Juergen Schmidhuber wrote
>But there _is_ a modern religion based on universal principles.
>More precisely, on universal computers.
I can give sense to that. I cannot give 3-sense to the word
"modern" though, it is a typical (plural) first person expression,
which add nothing to a "scientific" or "axiomatic" or 3-defensible
debate.
>According to the Great Programmer Religion, the Great Programmer wrote a
>very short program that computes all computable universes. One of them
>is ours.
So the Great Programmer is not even the UD. The UD, with Church Thesis,
wrote all programs and execute all programs. The problem consists in
explaining appearance of beliefs like "our universe" from possible
(consistent) inside point of views scattered in UD* (UD's work).
>Some disciples of this religion find it plausible because the short
>program is the simplest explanation of all observations.
The UD is the shortest explanation then. Although a shallow one
which only permits a (re)formulation of the mind body problem.
With comp the appearance of big bang and cosmos must be explained.
If you make the distinction between 1-person and 3-person you get
an "absolute" form of indeterminism (the self-duplication) making it
possible to define a measure on the consistent extensions, which,
I repeat are scattered in UD*.
You have still not explain to me how you predict your reasonably next
experience in the simple WM duplication. Nor, a fortiori, how the
1-person can take into account the delay of reconstitution
accessed by the UD (which is necessary to do if you want single out
one program in UD* and one computation. With comp we must take all
computations into consideration.
See
http://www.escribe.com/science/theory/m3044.html for links to
more explanations.
>
>Others find it increasingly plausible because our own, currently
>quite primitive virtual realities are getting more sophisticated all
>the time. We observe a speed-up factor of 10^6 every 20 years, possibly
>10^30 in the ongoing century. More and more people, especially kids, are
>in regular contact with virtual realities, and to them the new religion
>may seem just like a natural extrapolation.
Good argument, corresponding to a new version of the old
dream argument. Good argument for comp, but nocturnal dream
recall us also that certainty of being awake relatively to a
neighborhood does not entails that neighborhood is our "true"
neighborhood. Remembering dreams (especially the common non lucid
one) can help to anticipate the possibility of inconsistency (<>[]f,
A theorem of G* which is not a theorem of G, although G
proves <>t -> <>[]f). I mean "<>[]f" is a typical true and anticipable
but non provable proposition by self-observing sound machines.
Both the thought experiences and their arithmetical
translations illustrates our necessary ignorance about which
sheaf of coherent computations we belong to.
Comp, without QM, entails strong form of non trivial
indeterminacy and non-locality.
Look at
http://www.escribe.com/science/theory/m2855
for a purely arithmetical interpretation of Bell original
inequality (cf [] = godel provability predicate,
<> = -[]-, p represent \Sigma_1 sentence (= UD accessible).
Is that inequality violated in the UTM mindscape? Open
question (although there are argument showing that it
would be quite astonishing that those inequalities are
not violated.
Before going into such technics it would help us if you told
us what is wrong with the comp 1-person indeterminism
in the simple self-duplication experiment.
We discuss that before, but I am still not sure to
be able to make sense of your critics.
Bruno
Received on Thu Oct 04 2001 - 02:57:51 PDT