As I mentioned in my last post, Wolfram's first CA paper limits "valid" CA rules to a special form -- ABCDBED0. But in this article in Forbes, he talks about rule 30, and mentions exploring rule 107. Now, right off the bat, rule 107 does not live up to his special form for rules, mainly because it's an odd number. Rule 30 violates the rule as well, because 30 in binary (00011110) does not follow the ABCDBED0 form.

Also, in the aforementioned Forbes article, they talk about Wolfram reevaluating his earlier views on CA. And this is what I've run into. His original idea of what constitutes a valid CA rule is wrong, and he says so. And of course his earlier paper doesn't say anything about it.

So it looks like I need to read some of his later papers.

Oh, and it was actually quite fun to work out the CA by hand. It's easy to do with a 1D automaton and graph paper.

8:53:37 AM    comments ()  trackback []