Xconomist of the Week: Stephen Wolfram on Big Ideas & Companies
Stephen Wolfram is not one to be summed up in a few pithy quotes. Well, too bad.
Here is his life in a nutshell:
“I grew up in England and went to all sorts of good schools that I thought were completely irrelevant.”
“By the time I was 20 years old, I was a physics faculty member at Caltech, and I was building a big software system that was a forerunner of Mathematica.”
“Along the way, I learned a lot about what not to do in starting a company.”
“For about a decade I was almost a complete recluse, running the company from a distance, and spending every night working on basic science.”
“I don’t really have a boss. I just do what I want to do. The trick is not to have a private company that gets too weird and too pathological.”
That was a sampling of what the distinguished and controversial Wolfram had to say at our Xconomy Forum in Boston last week, called “6×6: Six Cities, Six Big Tech Ideas.” For those who don’t know, he is the founder and CEO of Wolfram Research, the creator of Mathematica and Wolfram Alpha, and the author of A New Kind of Science. He is also a recently minted Xconomist.
Wolfram, 52, set the table for the theme of our event, which involved some of the biggest ideas in technology and business from entrepreneurs and executives around the country. As only he could, the physics and software guru reflected on his 30 years in the tech industry and his contrarian approach to running big projects and building companies.
A few things really stood out to me in his talk. One was the importance of making mistakes early in his career. While at Caltech in the early ‘80s, Wolfram got into a “grisly early-IP-meets-university” battle over his software tools, he said.
“I ended up deciding I had to start a company around the software system I’d built, and of course I was just a physics kid. I didn’t know anything about starting companies,” he said. “I made lots of mistakes, like not running the company myself, hiring a CEO who was twice my age, and so on. The company quickly started doing things that I thought were silly and boring. In the end, after many trials and tribulations, it did in fact survive and finally went public and was gobbled up by bigger fishes.” (You can read more about his first startup, Computer Mathematics, which was venture-backed and later merged with Inference, here.)
Another thing Wolfram figured out early on was that university work was not for him. After spending time at Caltech and the Institute for Advanced Study in Princeton, NJ, he started a research center at the University of Illinois at Urbana-Champaign to study complex systems and complexity theory. “My plan A was to get lots of other people to help work” on the implications and applications of his findings, he said. “It was OK, but it was really slow. I got kind of frustrated and needed a plan B,” he said. “My plan B was to build the best possible tools I can, set myself up personally in the best possible situation, forget the idea of having everybody else help, and just have a blast figuring out a bunch of science for myself.”
So, 25 years ago, he started working on Mathematica, a symbolic programming language for scientific computing, via Champaign, IL-based Wolfram Research. “This time I was CEO of my own company, I didn’t take anyone else’s money,” he said, and things “worked out extremely well.” (Mathematica is a cash cow in the fields of education and algorithmic software development. Wolfram Research has been consistently profitable and currently employs about 700 people.)
Meantime, Wolfram moved to the Boston area and spent the ‘90s working on his book, A New Kind of Science (the details of which I can’t get into here), all in parallel with running Wolfram Research remotely. “The company continued to grow and prosper,” he said. “Despite some of my hopes that it would happen, there was no coup.”
If you’re thinking “crazy maverick genius,” you’re spot on—but it gets crazier. (Did I mention he’s saved every keystroke he’s typed in 22 years so he can analyze how to be more efficient?) Wolfram’s most recent project, called Wolfram Alpha, seeks to “take the knowledge of our civilization and make it computable, so that we can have something that automatically answers questions as well as or better than any human expert in any field,” he said. Basically, you can type in any query (on scientific, financial, economic, or other matters) and instead of giving you links like a search engine, Wolfram Alpha tries to compute the answer based on huge swaths of curated data, natural language processing, human experts, and 15 million lines of Mathematica code. (This sounds a bit similar to Paul Allen’s Project Halo, though there are big differences too.)
“I’ve worked on some pretty complicated projects and products over the years, but Wolfram Alpha is, from my point of view, kind of absurd,” Wolfram said.
