Jim Clark

It didn't take long for Clark to realize that Netscape's dominance of the browser market wouldn't last long, not with Bill Gates on the scene. Already, Microsoft's own Internet browser was whittling away at Netscape's market share. And because Microsoft's Windows enjoyed a near monopoly in PC operating systems, it would only gain more, Clark feared. So what did he do? He began looking for new ways to make money.

He soon turned his eye on the U.S. health-care market. In the early 1990s, he had been in a motorcycle wreck that crushed his leg. Later, he was diagnosed with a blood disease that required regular trips to the doctor. In both cases, Clark, like virtually all American health-care consumers, found himself frustrated with the long waits and complicated forms. Why not use the Internet to eliminate paperwork by linking all the health-care players -- the doctors, patients, pharmacies, health plans and benefits administrators -- to a central depository of information? Clark's company, dubbed Healtheon, would control that depository and collect a few pennies of each transaction. Considering that the $1.5 trillion health-care sector is the largest component of the U.S. economy, Healtheon could make billions, even if it only handled a fraction of those transactions.

Like most of Clark's ideas, this too was an audacious notion. Neither Clark nor his engineers claimed to understand the severe problems faced by the byzantine U.S. health-care system. But the idea had a certain elegance to it. It also was a good story. And most important, it belonged to Clark. As a result, venture capitalists who were shut out of Netscape were slavering over the idea.

Healtheon had a rough start. The software was late to the market, and the nation's health-care companies were less than eager to embrace new technologies. An initial attempt to go public was aborted amid last fall's stock market swoon. But by February, the markets had recovered and investors had regained their appetite for promising Internet companies that still happened to hemorrhage cash. And so it was that Healtheon saw its share price shoot to $33 and Jim Clark became the first person in the world to create three different multibillion-dollar technology companies.

Healtheon's future remains murky. With its purchase of Web MD, the company has become a comprehensive source of health-care information, the real money lies in commerce. Yet so far, only a handful of the nation's large insurance companies have adopted sophisticated computer systems and many still process claims by hand. Most doctors' offices are even less technically adept. The act of uniting them all in a cohesive Web seems daunting, to say the least.

Clark, characteristically, had little interest in hanging around. "Having articulated the new new thing," Lewis writes, "Clark intended to return to the important work of teaching his computer to sail his new boat."

That would be the Hyperion, a 155-foot, $30 million sailboat with a 197-foot mast, designed to be controlled by 25 high-powered Silicon Graphics workstations. Originally conceived as the basis for a new business that would create so-called smart homes whose every function was controlled by computers, the project has obsessed Clark -- and baffled those who know him -- for years. Eventually, he hopes to be able to control the Hyperion from his living room, via the Internet. That's some ways off. Last spring, on the boat's first transatlantic voyage, the on-board computers inexplicably shut down the engine. Unable to fix the glitch while at sea, the crew was forced to complete the journey the old-fashioned way.

Clark, not surprisingly, has shown no willingness to give up. And whether the boat actually will be sailed successfully by a computer seems almost beside the point. Instead, Clark seems drawn to Hyperion precisely because of the impossibility of ever pulling it off. Will he ever do it? Every instinct argues no, that perhaps there are some places where computers and the Internet truly do not belong. On the other hand, considering all that's happened over the past 18 years, it's very difficult to bet against Jim Clark.

• View blog reactions