Bleh!

So. I bought Stephen Wolfram's (the Mathematica guy) new book, _A New Kind of Science_. What a monstrous tome. Let me make a few initial remarks. First of all, Wolfram comes across as a complete egomaniac. It's off-putting. Secondly, I'm not sure if he just can't write or he has intentionally set out to repeat himself so often that it gets hammered in to your head. In either case, the writing style is fairly gross.

All that said, let me see if I can boil out the first 250 pages for you. I think the principle concepts are as follows: Instead of the traditional methods of modelling systems with math he suggests that simple computer programs are, in fact, a better method because they, despite their simplicity, be made to model all manners of complex behaviors. Additionally, he seems to be saying that there is a threshold at which adding more rules to a system does not cause the system to produce measurably more complex behavior. In fact as you add rules to a system you will see a progression of behaviors in the ouput until you reach that threshold. Apparently the outputs of various systems can be classified in to four simple types: systems which quickly stabilize to a specific output, systems which stabilise to a structured but repetitious output, systems which produce output with nested structures, and finally, systems that produce complex (almost random) output.

There's a little more to it than that but I think those are the high points. All of these begs an obvious question: so what? I'm hoping he'll get to that. The man, while obviously brilliant, is a master of circumlocution. He says (not implies) that he's going to present something that has huge, paradigm-shifting, consequences in all branches of science. We shall see. For my own part I just wish he'd cut to the chase.

I'll keep you posted.