Re: Riffing on Wolfram

2002-11-11 Thread Russell Standish
Eric Hawthorne wrote:
 
 Any comments? Can anyone point me to similar speculations?
 
 Thanks, Eric
 

 
  
 
 A collection of thoughts (very much a work in early progress) 
 provoked by chapters 9 and 12 of A New Kind of Science 
 by Stephen Wolfram.
 

... rest deleted ...

I truly like the idea of identifying Wolfram's fundamental CA with
the network of possible differences. If you look at the idea behind
the Schmidhuber ensemble, which is later expanded in my Why Occam's
Razor paper, you will see the set of all descriptions as being the
fundamental plenitude (we tend to use the Multiverse to refer to
solutions of Schroedinger's equation :).

That one gets a sequence of transitions between these descriptions is
simply the TIME postulate descibed in my Occam's Razor paper, which I
learnt recently is closely related to the Kantian notion of Prior -
propositions which must be true of themselves, yet not tautological
(for example I think, therefore I am).

That these transitions must follow a simple (rather than complex) rule
is simply the argument I present in Why Occam's Razor. Even if the
rule is complex, it is most likely indistinguishable from a simple
rule.

However, going beyond this position seems a bit of a flight of
fancy. For instance, I couldn't undertand why the CAs should be 2D. As
far as I can tell, any dimensionality is a possibility...

Cheers


A/Prof Russell Standish  Director
High Performance Computing Support Unit, Phone 9385 6967, 8308 3119 (mobile)
UNSW SYDNEY 2052 Fax   9385 6965, 0425 253119 ()
Australia[EMAIL PROTECTED] 
Room 2075, Red Centrehttp://parallel.hpc.unsw.edu.au/rks
International prefix  +612, Interstate prefix 02





Riffing on Wolfram

2002-11-10 Thread Eric Hawthorne
Any comments? Can anyone point me to similar speculations?

Thanks, Eric



 

A collection of thoughts (very much a work in early progress) 
provoked by chapters 9 and 12 of A New Kind of Science 
by Stephen Wolfram.

---
Caveat: The following was written hastily and in somewhat sloppy,
informal terms, with casual or vague use of some arguably 
pseudo-scientific terms, like de-quantized or classicized by which
I mean something like the process whereby a single state or average
of quantum probabilities seems to take on importance so as be considered
the actual state of some particle etc. after it is observed.
---

Wolfram postulates that space-time is a network (of nodes and connections),
manipulated by simple programs which have the 
characteristics that:

1. the only thing they do is make local adjustments
to the configuration of the network (e.g. replace a node by 3 nodes joined
by connections, erase a connection etc.)
2. They are order-invariant (causally-invariant he calls it) 
in their global effect. It doesn't matter which time-order the local 
replacement rules fire in.

and he goes on to begin to prove how relativity, gravity, matter etc. 
work out nicely in such a model. 
But he doesn't say what the substrate of the universe network is, 
and he cannot yet fit quantum theory into his model,
which got me to thinking:




---
Quantum Computational Cosmology??  - E.H. 2002
---

---
--- The universe is information. More specifically, it is emergent  
--- order within an infinite-bandwidth signal, or in other words, is just
--- a particular, priveleged view of all-possible information, all at once.
---


On reading Wolfram's book, and in particular the part about physics as CAs operating on
a network to produce space-time, matter, energy, I was prompted to have the following
ideas. Please excuse the lack of rigour. I'm just trying to convey intuitions here
and get some feedback on whether anyone thinks there's promise in this direction
or if there are other references people can point me to.

These questions arise: 
1. What would the network of nodes and arcs between nodes, in Wolfram's 
   spacetime-as-network be made of? i.e. what is the substrate of Wolfram's 
   universe network?

2. How do we define the time arrow and what makes the universe 
   as it appears to be?  

My essential concepts are these:


Principle 1
-
The substrate is simply (all possible arrangements of differences)
- 
or perhaps put another way, the substrate of the universe is
the capacity for all possible information, 



The fundament is the binary difference. Each direct difference is an arc,
and network nodes are created simply by virtue of being the things at either end of
a direct difference.

Let's posit that there is a multiverse, which we can think of as
all possible states of all possible universes, or as the information substrate
of the universe.

An information-theoretic interpretation of the multiverse might say that it is
defined as:

a universe with just one thing and no differences (boring) +
a universe with one difference (ergo, two things) +
all possible configurations of two differences +
all possible configurations of three differences + etc.



   ----- A binary difference - the fundamental unit of information
   
   
   
 A --- B   -- two things, A and B, created just by virtue of being defined
   to be at the opposite poles of the binary difference.



To define a particular configuration of the universe, that is, a network
of binary direct-difference relationships between a certain number of
postulated individuals, you can use binary bits, as follows: 
The individual things are denoted A,B,C...
A 1 in the matrix (below left) denotes that a direct difference exists between the
column-labeling individual and the row-labeling individual. 

  E D C B AB - C 
E   1 0 1 1   / \  
D 0 0 1   equivalent to  A - E
C   1 0   \ /
B 1D  
   
   
Every fundamental-level thing that exists is either at the end of a 
direct difference from another thing, or is reachable by some chain of
direct differences from the other thing. things which are not reachable
by a chain of direct differences from some other thing do not exist.

So why don't we posit that the Wolfram network that describes the form
of spacetime at its smallest-grained (i.e. plank-length) level is in fact 
comprised of nodes and arcs which have no other reality (no other material 
that they are made of) other than binary differences. i.e.