machines don't act like
single machines). and finally, 3. I would strongly prefer working on
ways to increase our human abilities rather than making an imitation
human: first, we have no reason to make competitors to ourselves,
and second, even if not competitors we already have many such 
devices, which aren't even imitations.

              Best wishes and long long life for all,

                  Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=25550