As a Caltech faculty member and researcher studying theoretical physics in the early 1980s, Stephen Wolfram often found himself frustrated — at a computer.
“I needed to rapidly compute equations for my research,” Wolfram explains, but he found the software available at the time to be lacking — a problem he believed he could solve by creating a brand new computer language.
The result, co-developed with a small group of handpicked recruits, was the computer algebra system Symbolic Manipulation Program (SMP). The effect was logarithmic. SMP released commercially in 1981, spawned a software company, helped earn Wolfram a MacArthur Fellowship, and sparked a lifelong fascination with computational complexity.
Wolfram regularly introduces a key concept with an illustration: A collection of very basic programs are coded with a set of rules. The programs — which Wolfram dubbed “cellular automata” — launch out like little worker bees to construct a very simple image resembling a pyramid. With only slight variations to their coding, however, the automata create an image wildly more complex, producing elegant, cascading, interconnected patterns Wolfram describes as “computationally irreducible.”
“When I first saw that, it came as a huge shock to my intuition,” Wolfram confesses. “A very simple program can produce a pattern too complex to predict. The only way to find its outcome is effectively to simply watch it evolve.”
In 1988, Wolfram turned his focus to business, releasing a new version of his computing software, Mathematica — now considered one of the standard software-language environments for scientific, technical, and algorithmic computation and software development. Wolfram continued his research behind the scenes, and in 2002 published A New Kind of Science, a 1,200-page book detailing his extensive study into computational complexity.
By 2009, Wolfram released Wolfram|Alpha, a muscular Web-based “computational knowledge engine.” Similar to a search engine, the service accepts freeform text queries, but instead of merely polling, it both sources and computes relevant data in real time using approaches outlined in his book. Wolfram’s ambitious hope is to make all systemic knowledge immediately computable and accessible to everyone.
Today, Wolfram remains as surprised and fascinated by the intersection of nature and computing as when he began, and he predicts it will help unlock ever-greater mysteries.
“Incredibly simple systems can do sophisticated computation,” he says. “I believe this has deep implications on issues such as biological processes, economies, artificial intelligence...the limits of science itself.”