[The Universe as a CA] - A New Kind of Science: The NKS ForumA New Kind of Science: The NKS Forum
Pages:1
The Universe as a CA
(Click here to view the original thread with full colors/images)
Posted by: Maxime Chevalier
Hello Everyone,
I'm an undergrad computer science student at McGill University. I'm pretty new here, because I just found out about this site today from a friend.
I just wanted to say I'm quite amazed. This idea, of the universe as an elementary computational machine, is an idea which I had been entertaining for quite some time, without knowing that there had been work on the concept. From the time I knew what a turing machine was, I kept thinking "The universe must be the largest turing machine and God should be reffered to not as The Great Architect, but The Great Programmer". I find it very interesting to see that this concept is in fact being seriously considered by many people. Cellular automatas also seem like a very interesting form of computational machien to me, which are probably more suitable for "universal" behavior to emerge.
Well, that being said, I haven't really had time to read the book from cover to cover, but from what I did read, I am in love. I always thought that the claims quantum physics make about the discrete nature of the properties of elementary particles were very good indicators that our universe was, in fact, built just like a computer simulation. Where everything is discrete, finite, and there are very real limitations to everything that can be. Virtual particles also seemed to me like they were a good indication that in such a simulation, a random factor could be used to keep things moving.
My only disappointment is that a cellular automaton mimicking the universe hasn't yet been found!
I was just wondering, however. How far has this investigation been taken? Has anyone worked with 3 or 4-dimensional CAs? Has work been done on multiple parallel CAs? For example, a separate CA for handling gravity.
I'm tempted to program my own 3D CA in C++. 2GB of RAM on this PC, should be enough to have my own tiny 1024^3 micro-universe :)
Posted by: Ed Hanna
Dear Maxime,
Funny you should ask about a cellular automaton mimicking the universe.
I started playing around with a model of discrete space-time based on Feynman's checkerboard, and it developed into something that reminded me of an article in
Scientific American some years ago about Stephen Wolfram's work, so I came here looking for updated information on CA's.
I'm not even sure if this model is anything like a real CA, since the primary rule is that a particle (a filled cell) moves to the next cell with a given probability, rather than
according to a certain rule.
In any case I have a 250 x 2000 cell model that exhibits certain interesting properties found in the real universe: the speed of light, a Lorentz equation, an uncertainty
principle, and a wave / particle duality.
I know this sound like a crackpot model, but to me, it sure seems to mimic a tiny 250 x 2000 piece of the universe.
I've written a paper on it, and I'd be curious to hear people's reaction to the paper, and whether the model could be adapted to a CA.
The model and the paper are on the web at www.geocities.com/stq50
I am also looking for an endorsement for an arXive posting, if anyone can help me there.
Posted by: Maxime Chevalier
I was actually thinking of doing the same thing (moving particles). I don't think you should necessarily looking to make this a "standard cellular automata". After all, cellular automatons, in the way they are usually defined, are just one of many possible equivalent models... And it's important to realise that whatever relation is more convenient is often better to use.
I'm thinking of doing a 3D cellular automata with between 256^3 cells and 2048^3 cells, depending on the speed I can get it to update at (and how efficiently I can manage to render it).
Maybe we could chat at some point to exchange ideas if you have MSN messenger. My address on there is roue at sympatico.ca.
Posted by: Garrett Neske
In his chapter on Fundamental Physics in NKS, Wolfram tends to deemphasize cellular automata as appropriate models for the universe, instead focusing on networks. The reason for this is that a valid simple-program model of the universe must have as little rigidity initially built in as possible. In this way, one divorces any preconceived notion of what space and time actually are, focusing instead on causality.
While Wolfram is quite far away from a computational model of the universe, as he rightly exclaims, it seems as though his methods may succeed where traditional mathematics have failed.
Posted by: Ed Hanna
Dear Garrett,
I understand what you are saying about avoiding any preconceived notion of what space and time actually are (as a general case), but in my specific case I *started* with the question that if space-time were a discrete lattice rather than a space-time continuum, how would it act, or how would it be different from a continuum.
The original question I was trying to answer was something along the lines of: It is generally accepted that matter and energy occur in discrete quantum units while space and time is a continuum. But if matter and energy are in discrete quantum amounts, then how can space and time not be discrete, since space, time, matter, and energy are so inter-related. One familiar example that everyone knows is E=M*C*C. If you rewrite this equation as E/M=C*C then the units of energy and mass on the left are discrete, so how can the units of distance and time on the right be a continuum? (A more concrete example is that if the potential energy of a suspended weight is in quantum units, there must be quantum units of height that it can be suspended from.)
From my (biased) point of view, the universe ought to be made of discrete quantum units of distance and time, each of which is either occupied or empty.
Surprisingly, a simple version of Feynman's checkerboard seems to mimic / duplicate several key attributes of the real universe.
Posted by: Ed Hanna
The period at the end of the sentence attached itself to the URL in my original reply. The paper and model are on the real website at www.geocities.com/stq50 Sorry for any confusion.
Regards,
Ed Hanna
stq50@yahoo.com
www.geocities.com/stq50
Posted by: Joshua
On the applied NKS I have posted a simulation of Feynman Diagrams in a cellular automata. I am cross posting to this because it might be a pure CA idea .
Posted by: Joshua
Rule 10 in the paper should read real particle paths are set to zero.
Forum Sponsored by Wolfram Research
© 2004-2013 Wolfram Research, Inc. | Powered by vBulletin 2.3.0 © 2000-2002 Jelsoft Enterprises, Ltd. |
Disclaimer
vB Easy Archive Final - Created by Xenon and modified/released by SkuZZy from the Job Openings