Path: utzoo!attcan!uunet!tut.cis.ohio-state.edu!ucbvax!bloom-beacon!eru!luth!sunic!mcsun!ukc!edcastle!aipdc
From: aipdc@castle.ed.ac.uk (Paul D. Crowley)
Newsgroups: comp.society.futures
Subject: Re: expert systems & work ethics
Keywords: expert systems, AI, fundamental limits
Message-ID: <2154@castle.ed.ac.uk>
Date: 8 Feb 90 15:51:52 GMT
References: <93@macuni.mqcc.mq.oz> <9001152252.AA26560@yang.cpac.washington.edu> <127@macuni.mqcc.mq.oz> <4863@crdgw1.crd.ge.com> <142@macuni.mqcc.mq.oz> <439@unicorn.WWU.EDU> <772@geovision.UUCP>
Reply-To: aipdc@castle.ed.ac.uk (Paul D. Crowley)
Organization: Edinburgh University Computing Service
Lines: 26

>In article <439@unicorn.WWU.EDU> n8445388@unicorn.WWU.EDU (treon verdery) writes:
>>In the same paper I also wrote about a possible political dilemma:
>>a computer model that could accurately simultate & retroactively 
>>predict the economy & political reality of the USA.
In article <772@geovision.UUCP> gd@geovision.UUCP (Gord Deinstadt) writes:
>Such a system would, by its very existence, *change* the economic and
>political reality.  So its simulation can't be perfectly accurate.  (See
>Heisenberg.)
>
>Unless, of course, it's prophecies are self-fulfilling.  But then it is
>no longer a simulation.

Not necessarily. The information on the system could be kept secret by
the reader. Or the system could be manufactured by Cassandra Computing.
Or there could be a kind of inevitability about the behavior of the e
and p reality, such that it would not be greatly changed by being
accuratly predicted.

Anyone want to write a short story assuming the existance of one of the
above such systems?
 

-- 
This posting contains logical punctuation, for which I make no apology.
It may be reproduced freely in part or whole correctly accredited.
Paul D Crowley aipdc@uk.ac.ed.castle ---  It wasn't me, it was my aardvark.