From newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!malgudi.oar.net!ucbeh.san.uc.edu!poremba Tue Jun  9 10:08:00 EDT 1992
Article 6163 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!malgudi.oar.net!ucbeh.san.uc.edu!poremba
>From: poremba@ucbeh.san.uc.edu
Newsgroups: comp.ai.philosophy
Subject: Observation and QM
Message-ID: <1992Jun8.171546.1403@ucbeh.san.uc.edu>
Date: 8 Jun 92 17:15:46 EST
Distribution: world
Organization: Univ. of Cincinnati
Lines: 126


    I'm fascinated by this discussion of QM and AI. Some have suggested    
that the two are quite separate from one another, but I read something
a couple of months ago which suggests otherwise. In all of the discussion
that I've seen on the net and with respect to all of the other stuff I've
been encountering concerning cognition/AI/perception, this is by far the
most intriguing stuff I've seen. Here's an excerpt from some correspondence
I had with Randall Holmes, which Mark Shanks suggested I post. Holmes
doesn't buy into this stuff, but I'd like to know what some other people
think. Perhaps someone could poke around at the UoCal, Irvine, and try to
get one of the authors to comment.



>    Ever since my Modern Physics classes (2nd year physics for 
>engineers) I've had the feeling that mathematics was turning up in the 
>physics. I mean, where I expected raw explanation of the universe, 
>things related to my more abstract thoughts about mathematics and sets 
>were seemingly appearing in physics. I thought it odd. More recently, I 
>picked up a book called _Obeserver_Mechanics:_A_Formal_Theory_of_ 
>_Perception by Bennett, Hoffman & Prakash (UoCal, Irvine) in which it 
>was proposed that quantum mechanics might result from the nature of 
>perception. I loved hearing that. It make so much sense to me. I think 
>that over the next century or two we can expect theoretical links 
>between logic and physics--the limits of which it will be found are 
>quite similar, given the way cognition works.
>    Here's an excerpt from an informal discussion in the back of the 
>book:
>
>  "It happens that many physical properties, like spin and mass, turn
>  out to be properties of the representations of algebraic groups. We 
>  mentioned, you'll recall, groups, Hilbert bundles, systems of 
>  imprimitivity and the like in our colloquium. Well our bet is that the
>  groups that crop up in physics are intimately related to the groups
>  that crop up in our reflexive observer frameworks and to the symmetry
>  groups of the transition probabilities of observer dynamical systems.
>  And it appears that the Hilbert spaces so ubiquitous in quantum theory
>  might arise from the linearization of specialized reflexive frameworks.
>  If so, then the notion of a physical state--namely a measure on the
>  logic of subspaces of an appropriate Hilbert space--might be grounded 
>  in observer theory. And then quantum measurement theorists and
>  perceptual theorists might have something substantial to talk about."        
>                                                       (pp. 258-59)



Some comments on AI and these threads:


(i)
    One of the questions that has been discussed here on the net is, 
"To what degree is consciousness dependent upon the stuff in QM?" I
think that Bennett, Hoffman & Prakash might ask the question the other
way: "To what extent is QM dependent upon the nature of congnition and
perception?" I like that approach better. 


(i.i)
    The more I think about this stuff the more I really really believe
those guys have hit upon something wonderfully important. They've taken
the stand on `observation' that the QMists haven't been able to take.


(ii)
I'm surprised that some people are talking about humans as being machines.
Webster's 2nd definition in the 9th New Collegiate says "a living
organism or one of its functions," but I would judge this as being a
definition accepted because of the frequent conceptual abuse that
living beings have been introduced to. Random house offers no such
alternate definitions. All of the entries in that text come closer
to Webster's first definition: "a constructed thing whether material
or immaterial."
    I have a notion of how people could be called machines, but it relates
to my understanding of cognitive modeling, as I'll call it, rather than
being based on a judgement about the nature of the universe. I believe
when most people refer to us or describe us as machines they are basing
it on a fundamentally mistaken view of nature. I think there was a book
recently (past 2 years) published on the history of this fallacy which
illuminates its error. I don't remember the title.


(iii)
    I can't buy into the resurrection of possible (many) worlds
interpretations of reality. It didn't work with Leibniz, and it won't
work now. Theories on cognition and reality which are based on 
psychology and neurology are more promising than those which ignore 
these research areas.


(iv)
    Isn't it fascinating (psychologists) how some people get so
aggressive in trying to defend their world view? Perhaps something in
their infancy or early childhood. Possibly a bad parental relationship.


(v) 
    Although I see much potential for AI, I doubt that anything like
human intelligence will be successfully modeled without including the
introduction of massive continual sensory stimulous as we do. Isolation
tanks (sensory deprivation) result in the unanchoring of consciousness.
I read this as saying something about the critical importance of sensory
stimulation to what we understand as consciousness. But, I'd like read
the Brooks publication mentioned on one of the threads before saying 
anthing else on this.



Does anyone know any of the authors of the book I quoted from,
_Observer_Mechanics_? Here's the reference:

  TITLE        Observer mechanics : a formal theory of perception / Bruce M.
                 Bennett, Donald D. Hoffman, Chetan Prakash.
  PUBLISH INFO San Diego : Academic Press, 1989.
  DESCRIPT'N   xii, 285 p. : ill. ; 24 cm.


Mike

  +------| poremba@uc.edu |--------------+
  | @@@                       d          |    "Perhaps nobody yet has been
  | @  @                      d          |    truthful enough about what 
  | @@@  uu  ccc ...  eee ee  ddd   uuu  |    `truthfullness' is."
  | @   u  u c   :..  e  e  e d  d u  u  |                  - F. Nietzsche
  | @    uu  c   :..  e  e  e ddd   uu u |
  +--------------------------------------+