In the spirit of understanding how and why to tackle such a huge idea, Wolfram shed some light. “Ever since I was a kid, I have been thinking about how to systematically organize knowledge and make it computable. I thought I’d have to solve the general problem of artificial intelligence. About every decade or so, I came back to that and always thought, ‘It’s just too hard.’ But from [A New Kind of Science] and thinking about the nature of intelligence, I realized I was not thinking about it correctly,” he said. “And that in some fundamental sense we already had everything we needed for computational knowledge, just from computation.”
What he means is he was able to tackle Wolfram Alpha because of decades’ worth of software framework built up at his company. Crucially, he also had the freedom—financially and intellectually—to pursue a big, risky project like this because of the company’s success to date. Of course, it takes a certain audacious mindset too.
“I think the single most important thing in that project, and in other impossible projects, is the simple point of believing that one can do the project,” Wolfram said. “Having the confidence, maybe the arrogance, to believe that even though other people have tried and failed, and lots of people tell them it’s impossible, that one can nevertheless actually do it. And then being able to lead other people to throw themselves into it too.”
It remains to be seen how well Wolfram Alpha will ultimately perform, but it’s been quite a ride already. The project has advanced far beyond the experimental stage: Apple’s latest iPhone uses the technology in its virtual assistant, Siri. “It’s always a strange thing,” Wolfram said, “starting with just an idea, gradually turning it into something real, and inventing along the way all those things that start small and eventually become huge structures that everybody assumes were always there. I always think that anyone who’s been involved in such a thing has a kind of glow of confidence that lasts about a decade or so—realizing that, yes, with the right effort, nothing can turn into something.”
But let’s come back to Earth here. There was a pretty strong overcurrent of do-it-yourself, anti-VC sentiment in Wolfram’s company-building comments. (To be fair, some of his best friends are probably VCs.) Is any of that applicable or realistic for most tech entrepreneurs these days?
“Our company is pretty different from a lot of the industry,” Wolfram said. “Our goals are a bit different about building these giant things that are interesting, independent of the difficulty of doing it.” Plus, Wolfram Research has been around for 20-plus years and is still privately held. “That’s an unusual thing in the technology business. Usually people get venture capital, there’s a certain cycle of things happening, it either gets very big or it blows up, or it gets bought by somebody,” he said. “For Wolfram Alpha, I can imagine some venture capital-type pitches for that project—and I can’t imagine they would have ended very well.”
Another insight, which is more applicable to medium-size and larger companies, had to do with hard-wiring innovation into the corporate culture. Wolfram has a theory about “the rhythm of innovation,” he said. “The notion is, we’re always going to be doing something qualitatively different every couple of years, and just get used to that. There will always be a new thing happening, and there will always be people who will be pulled into that new thing.” He added, “So I get to do the scary thing to folks at the company, which is with great regularity saying, ‘I have a new idea. Let’s see who can work on it.’”
Overall, Wolfram seems very optimistic about the progress of technology. Given the current state of the industry (and his own role in it), he said, “We have sort of a Cambrian explosion moment. I’m certainly having more fun and feeling more productive than at any time I’ve been involved in the technology industry.”
One of the questions from the audience had to do with Wolfram’s philosophy on what can be predicted about the world. “When it comes to what happens in society, there’s sort of a mixture of things you can compute and things that are irreducibly hard to compute,” he said. “Basically the only way you can work out what’s going to happen [in the latter case] is to directly simulate them, or to just watch what they do.”
Perhaps an example of that comes from Wolfram’s home life. “I’m always trying to figure out where things are going, and I make these predictions about what will happen,” he said. “My wife is always reminding me of the following story of what I’d said in the early ‘90s. We had been building a house. OK, how do we put a television in this room? I said, ‘Well, you just make this little niche in the wall, 4 inches thick, that will be fine for a television.’ And, well, of course, there were eventually flat screen televisions, but it wasn’t in the early ‘90s.”