Wednesday, March 24, 2010

A New Kind of Science

I picked up this book in 2002 when Stephen Wolfram brought his traveling show to CMU, and I finally got around to finishing it. The whole approach of the book is absolutely fascinating. I mean, the idea is simple enough -- there are many aspects of science that can better be studied and described via algorithms than via formulae. I don't mean to trivialize it -- it is an important conceptual shift. And so, to some extent, I can understand Stephen's bold approach. And further, I certainly understand how making bold provocative statements can draw attention to matters that are much in need of discussion and debate. And, finally, no one can say that Stephen did not do his homework on this -- the book is 1200 pages long, and goes into very, very thoughtful detail about the analysis of repeating algorithms, and their implications. But, man, his approach completely backfired, I think. This note in the back of the book sums things up:
Clarity and Modesty. There is a common style of understated scientific writing to which I was once a devoted subscriber. But at some point I discovered that more significant results are usually incomprehensible if presented in this style. For unless one has a realistic understanding of how important something is, it is very difficult to place or absorb it. And so in writing this book I have chosen to explain straightforwardly the importance I believe my various results to have. Perhaps I might avoid some criticism by a greater display of modesty, but the cost would be a drastic reduction in clarity.
Can you believe that? "My work is so incredibly important that to present it in an objective way would be confusing." Don't get me wrong, there is some really great stuff in here. It must have taken at least ten years to write this. The whole concept is meaningful and thought-provoking. But the presentation is such that it is hard not to roll one's eyes and say, "Oh, come on..." half the time. I mean, even the title! The book is the picture of a privileged, self-important monomaniac who looks down on, well, everyone. I hate to say something so unflattering about someone I don't really know, but that's how he comes across. And frankly, it ticks me off a bit, because I feel like if he had presented his work in a more collaborative, inclusive, humble manner, I think it would have served as 100x the inspiration, and moved the field forward significantly. But, as it stands, here it is, eight years later, and no "algorithmic revolution" has taken place, I think partly because it is a little embarrassing to be associated with this arrogant book.

Maybe I've got it all wrong, but I think the amazon reviews tell the story.


  1. Completely agreed Jesse; well said. I never did "finish" the book, but it didn't take consuming the whole thing to get Wolfram's point, or to be lashed repeatedly by his ego. His biggest failing with this book IMO was to the conspicuous lack of any references to others' work: this made it impossible to evaluate some claims or to consider his work in any real context, and of course it simply underlines the monomaniacal nature of this in that he apparently did not think there were any other sources worth referencing!

    I believe that the "algorithmic revolution" is a significant step forward: it enables ''generative social science' (Epstein and Axtell's phrase preceding ANKOS) via models of what cannot otherwise be modeled, as well as freeing us from an over-reliance on formula and over-arching theory for systems (e.g. neural and social ones) that may in Dreyfus' words be a-theoretic. It's also an interesting meta-comment on what science can be. Unfortunately, the ponderous, self-important layers these thoughts are wrapped in have all but smothered it. I expect we'll get to the same place in terms of modeling and algorithms (though not necessarily based on Wolfram's classifications of cellular automata), but it will take a lot of work and publications by others to get us there.

  2. You know something Jesse, I've also observed that those who are greatest at what they do are usually the most humble in their approach to things. And so I too appreciate a more modest approach. Though, that being said, I do think that Wolfram understands the power of rhetoric and that sometimes a more *meaningful* (bias, opinionated, non-scientific) approach is necessary for non-domain experts to see the /point/ of your work.

    Of course, anyone who truly understands what you're saying is going to end up being annoyed with you for thinking you're so important in the first place! ;-)