Newsgroups: comp.ai.games
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!news.kei.com!newsfeed.internetmci.com!EU.net!ieunet!news.tcd.ie!rwallace
From: rwallace@tcd.ie (russell wallace)
Subject: Re: TownSim Models
Message-ID: <DIpxqL.Fyo@news.tcd.ie>
Sender: usenet@news.tcd.ie (TCD News System )
Organization: University of Dublin, Trinity College
References: <484uqc$ekq@mhost.ida.his.se> <48dmi5$74h@rc1.vub.ac.be> <48j113$nr3@news.worldlinx.com> <48o8qg$i45@hermes.cair.du.edu> <DILM7u.9EJ@actcom.co.il> <DIM7tC.5wK@news.tcd.ie> <49aql9$fn7@case.cyberspace.com>
Date: Mon, 27 Nov 1995 20:18:20 GMT
Lines: 26

In <49aql9$fn7@case.cyberspace.com> rtfm@case.cyberspace.com (Dan Thies) writes:

>russell wallace (rwallace@tcd.ie) wrote:
>: The only problem I can see with all these proposals about GAs, different
>: species mating and so on is that the result will be JungleSim, not
>: TownSim.  99% of the activities carried out by modern-day humans are
>: done for reasons having nothing at all to do with number of surviving
>: offspring, so if you want to create a simulated town, GAs won't do the
>: job.

>The reason they do things may not have anything to do with the number of 
>surviving offspring, but the future population will be the surviving 
>offspring regardless of intention.

>You can use any fitness function you want - it could be income.

To be sure.  My point is merely that whatever fitness function you use,
the result will bear no resemblance at all to a town (as opposed to a
jungle).  Not that the result won't be a perfectly good and valid
evolutionary programming project, it just seems to be an error to call
it TownSim.

--
"To summarize the summary of the summary: people are a problem"
Russell Wallace, Trinity College, Dublin
rwallace@vax1.tcd.ie
