Originally published 8 October 2002
Stephen Wolfram’s A New Kind of Science arrived earlier this year in a blizzard of hype. As significant as the works of Newton and Darwin, said breathless touts, and the author seems to agree.
Science has been barking up the wrong tree for hundreds of years, he says. His “new kind of science” will set things straight, explaining everything from elementary particle physics to the economy to the weather to human consciousness and free will.
The universe is not myriad particles acted on by forces, moving in continuous space and time, as described by all physicists since Newton. Rather, Wolfram states, the universe is a computer program — and the program might be incredibly simple.
Not hardware, but software.
A stunning claim, and you may want to check it out. But you won’t read this doorstop of a book overnight. If you want to have a serious go at it, set aside a couple of months on a desert island.
Stephen Wolfram was born in London, and educated at Eton, Oxford, and Caltech. He received a doctorate in theoretical physics at age 20, having already made significant contributions to particle physics and cosmology. In 1981, at age 22, he received a MacArthur “genius award.”
In 1986, he founded his own company to develop and market Mathematica, the world’s leading software for scientific computing, and made himself a rich man. Meanwhile, he was busy investigating a branch of computational mathematics called cellular automata.
Cellular automata are the heart and soul of Wolfram’s “new kind of science.” Let me give an example.
Imagine a grid of cells, like the squares on an endless checkerboard. Each cell can be in one of two possible states, black or white.
Start with any pattern of black squares, perhaps something very simple. With each tick of the clock, let every square in the grid simultaneously respond to the following rule: If a square has two black neighbors (among the eight adjacent squares, including diagonals), it stays the same color; if it has three black neighbors, it becomes black; any other number of black neighbors, it becomes white.
Set the clock running — tick, tick, tick — and watch as amazing, unpredictable and apparently inexhaustible patterns evolve.
This particular two-dimensional cellular automaton, called the Game of Life, was invented by the British mathematician John Conway in the late 1960s, and has given computer geeks endless hours of fun.
Cellular automata can be of one, two, three, or more dimensions, have more than two possible states for each cell, and evolve according to any conceivable rule. What they all have in common is a grid of pixels that change with every tick of the clock.
And here’s the kicker: Even extremely simple cellular automata can quickly lead to patterns of astonishing complexity.
Wolfram spent 10 years investigating cellular automata on his computer, and describes them by the hundreds in his book. The patterns they generate are indeed dazzling, frequently bearing uncanny resemblances to patterns in nature, such as snowflakes, the stripes on zebras, the markings of seashells, curling smoke, or fluctuations of the economy.
These resemblances are not accidental, Wolfram states. The universe is itself, perhaps, a cellular automaton, a multidimensional grid of discrete pixels, ticking away according to some simple rule or rules, yet to be determined.
Whoa! That’s one big jump. From the admittedly mind-boggling patterns that can be generated with simple computer programs to “a new kind of science.”
For one thing, the “science” is not as new as Wolfram makes it seem. Cellular automata have been around for more than half a century, and many people have investigated their properties. Wolfram is not the first to suggest that the universe might be usefully described as a cellular automaton.
For another thing, Wolfram’s “new kind of science” hardly qualifies as science. The book makes not a single prediction that can be tested in the usual way. The universe might be a cellular automaton running a simple rule, but, as Wolfram demonstrates, any number of different rules might lead to the universe we live in. The only way to predict what sort of universe will evolve from a given rule is to run the program from the beginning, which is equivalent to re-running the universe itself.
Wolfram also demonstrates that all complex cellular automata are equally complex, which means — if the universe and everything in it are cellular automata — that the universe, the human brain, and a bowl of Cheerios are equally complex. The human brain therefore has as much chance of explaining the universe as a bowl of Cheerios has of explaining the human brain.
A New Kind of Science is not new and not science, but it is a dazzling compendium of computational play, an important contribution to the theory of cellular automata, and, yes, a work of genius.
Is the universe a computer program? Why not? All of our explanations are metaphorical. For Newton, the universe was a mechanical clockwork. As computers get ever more sophisticated, and dominate our lives ever more completely, it is inevitable that they will influence the way we think about the universe and ourselves.
So, if you wish, think of yourself as a pattern of flickering pixels.
But don’t expect cellular automata to replace the old kind of science any time soon. Ironically, Wolfram’s big book demonstrates rather convincingly exactly why that’s not about to happen.