From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!newshost.uwo.ca!torn.onet.on.ca!utgpu!cs.utexas.edu!uunet!portal!cup.portal.com!PLai Tue Jun  9 10:06:18 EDT 1992
Article 6031 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!newshost.uwo.ca!torn.onet.on.ca!utgpu!cs.utexas.edu!uunet!portal!cup.portal.com!PLai
>From: PLai@cup.portal.com (Patrick L Faith)
Newsgroups: comp.ai.philosophy
Subject: Re: Quantum mechanics (no AI here, sorry)
Message-ID: <59961@cup.portal.com>
Date: Mon,  1 Jun 92 22:16:33 PDT
Organization: The Portal System (TM)
Distribution: world
References: <1992Jun1.201556.24184@news.media.mit.edu>
  <1992Jun01.234940.40210@spss.com>
Lines: 39

>>(Nick Cassimatis) writes:
>>Even if this is true, so what?  Laplace's statement is based on a
>>counterfactual: "If I knew the state of the universe...."
>>Heisenberg's principle only suggests that we can't know the state of
>>the universe, it says nothing about what could be done *if* we did
>>know the state of the universe.

Good insight !!! Took me awhile to completely understand the ramifications
of assuming an initial condition, a extremely non-trivial problem.

>Now, your idea that the universe does have "real states" even if we can't
>(yet?) observe them is not new; such "hidden variable" theories can be
>built which are compatible with quantum facts.  However, it's been
>pretty well established that such theories can't explain the facts without
>recourse to faster-than-light signalling.

Agreed, and the fact that it always points to finer granularity of space/time
should tell us something.

>For further reading, I'd recommend Richard Feynman's _QED_, Nick Herbert's
>_Quantum Reality_, and J.C. Polkinghorne's _The Quantum World_.

Not going to find what you're looking for in these books, try looking
for Einsteins later notes (especially discrete space/time), or general
relativistic pregeometry ( is a graduate level concept but has a lot
of good computer science).

I was suprised to find this thread in the ai loop, but pleasently because
it belongs in here. There is a lot of work going on in connectionist 
oriented geometry (i.e. n-dimensional geometries) and other so such
weird things that require ai techniques to fully pull them off - for example
if every event had a connection (dimension) with every other event, then
computationally you would have to throw out connections and make new
connections to model (with error) any complex event with a computer, the
ability to chunk and generalize being the ai techniques I think that are
most useful working on such problems.


PLai 


