Quest for a Code of Nature

A Beautiful Math: John Nash, Game Theory and the Modern Quest for a Code of Nature

By Tom Siegfried

Joseph Henry Press

ISBN 0-309-10192-1

One of the reasons human beings are such wonderful creatures to be around is because they are so endearingly unpredictable. Get two of them in the same room and watch the sparks fly. Fill a planet with them and it's just one big party. Or, at least, that's the way it should be. Individual and group behaviour can also cause calamity and suffering on often vast scales, which is reason enough to build a model to render it quantifiable and predictable. Tom Siegfried believes game theory holds the key to solving this 'Code of Nature'.

Siegfried - a science writer, not a scientist - found his inspiration in Isaac Asimov's Foundation trilogy, where character Hari Seldon's 'psychohistory' method is called upon to predict the future. Sounds cute, but it's science fiction, right? Hang on. It turns out Asimov's method relied on statistical mechanics, used by physicists to make predictions about the movement of molecules based on probability distributions. Substitute humans for molecules and the thinking is still sound. Throw in John Nash's game theory, modern brain research, statistical physics describing networks and societies, and empirical methods that measure players' psychology, and the idea starts to look even more robust. In fact, it turns out there are scientists today thinking exactly the same way Asimov did 50 years ago.

Siegfried is enthusiastic and builds his case as though it were a whodunnit. It's a great cast, including Adam Smith, Darwin, game theory's early architects John von Neumann and Oskar Morgenstern, all the big names in statistics and Nash himself (although nowhere near as much as you'd suspect by looking at the book's cover). The author has also interviewed many academics at the coalface in this field, but detractors are nowhere to be found, and there are moments when it all seems a bit hopeful — a theory-in-progress. After all, a Code of Nature machine will surely only be able to produce probability distributions of how society will act — and what about that wonder theory of the 1980s, chaos, that suggests even an imperceptible change in initial conditions could throw all these nice uniform statistics into disarray? On the other hand, a leap in computing power and a tweak of the model is all it might take to craft a tool that will stop the next war (or revolution!).

Reviewer: Jeremy Chunn

www.cosmosmagazine.com