home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!know!cass.ma02.bull.com!mips2!news.bbn.com!usc!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!news.sei.cmu.edu!drycas.club.cc.cmu.edu!cantaloupe.srv.cs.cmu.edu!kck
- From: kck+@cs.cmu.edu (Karl Kluge)
- Newsgroups: talk.origins
- Subject: Re: Random chess moves
- Message-ID: <BxxspJ.105.2@cs.cmu.edu>
- Date: 19 Nov 92 00:15:18 GMT
- Sender: news@cs.cmu.edu (Usenet News System)
- Organization: School of Computer Science, Carnegie Mellon
- Lines: 40
- Nntp-Posting-Host: g.gp.cs.cmu.edu
-
- > Subject: Random Chess Moves
- > From: David.Rice@ofa123.fidonet.org
- > Date: 17 Nov 92 09:05:32
- >
- > Who1: ksand@apple.com (Kent Sandvik )
- > ID: ksand-121192183050@wintermute.apple.com
- > Who2: lionel@cs.city.ac.uk (Lionel Tun)
- > And: maguire@sun.soe.clarkson.edu (Bill Maguire)
- >
- > BM> Here's something you could do with this game. Take a set
- > BM> of these games, each with different random weights. Have
- > BM> a little tournament then throw out the games that did the
- > BM> worst. Duplicate the remaining games and modify the weights
- > BM> in the duplicates by a small random amount. Repeat until you
- > BM> get bored. Think that final set of games is any better than
- > BM> the first batch Lionel?
- >
- > LT> A number of people have pointed out this sort of use
- > LT> of `randomness' in the design of a program. But please note that
- > LT> the program has been specially designed in this way. The use of
- > LT> randomness here is very specific and very tightly controlled. It
- > LT> is being used as part of the _design_ process.
-
- What Lionel is forgeting here is that natural selection acts on competing
- _representations_ early on in the process -- i.e., those genetic codings
- which are "fragile" in the face of forces causing variation (mutation,
- crossover, inversion, etc.) will get selected against (and selected against
- very strongly).
-
- By the way, the probability argument seems to have dragged on far too long.
- We've been down this road before. It's not even a probability argument,
- really, it's the same old info theory/thermo objection that Creationists
- keep flogging. First, there is a formal equivalence between thermo and
- Shannon info theory (albeit, I've never understood it in detail). Second, we
- seem to all be able to agree that in discussing the feasibility of evolution
- from some initial state you have to code that initial state as well as the
- rules by which it evolves. Eh, bien. This just takes us back to the old
- question of whether the local decrease in disorder represented by the
- biomass on Earth is explicable given the much larger increase in disorder
- created by the emission of energy from the Sun.
-