A New Kind of Nonsense
I still have not finished reading A New Kind of Science, progress is very slow, I'm on Chapter 4, because I am very carefully checking most of his claims, reading about and understanding how Mathematica has taken quite a bit of time. Nevertheless, I've already found so many wrong things with it in so many levels, that I decided I'm going to start writing this multi part review now. I'll try not to be unfair and not make decisions about the whole book, but the first few chapters are not encouraging.
The main thing I noticed about this chapters is that they are rehash of Wolfram's work on Cellular Automata from the early eighties, still is not as bad as it seems, because modern computer techniques, in particular the use of Mathematica, which is an amazing program, makes the subject a lot more accessible and interactive.
My main problem with this part of the book is that it sets up an straw man argument. Wolfram claims, over and over, that "common intuition" from current science says that complex systems can not be achieved through simple programs. This is absolutely false, not one person I ask, most of them scientists and engineers, believe this! All of them believe that simple programs, may produce extremely complex behavior, most of them agree that it is entirely possible that complex programs or rules are not needed to produce a single complex natural phenomena. They all agree with me that that it's not the issue, the issue, the difficulty, is how to discover the simple rules and programs, just from watching the complex behavior. Everybody thinks it takes genius to make this type of discoveries
I don't know if Wolfram has this types of discoveries later in the book. So far he has just presented evidence that simple programs can produce complex behavior and claims this demonstrates the "common intuition" he made up is wrong. This is actually a false conclusion. His "common intuition" is not that simple programs can't produce complex behavior, but that complex behavior, in nature, is not produce by simple programs or rules. Nobody can argue that this program I'm writing on, Radio, is complex and the rules it comes from are not simple, in the sense of Wolfram. His argument can only be applied to natural phenomena. But even in this case he got it wrong, any person familiar with dynamical systems, chaos, cellular automata, and countless other areas of science, know that very simple rules can produce behavior as complex as anything found in nature. But more interesting is that nobody I know makes the claim that complex rules are necessary... the claim is that they are hard to discover. Period.
A New Kind of Science, in my mind, would be something that makes these discoveries more accessible and common. This is not, so far, what is presented in this book.To be continued
© Copyleft 2006 Alfredo Octavio.
Last update: 1/19/06; 3:38:24 PM.