Rough draft...  posted 7/16/2006

                      LEARNING, CERTAINTY AND CAUSALITY
 
                                  LCC-0
                               15 July 2006
 
                      Copyright (C) 2006 Homer W. Smith
           Redistribution rights granted for non commercial purposes.
 
 
 
     ABSTRACT

     This paper is about the Machine Certainty Theorem (MCT) and its
possible ramifications.

     The MCT states that a machine can not learn anything with
certainty.
 
     This statement depends heavily on the definitions of machine,
learn, and certainty.
 
     A machine is defined as any system of parts interacting via cause
and effect across a space time distance.

     Learn is defined as come to know.

     Certainty is defined as perfect certainty, that which can not be
wrong.
 
     Among other things a machine can not learn with certainty whether
or not it or anything else exists, nor whether or not any effects are
caused.

     This is because a machine learns by being an effect of causes,
even when it is learning about itself, and since effect does not prove
cause, a machine can never learn with certainty if cause exists merely
by studying alleged effects.
 
     Cause and its nature forever remain a theory to a machine.

     An unnecesssary one at that, because machines ultimately are only
interested in correlations, not causations.  And correlation does not
prove causation.
 
     The application of the MCT is in the reverse, since consciousness
can learn with certainty a number of things, including its own
existence and personal agency (causation), one is led to conclude that
consciousness is not a machine.

     The formal statement of the MCT and its application are as
follows.

 
                       MACHINE CERTAINTY THEORM
                            (The "Proof")
 
     Learning biconditionally implies learning with certainty or
learning with not certainty.

     Distance and learning implies learning by being an effect.

     Learning by being an effect implies not learning with certainty.

     Learning with certainty, therefore implies Learning but,

     Not by being an effect, and

     Not across a space time distance.

 
     INTRODUCTION

     This work is about Learning, Certainty and Causality,
 
     Specifically it is about learning with certainty about causality.

     Historically it has been admitted that the physical sciences do
not provide certainty of truth, only high probabilities of workable
dependability.  What other certainties we might profess have generally
been relegated to faith.
 
     Put simply, in a mechanical space time universe, causation is not
sufficient to witness causation.

     That is, learning by being an effect is not sufficient to prove
the existence of cause, nor does learning by being an effect allow us
more than to model or theorize about the true nature of that possible
cause if it does exsit.

     However there is a certainty of interest that resides between the
absolute uncertainities of the physical sciences and the world of
faith.

     This is the certainty of consciousness, the certainty that we are
conscious, that we exist, that we perceive, that we care, and taht
we can exert personal agency (cause) into the world.

     Tell a conscious unit there is no cause in the world, and you
have told him he does not exist.

     The search for causation in the physical universe is actually an
anthropomorization of directly perceived causation in ourselves.
Since we can see by direct perception, and thus with perfect
certainty, that there is causation within our consciousness, we
conclude that there must be causation out there in the physical
universe.
 
     [Some may be tempted to say that our perception of personal
agency is just that, simply a perception, and it doesn't indicate the
actual existence of such a thing as cause with certainty.  They are
saying that inspite of our direct perception of cause in our selves,
perhaps there still is no cause anywhere in the universe.  Think about
that for a while.]
 
     It is this certainty of our own existence, perception, caring and
agency that makes consciousness special and leads us to some startling
conclusions about its non mechanical, non space time, nature.

     The material of this paper started with the question could a
machine learn with certainty of its own existence or of anything for
that matter.

     As will be developed, the answer is no, and yet as a conscious
unit we are quite certain we exist and are agent and thus we are
forced to conclude that we as a conscious unit are not a space time
machine, although we may be interfaced with one, namely the body/brain
system.

     Homer