James wrote (recently):
>It seems we're in agreement. You are better placed to spread the gospel of
>the short program - didn't someone say the 'universe is counting' or
>something? And you're the best person to complete the all-important
>exercise:
>
>Exercice : From the statement
>
> "LET A=A+1 GOTO START" generates the universe.
>
>Prove (and make precise) that
>
> Universes are related to
> other universes only by correlation,
> which is a subjective feature.
>
>Regards
>James, Eagle and Serpent.
Yes the "exercice" is all important with respect to this discussion list.
Remember Wei Dai's old question:
(
http://www.escribe.com/science/theory/msg00002.html).
You can also try to guess what is missing in a typical "Schmidhuberian"
reply. (for exemple:
http://www.escribe.com/science/theory/msg00082.html).
Let me try to put things another way:
OBSERVATION gives Standart QM, which is [Schrodinger Eq + Collapse].
EVERETT Theory is given by an ontology obeying [Schrodinger Eq];
THEN Everett & Co. gives an phenomenological explanation why we (as
machines) will observe the Collapse (with the right probability).
WEI DAI-SCHMIDHUBER-HIGGO-... Theory is an ontology given by [the
counting algorithm], including the minimal amount of number theory or
computer science to provide meaning to the counting algorithm (of course).
That is a very nice Pythagorean Theory, but, what I say, is that, if we
pretend that it is an explanation for "everything", then we MUST provide
a
phenomenological explanation of BOTH Schrodinger eq. AND the collapse.
5O% of my thesis (
http://iridia.ulb.ac.be/marchal) makes that obligation
precise. I mean 50% of my thesis explain why, if we believe in comp, we
really must and can derive a non trivial part of QM from classical
computation theory. The other 50% gives a precise attempt toward such a
derivation. The "easy" part is the quantum qualitative aspects
(indeterminism, non-locality, many-worlds phenomenology, quantum logic,
etc.). The difficult part is the (classical) hamiltonian smoothness.
BTW I discover recently some interesting clues toward that goal in the
impressionning paper by Tommaso Toffoli "Action, or the Fungibility of
Computation" in the nice book edited by Anthony J. G. Hey "Feynman and
Computation", Perseus Books, Reading, Massachusetts, pp. 349-392.
I recommand the whole book for those interested in the "information or
computation" thread.
A+ Bruno.
Received on Thu Mar 18 1999 - 01:37:36 PST