Silicon Insider: Sweet Music From a Simple Theory

Sept. 15, 2005 — -- Want to see and hear something very cool? Check out this Web site: http://tones.wolfram.com/generate/.

Pick a style of music and press the button. If you want, mess with the pitch or the choice of instruments or the time signature. Interesting, isn't it? The music is amazingly complex, not the kind of stuff you would hear from a standard music generator. There is a randomness here, but at the same time, a structure -- listen to the melodies emerge in the classical music, the riffs in the rock 'n' roll -- the kind of music that suggests an intelligence has had a hand in the composition.

And yet, if you read the accompanying explanation, you'll discover that all the music you are hearing -- and you can generate literally billions of tunes, enough to fill a lifetime of continuous listening -- is being generated automatically (literally) by a computer using a handful of simple rules.

How is that possible? Well, therein lies a tale …

A Boy Genius's Background

You may remember the name Stephen Wolfram. There was a flurry of press coverage about him three years ago when he published his magnum opus, "A New Kind of Science." The book, and the 10 years of secret research it described, was to be Wolfram's equivalent of Isaac Newton's "Principia," a scientific breakthrough so profound that it would turn our established model of the universe on its head, and send generations of young graduates racing off to spin out all of its implications.

"A New Kind of Science" would also, as a side benefit, finally enable Wolfram to answer the question: Why had the most promising young mathematician and physicist of his era -- the kid who had published his first scientific paper at 15, who had dazzled his professors at Oxford and earned his doctorate in theoretical physics from CalTech at 20, and had been, at 21, the youngest-ever MacArthur "Genius" Prize recipient -- suddenly left academia and "sold out" to the corporate world?

Not that Wolfram was a failure in the business world. On the contrary, he was a raging success. He founded Wolfram Research, which packaged his extraordinary software program, "Mathematica," that enabled users to instantly convert any mathematical equation into a geometric image on the computer screen, and vice versa.

Mathematica, one of the most successful educational programs of all time, transformed the teaching of mathematics in schools and built a gigantic cult following. It also made Wolfram very rich. Rich enough, it was assumed, that he had left his academic career far behind and was living the life of a wealthy chief executive officer somewhere in the distant suburbs of Chicago.

But Wolfram had a secret, and by a weird confluence of events, I was the first reporter to discover it.

As it happened, I knew Wolfram. In one of my early incarnations as a PR guy, I directed the introduction of Mathematica. I had never met Wolfram before, and wouldn't see him again for more than a decade. I found him brilliant, eccentric and more than a little difficult to work with. At one point I even sent him off shopping for a suit with a woman of considerable taste -- only to have him show up at the press conference in the same old Birkenstocks and ratty sweater.

But the introduction was a big success -- even Steve Jobs showed up to acclaim Mathematica -- and in the years that followed, I occasionally checked to follow the company's ongoing success.

An Intriguing Interview

Then, one day in 2001, I bumped into a publisher friend, a neighbor since my childhood, who casually informed me that he was going to publish Wolfram's new book. He told me the title. "Ambitious, huh?" he asked. I figure it's either going to change the world or make him a laughingstock.

I was intrigued. And, being at the time the editor of the world's largest circulation technology business magazine, I was actually in the position to do something about it (over the concerns of most of my staff). I contacted Wolfram -- not an easy task -- and based upon our past connection, got him to agree to an interview.

And that in turn led to one of the weirdest experiences of my professional career. I was picked up from my hotel in downtown Chicago at 8 p.m. by a publicist from Wolfram Research, then driven on a circuitous route -- during which time she twice pretended to be lost -- that finally delivered me, at 11 p.m., somewhere in the woods in Illinois, to a large, brightly lit house. And there, in the doorway, was Stephen Wolfram.

The Theory of Cellular Automata

While the publicist, and Wolfram's family, slept downstairs, Stephen and I spent five hours discussing his work. It seemed that he had spent almost a decade running his company by day, doing his research by night, developing what he believed was a fundamentally new way of looking at reality. It was based upon a field he had help invent: cellular automata. This is the theory that, using very simple items -- say, black and white tiles -- and a handful of equally simple rules, you can create structures of amazing complexity.

What Wolfram said that he had discovered, thanks to the power of the computer, was that if you ran these rules a million, or a billion, times, very strange things began to happen. Not always -- usually the design got real boring and predictable real fast. But every once in a while, for example with a rule he named "# 30," all sorts of weird and interesting randomness began to pop up in all that order.

To illustrate his theory, consider this paraphrased example:

Imagine you are standing on the 50-yard line of a football stadium, and your job is to tile the field with 1-inch-square bathroom floor tiles, half of them black, the other half white.

By the time you cover the insignia at the center of the field, you've only used 10,000 tiles. You realize this boring job is going to take years. So, to kill time, you make up some simple rules -- such as white tiles can line up to make an "L" shape but not a "T." You come up with a half dozen of these rules, and because they amuse you, you stick with them.

Ten years pass. Suddenly, you're done. Four billion tiles. You can't believe it's all finished. Since you've spent all of those years on your knees on the tiles, you have no idea what the whole thing looks like. So you climb to the top of the dome and look down, expecting to see something like sand on a beach.

Instead, you see a flower. A giant flower so perfect that it looks like a black and white photograph. So perfect that it seems alive.

This column is too small a space to go any further into the scientific, even theological implications of Wolfram's theory, except to say that it suggested an underlying structure to the universe that none of us knew was there.

Just Listen

I published my article, which led to a brief flurry of reporters chasing down Wolfram. Then, at last, three years late, he published his book (all 1,200 pages of it) to another burst of press coverage. He even went on a book tour (appearing, memorably, on "Charlie Rose"). Then Wolfram sat back and waited for the explosion to tear apart the scientific world.

But it didn't happen.

Perhaps there is a doctoral candidate out there doing his dissertation on extending Wolfram's work, but I haven't heard of it. And obviously Wolfram's New Science has hardly taken the scientific world by storm. It must have been both heartbreaking and infuriating to Wolfram.

And so now, WolframTones. Wolfram once told me that his goal with his New Science was to plant a seed in the mind of mathematicians and scientists, to ask questions they couldn't answer, and provoke them to pursue this new course. Instead, they merely shrugged.

So, it appears, unwilling to wait, Wolfram has decided to pursue one of these paths to see if he can start that wildfire. What he has done is take some of his mosaic pyramids created by cellular automata -- the ones with the interesting randomness -- taken a slice out of their centers, set them on their sides, and used them, according to equally simple rules, to generate musical notes.

Listen, he seems to be saying, this isn't some dumb, random computer-generated sound. This is real music -- the Music of the Spheres. It is alive, and is so because it taps into the underlying structure of you and me and the universe itself.

Just listen … and consider what it means.