Message-ID: <3240866D.22C8@panspermia.com>
Date: Wed, 18 Sep 1996 16:31:57 -0700
From: Brig Klyce <bklyce@panspermia.com>
Reply-To: bklyce@panspermia.com
Organization: Acorn Enterprises, LLC
X-Mailer: Mozilla 3.0 (Win16; I)
MIME-Version: 1.0
Newsgroups: comp.ai.genetic
Subject: ?Programs that mimic evolution: Results
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: 206.26.142.2
Lines: 34
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news4.ner.bbnplanet.net!cam-news-hub1.bbnplanet.com!howland.erols.net!newsfeed.internetmci.com!pull-feed.internetmci.com!chaos.magibox.net!

Three weeks ago I polled this and a few related newsgroups with the
following question: "According to modern neo-Darwinism, random mutation
and natural selection are sufficient to compose the new genes required
for evolutionary advances. It should be possible to achieve analogous
results _in silico_. Is there any software that can, by random mutation
and natural selection, compose the analog of new genes -- that is,
longish strings of code with unforseeable, useful instructional
meaning?"
	Several replies correctly pointed out my failure to mention
recombination, and it was added to later versions of the question.
	Four replies mentioned Dawkins's "Biomorphs," which definitely does not
qualify. Several people mentioned "Tierra," which may exhibit
parasitism, but does not compose the analog of new genes. Two replies
mentioned Koza, with whose work I a not familiar. Related discussion by
Steve McGrew mentioned genetic algorithms such as "Evolver" and "Gene
Hunter". But as Larry Moran pointed out, these solve predefined
problems. That they qualify is questionable.
	For computers to mimic evolution, they should be able to generate, on
their own, something like the files necessary to upgrade WordPerfect 6.0
to 6.1. Or the files to create an "App" necessary for some other program
to run, such as Netscape offers to provide when needed. Both of these
developments would improve the "fitness" of the pre-existing computer or
software. 	Apparently mutation, recombination and natural selection in
software cannot create the analog of new genes. The logical conclusion
for biology is that new genes must come from outside the system -- the
system is not closed.
	How is this reasoning incorrect? For a more complete discussion of this
position see "Can Computers Mimic Evolution?" at the website "Cosmic
Ancestry", address below. Thanks.
-- 
Brig Klyce * 1503 Union Ave #216B * Memphis, TN 38104 * USA
email: bklyce@panspermia.com
author of "COSMIC ANCESTRY: Life Comes From Space"
at the URL http:\\www.panspermia.com\1cosmic
