Re: Computational irreducibility and the simulability of worlds

From: Eric Hawthorne <egh.domain.name.hidden>
Date: Fri, 16 Apr 2004 23:29:33 -0700

We're just doing models or thought experiments here when we postulate
that a universe is (could be)
a simulation in a computer running a cellular automaton, are we not?

Whatever explanation we do come up with eventually is going to have to
explain
1. the "memory" cells themselves, and
2. what a "state transition of which the dance of shiva is composed" is,
in itself.
If we can't explain those things, we just get one of those "I don't
know, but it's turtles all the way down" kind of
unsatisfactory theories.

This seems a hard problem to me. There seems to be a fundamental
connection between information
and representation of state. But we are after an explanation of
"physical state" itself, i.e. the presence of state,
and not just after the secondary thing; the "re-presence" of state.

Can we come down on the side of saying "There's really nothing but
information"? I find it better
to say "No no, information is a RE-PRESENTATION of differences." What
there really is is some kind
of medium with a capacity for being described as having (arbitrarily
combinable) differences. Given such
a medium, information theory and computability theory is probably
sufficient to ensure that everything
observed with certainty will be logically consistent with everything
else observed with certainty, by any
co-existing observers in the medium.

But what the heck is the medium (the "aether" of old)? I guess comp sci
tells us it could be any
turing machine equivalent.

So does that mean we just say "think of the substrate of the universe as
being a turing machine equivalent",
any old turing machine equivalent. Ok, but still, you have to admit that
every "easy to think of" instantiation
of a turing machine (e.g. a PC with a lot of time on its hands) is a
terribly implausible universe substrate.
For heavens sake, the PC with a lot of time on its hands presupposes
time (and space (i.e. different localities,
with notions of adjacency), in which to write the tape). Classic
chicken and egg problem.

Does anyone know the way out of that particular conceptual pickle?

--------------------



Hal Ruhl wrote:

> Hi Stephen:
>
> Observers:
>
> Accepting as a starting point the earlier argument that our universe
> is an interpretation of a collection of the automaton cells considered
> there then going further: What can the collection look like in order
> to have an interpretation compatible with our universe?
>
> It is entirely reasonable that collections of cells extending beyond
> nearest neighbors have state histories that are coordinated to some
> degree. I have called these coordinations dances since the particular
> state progressions can move from cell to cell without the cell itself
> moving - the cell does not change nearest neighbors. Some dances look
> like this.
>
> For example what we call a photon can be such a dance moving through
> the grid of cells.
>
> Dark matter could be just a very large scale dance somewhat kept in
> place by smaller dances within and in turn influencing these internal
> dances.
>
> Further it must be remembered that the progression of states in any
> cell is a succession of discontinuities. One state shifts to another
> with no continuity of states between.
>
> In this venue dances interact and change each other discontinuously by
> mutual collision or by exchanging smaller dances.
>
> How then does a human differ in kind from a rock? Should we expect
> them to differ in kind?
>
> Yours
>
> Hal
>
>
>
>
>
>
Received on Sat Apr 17 2004 - 02:45:50 PDT