Registered: Dec 2008
Posts: 2

An attenuated version of Mathematica dedicated to the exploration of CAs (cellular automata) could accelerate NKS discoveries.

A version of Mathematica which walls off the mathematics features of the program so that the user is restricted to a well understood discrete tool kit that allow only the creation of CA`s through the rudimentary non-mathematical logic operations used in their implementation could provide the following benefits.

1. Most of the worlds population is not proficient in advanced mathematics which is the most fundamental motivation for the use of Mathematica. Creating an attenuated version of Mathematica renamed to focus on CA`s will greatly reduce the intimidation factor associated with higher mathematics and expand direct hands on interest outside of the Mathematics community. (For now lets just call the product `CA Manipulate` )

2. Steven Wolfram has expressed that CA`s and other similar programs are simpler and are in a sense more fundamental than mathematics, `CA Manipulate` will showcase this concept.

In some sense this addresses a marketing challenge framed such as: If CAs that utilize very simple rules can be used to model arbitrarily complex systems why do I need the most advanced mathematics program on the planet to research and create CAs?

3. By segregating and attenuating the math functions of Mathematica, CA Manipulate could be provided at low cost without adversely effecting Mathematica sales.

4. CA Manipulate, as a small well defined subset of Mathematica will greatly facilitate the communication and interaction of CA researchers by establishing a easily understood well bounded language and manipulation tool kit.

5. Because even simple CAs can be computationally equivalent/ universal and as a result their nature can only be observed through direct implementation, it seems that NKS advancement will be accelerated through the dissemination of easy to use low cost research tools. This will motivate many more people to participate, observer and report findings.

Note: Mathematica Player Pro does not qualify as this product since it does not allow a person to define the logic operations with which to implement CAs,

Report this post to a moderator | IP: Logged

Registered: Oct 2003
Posts: 167

Kevin, I've been a Mathematica user off and on since version 1.0 but it is overkill for almost all my current projects.

It is also an open secret that I've been making some amazing discoveries since November with a rule from a simply defined extensible family 'Generations' that I would not want to start trying to represent in Wolfram's rule encoding scheme.

The tool I use for this is Golly which went 2.0 at the start of December and provides cross platform capabilities for several new families of CA as well as being highly optimised for the needs of those who engineer vast constructions in Conway's Life.

You can get it at http://golly.sf.net/

(I will be posting here about my discoveries as soon as I can get my presentation of them into half reasonable shape. Anyone too impatient can always check http://www.transforum.net/m.cgi?num=2779 where I'm semi-blogging about it.)

__________________
Tony Smith
Complex Systems Analyst
TransForum developer
Local organiser

Report this post to a moderator | IP: Logged

Registered: Dec 2003
Posts: 179

Interesting thread.

Since you bring it up...

I'm offering a monad research tool for $2.

http://cloudmusiccompany.com/monads.htm

__________________
Information Science, Neuroscience, Quantum Mechanics, and Leibniz
http://www.cloudmusiccompany.com/paper.htm

Report this post to a moderator | IP: Logged

Registered: May 2005
Posts: 6

Originally posted by Tony Smith
Kevin, I've been a Mathematica user off and on since version 1.0 but it is overkill for almost all my current projects.

It is also an open secret that I've been making some amazing discoveries since November with a rule from a simply defined extensible family 'Generations' that I would not want to start trying to represent in Wolfram's rule encoding scheme.

The tool I use for this is Golly which went 2.0 at the start of December and provides cross platform capabilities for several new families of CA as well as being highly optimised for the needs of those who engineer vast constructions in Conway's Life.

You can get it at http://golly.sf.net/

(I will be posting here about my discoveries as soon as I can get my presentation of them into half reasonable shape. Anyone too impatient can always check http://www.transforum.net/m.cgi?num=2779 where I'm semi-blogging about it.)

Report this post to a moderator | IP: Logged

Registered: Apr 2005
Posts: 11

Tony,

Here are a few patterns for rule 345/3/6.

http://linuxenvy.com/bprentice/Rule_345_3_6.zip

Of particular interest is puffer 25 (P025.mcl) which generates puffer 2 (P002.mcl) on generation 6080. Could you share with us the seed which generates the interesting organization described in your blog. I look forward to the presentation of your discoveries.

Brian Prentice

Report this post to a moderator | IP: Logged

Registered: Dec 2008
Posts: 2

Tony, Mike, Brian, thanks for showing us your work and pointing out these nice CA software tools. Your substantive entries elevate the blog and contribute to the exploration of NKS.

I feel that great insights can be realized from the exploration of very large implementations of Conway`s Life and the single dimensional CAs as described by NKS, as you have demonstrated.

I am also interested in hearing from individuals that are exploring CA`s where each cell is following a more complex rule set.

The biological world with the exception of viruses is dominated by CAs that consist of cells that exhibit complexity along another dimension outside the number of interacting cells all following a small number of simple rules.

Even the simplest bacterial cells implement a very large number of rules housed in its genome and its epi-genome.

A question that emerges from contrasting simple and complex CAs is: Has evolution (natural selection) favored complex rule set CA`s over simple rule set CAs for reasons of efficiency (fewer cells required to produce the same result)? This conclusion seems likely since even simple rule set CAs appear to be universal. Are there other plausible reasons for the total dominance of complex rule set CAs in the biological world?

Has anyone examined or attempted to quantify this effect looking at the number of generation and cell required to produce a specific result using a simple rule set CA as compared to a Complex rule set CA?

Does anyone know of CA software platforms that allow the user to implement arbitrarily complex rule sets so that these questions can be explored?

Report this post to a moderator | IP: Logged