Artificial Knowledge Is Power
by Chris De Benedetti
Creating a cutting-edge program based on the ideas of a 300-year-old British minister sounds like the punch line to a Web geek's bad joke.
Meet entrepreneur Michael Lynch. When Net experts call him the Bill Gates of Europe, they're not kidding. He's the founder of Autonomy, a knowledge-based software start-up, and Britain's first billionaire at the ripe old age of 35.
When Lynch studied the Rev. Thomas Bayes' mathematical theorems, he knew the 18th-century minister's formulations could change his doctoral research on pattern recognition. What he couldn't know was that Bayes' ideas would
inspire Lynch's two successful start-ups and possibly revolutionize artificial intelligence on the Web.
|Prime Mover: Michael Lynch|
Founder and CEO
Founded Neurodynamics Ltd. Then founded Autonomy in 1996. Created an algorithm that explains how to predict occurrences based on prior related events.
Cambridge University — Ph.D., 1988
Based on Bayes' probability theorem — which mathematically explains how to predict occurrences based on past, related events and present conditions — Autonomy licenses artificial-intelligence software to more than 200 of the world's biggest companies.
"Computers are very obsessed with black and white ... and ideas are not like that, they're expressed in different ways," says Lynch. "What you need is a technology which understands the shades of gray in the world. Somehow, Bayes understood that 250 years ago."
Lynch's software organizes unstructured information from complex company databases by extracting meaning from its words and context. The software has myriad uses, including e-mail routing, marketing and sales automation, and business intelligence.
Enough with the egghead talk. As your grandfather might say, cut to the chase. Well, that's the point, according to Lynch. In an industry overloaded with text, Autonomy's software takes all that messy, unorganized data and streamlines it for large companies in a wide range of ways.
But, the real kicker is that the software hinges on mathematics — and is not limited by language or culture — thus, it truly has global, uniform appeal. "We realized that if you viewed it (software) as a mathematical problem ... you have a system that can deal with the real world," explains Lynch. "It can learn slang, it can learn ideas expressed in different ways. That was a magical moment."
Silicon Valley is virtually littered with the corpses of brilliant ideas that couldn't find a niche. Autonomy is not one of them. A first IPO in Europe netted $165 million, and a second offering in May raised another $109 million. The 4-year-old company already has a market value of more than $6 billion. Though Bayes' theorems may have lain dormant in Cambridge's musty halls for 250 years, all it took was a mathematician nearly as brilliant to find practical applications for it. Without the global proliferation of the Web, Lynch's inspiration might have met the same obscure fate.
"You may have a wonderful piece of technology. And then you realize that actually can count for very little," says Lynch. "...So, what is nice is when you do have a technology which is really of the next generation, and being able to have the infrastructure around it to actually exploit that and put that into place." Given his prompt success, it's not surprising Lynch imagines even greater benefits from the Internet. "I think it's actually more powerful than the Industrial Revolution," he says.
"(That) was about muscle and organization. Here, it's about leveraging something which is far more magical, which is the human brain."
Page:Economics, Creativity and Technology