Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: To Robert Hyatt, Dan Corbit, Christophe Theron , And Other Experts.

Author: William H Rogers

Date: 12:14:05 08/05/02

Go up one level in this thread


I few years ago when I was activately investigating AI I read some reports from
a couple of the larger universities in this country in this area.
In one of the studies, the programmers had input electronic circuitry into the
computer so that it could design new circuits as needed. In going over some of
the advanced designs that the computer had printed out, they discovered that the
computer had come up against a block and then did something that supprised all
of them. It had designed a new transistor to make the circuit work. The amazing
thing about this transistor was that it was a three-deminsional design. A patent
was applied for in the computers name and granted. Just in the last week or so I
read that either IBM or INTEL have now been experinting with 3D circuits in
computer designs, that is the circuits can new be 'layered' rather than just
built on a horizonal level.
In the other university, the computer had some basic mathematical functions
programmed in and then programmed to experiment and to learn. At any point in
its learning, they could ask for a printout of its theorms and axions used to
prove it findings. They went on to say that in no time at all the computer had
taught itself algebra, then geomerty, trig and calculus. At the last reporting
they stated that the computer program had advanced to a level that was beyond
anything that humans had ever seen or understood before and not even their
mathematical Phd's could even began to understand what the computer had figured
out even with the proofs for them to read.
One last thing on computers in general, although we have been developing digital
computers and indeed their circuitry has been improving at a very fast rate,
there still exits another kind of computer circuitry that had bearly scratched
the surface and that is the analog computer circuits. Analog circuits in many
ways mostly immulate the way a human brain works. Who is to say that someday
they will not advance to a higher level that we can obtain as humans.
It will take humans or maybe digital computers to design such a machine, but
then once it works who is to say that it might not redisign itself even more?
Bill



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.