At 22, Stephen Wolfram already had a great career as a physicist.
So why did he ditch the academic life to make software? He didn't really like school, and he didn't really like doing calculations. And a multitude of scientists, financial analysts and engineers are glad he did because Wolfram's new path lead to the creation of one of the most famous and powerful software programs for highly technical mathematics.
Today, the 56-year-old Wolfram runs his own company called Wolfram Research, which created the Wolfram Language — a programming language embedded in products such as Raspberry Pi computers and Intel's Edison — made for wearables and "internet-of-things" devices. His search-engine-on-steroids, Wolfram Alpha, is used by Apple's Siri to answer computational questions. And the company's original product, Mathematica, remains one of the most highly regarded programs for mathematics and computation.
All of this grew out of two things: Wolfram's frustration and boredom with the academic life and his desire to make complex mathematics and programming easier and more efficient, even for people without technical backgrounds.
From an early age, Wolfram showed brilliance and a talent for physics. He published his first scientific paper at age 15. At 20, he was the youngest person to ever receive a Ph.D. from Caltech. Two years later, he was the youngest to receive a MacArthur grant — sometimes known as the "genius grant."
But academic life did not suit him. He liked solving problems and building things, and wanted to be able to build tools that would help him, and others, do better scientific research more efficiently.
Computers were another great passion. Wolfram had used them since the age of 12 — when they were the size of a desk.
Still, he said, he "wasn't that excited about, or good at, doing mathematical calculations," and software helped automate much of that work.
But by the time he received his doctorate, he was already reaching the edge of what was possible with available systems — he had even built one of his own with colleagues while still at Caltech.
So when he left academia, he began working on a computer program that would automate complex mathematical tasks. While intended for a technical audience, it would be accessible even to people who were not deeply experienced with computers.
He founded Wolfram Research in 1987 with his own money, some of it from his MacArthur grant. The company would be the home for Wolfram's new program.
His friend Steve Jobs suggested Wolfram call it Mathematica.
Wolfram had considered the name, but wanted something pithier. But Jobs "had a whole argument that you have to take a generic name and romanticize it," Wolfram told CNBC. "And I argued 'gosh, it sounds rather long and ponderous.' And his response was 'well, so does Macintosh.'"
The software was considered revolutionary — Wolfram made his first deal with Jobs, who bundled the program with the high-end technical computers Jobs was making at NeXT Corporation, the company he founded after he was fired from Apple. Deals with Sun Microsystems and others followed.
By 1995, more than a million people were using Mathematica. Now in its 10th edition, it is still Wolfram Research's flagship product.
Wolfram Research is perhaps unusual, and fortunate, in that it has created a kind of virtuous circle, where one product is used to research and build other products.
From Mathematica, the company spun out dedicated platforms for data science, finance, and programming, among others. In 2009, it debuted a sophisticated search engine called Wolfram Alpha, which the company used Mathematica to build. Wolfram calls it a "computational knowledge engine."
Wolfram Alpha can give information on nearly anything. It can provide definitions, calculations, and visualizations.
Following Wolfram's philosophy that computers should automate as much as possible, users can ask questions in ordinary language, and the system is meant to interpret what they need. If you ask it to tell you what an "astronaut" is, it will not only give you the definition, but it tell you how many points the word "astronaut" is worth in the board game Scrabble.
That ease with natural language serves it well — it is one of the sites Apple's voice-activated Siri technology uses.
The newest major product is the Wolfram Language — a powerful, cloud-based computer programming language that incorporates elements of Mathematica and Wolfram Alpha.
Like Wolfram Alpha, you can ask it things — what the weather will be tomorrow, what your Facebook friend list would look like on a scatterplot, all of the capital cities of Western Europe, and so on. It can access maps, photographs, and charts. Users can build programs with the information they summon.
Wolfram hopes the language will appeal even to those who have little to no programming experience. "The language is automated enough that the fancy R&D types who use it don't necessarily have any advantage over kids who are trying out programming for the first time," he said.
There is a free version on the cloud, with limited functionality. "You can't design the next great particle accelerator with the open cloud," Wolfram said, but "people who are just getting started programming can write their first programs by walking up to their web browsers and typing, so to speak."
Wolfram now uses the Wolfram Language to build Mathematica.
Other companies have taken note. Wolfram Language is now bundled with the Raspberry Pi — the cheap minimalist computer intended to help teach young people and hobbyists the basics of computing and programming. Intel has also included it in its Edison computer-on-board, a device used for internet-of-things devices and wearable computers.
Wolfram Research employs "700 people across all companies and including contractors" according to a spokesperson. The company is controlled by Wolfram himself, does not share financial information, and is secretive about aspects of its business.
This, no doubt, allows Wolfram to pursue ambitious projects without worrying about irking shareholders. When they were building Wolfram Alpha, Wolfram himself was not even sure if the project he envisioned would be possible.
But, he said, "I am lucky enough to have a successful private company where I don't have to answer to anyone else for what I do, so I can do crazy projects like that."
UPDATE: This post has been updated to reflect employment information from Wolfram Research.