Re: Interpretations, subjectivity

From: SLP <SLP.domain.name.hidden>
Date: Thu, 08 Jul 1999 06:47:35 -0500

Hans Moravec wrote: "Attribution of subjective experience is a
festering issue……"

This is actually the key point. We don't know what subjectivity is.

Passing the Turing Test is a necessary but not sufficient
condition for a consciousness. A sufficiently responsive mechanism, but
lacking self-awareness with qualia, could "outwit" any possible Turning
test. A proper definition of consciousness must await our understanding
its physical and computational basis.

We don't know whether there will be a smooth reduction of our common
(and admittedly folk-psychological) concept of consciousness to strict
physics (the position called "reductive materialism") or whether this
concept is so flawed that it will have to be junked completely
("eliminative materialism'). Until we actually understand how brains
work, and what we crudely call "consciousness" is generated, it is
premature to attempt an answer.

Talking about "interpretations" and "consciousness" ( or "subjectivity")
at this point therefore seems wholly premature.

On the other hand, we CAN say what reality is like, because we do
understand its math, at least to a certain level, confirmed by massive
amounts of experimental data. We can say, for example, that reality is
deterministic. Consider the following 3 points:

(1) QM really is the Everett formalism, and this implies determinism.
And if we look
at work by Gell-Mann and Hartle, which extends Everett's and explains
approximate classicity at the coarse-grained level, we see the
determinism more clearly when we consider the full cosmological context,
where literally EVERYTHING comes about as a consequence of the form of
the initial density matrix. I would also argue, although this may be
controversial, that by the understanding given of equivalence classes in
GMH, quantum state vectors are seen to be incomplete descriptions, and
the uncertainty principle, even though it is a rigorously derivable
consequence of the formalism of QM, therefore comes across (in some
appropriate sense) as only epistemological. I find it bizarre that
Gell-Mann's merely verbal statements seem to go against what his math is
saying. How, for example can he say that QM is somehow
"indeterministic" even though everything-the entire evolution of the
universe-is completely specified by the initial density matrix? And how
can he claim, so arbitrarily, that the particular quasi-classical set of
correlations that we call our macroscopic reality is the only real one
or even somehow (except from our very parochial point of view) a
particularly special one?

 (2) The "view from nowhen" of QM, almost by definition, means a fully
determinate block universe.

(3) It would appear, on the basis of the existence of quantum
correlations, that the only possible "true" wavefunction can be the wave
function of the entire "Universe," i.e., of all that exists. This
Universal wavefunction cannot depend on time or space. One easy way to
see this is by bringing in general relativity. If the Universal
wavefunction did depend on time or space, then we would have a
dependence on how we chose space or time coordinates, which is
inconsistent with general relativity. Space and time must therefore be
derivative entities whose "information content" is already contained
within the timeless and metaphysically determinate Universal
wavefunction. What is, is--complete determinism.

In a fully deterministic universe, what brings about the determinism?
We can SPECULATE the answer is pure logical consistency. Doing this
gives us a generalization of ordinary MWI to an "everything "
hypothesis. And in an appropriate form (e.g., Tegmark's) this
hypothesis is testable.



Steve Price, MD
Received on Thu Jul 08 1999 - 04:57:55 PDT