The Edison Of The Internet Sun Microsystems mastermind Bill Joy is on a 20-year streak of innovation that has laid the groundwork for a new technological era--one in which businesses will farm out the hassles of owning computers and people will be more wired than ever. No wonder he's considered the smartest man in Silicon Valley.
By Brent Schlender

(FORTUNE Magazine) – It's a vintage Bill Joy moment. Here he is, co-founder and chief propeller-head of computer powerhouse Sun Microsystems, down on his hands and knees, fiddling with the audio and video gear that drives his home theater system. Having just spotted the twister bearing down on her Kansas farm, a stricken-looking Dorothy, Toto in her arms, stands paralyzed in freeze-frame on the projection-TV screen.

"Daddy, just let Dorothy play!" pleads his toddler daughter, a Wizard of Oz junkie.

Joy, who pauses to give little Maddie a quick tickle, has something else in mind, namely a demo. He's trying to show a visitor how a certain scene in the DVD version of the movie classic looks so much sharper than the same sequence on laserdisk that it actually reveals a subtle technical flaw in the projection TV itself. (Don't ask why. It's a geek thing.)

The only problem is, Joy owns so many different kinds of video players that he has to swap cables manually before he can show these video snippets back to back. "This is precisely why we invented Jini," he grumbles, referring to the radically new computer networking technology he developed for Sun. "My home entertainment system should be a network of components that just works. I shouldn't have to be a rocket scientist."

Which is, of course, just about what Bill Joy is. In fact, many consider him the smartest guy in Silicon Valley, this despite the inconvenient detail that he lives 750 miles away on a mountainside overlooking Aspen, Colo. You may never have heard of him, but mention his name to a programmer and you'll get a knowing nod: This towering, tousle-haired software wizard runs a close No. 2 to Microsoft's Bill Gates as the most influential person in the computer industry. As unassuming as Gates is rich, Joy, 44, has earned his reputation not by dominating markets or amassing tens of billions of dollars, but by dint of a remarkable 20-year record of invention born of late-night programming epiphanies, way-outside-the-box thinking, and uncanny technological clairvoyance.

Here's the short list of why he's a legend among computer cognoscenti. As a graduate student in the late '70s, Joy almost single-handedly shaped AT&T's Unix into a versatile, industrial-strength computer operating system, one that today is the chief competitor to Microsoft's heavy-duty Windows NT in corporate America. Just as important, Joy gave Unix the networking underpinnings that would ultimately foster the emergence of the Internet. Or how about this: While most programmers have only a rough idea how computer chips work, it was Joy, and not some chip architect, who designed the most crucial circuits in Sun's SPARC microprocessors, which are the brains of its entire $10-billion-a-year line of workstations and servers. Then there's Java. Five years ago he spotted the potential of an obscure but inspired programming language in Sun's labs and masterminded its transformation into a software lingua franca that has attracted a million practitioners, brought razzmatazz to the World Wide Web, and thrown a monkey wrench into Microsoft's plans for Internet dominance. We'll get back later to his newest brainchild, Jini.

Joy's accomplishments not only have rocketed Sun into the front ranks of a legendarily competitive industry but also have amplified the explosion of the Internet, both as a communications medium and as a social and business phenomenon. Indeed, of the "ten tech trends to bet on" enumerated in the preceding story, fully half bear the unmistakable fingerprints of computerdom's "other Bill." Says David Gelernter, a Yale computer scientist who is himself world renowned: "Joy is certainly one of the most influential people in the modern history of computing. He's a sharp merchant of ideas, and he's a deep thinker." Joy's boss, Sun CEO Scott McNealy, who admittedly is a mite biased, puts it another way: "AT&T has Bell Labs, and we have Bill Joy. We get a lot more for our money."

Aspen Smallworks is the name Joy gave his aerie in downtown Aspen, three blocks from the ski lifts (he complains he doesn't get enough time on the slopes). There he and a handful of Sun associates "incubate" their creations far from Silicon Valley distractions such as traffic jams, back-to-back meetings, and McNealy. Says Joy: "I moved out here in 1991 because back in Palo Alto, when my office was next to Scott's, I couldn't even think. Every ten minutes he'd bother me with another one of his crazy ideas."

Most of the crazy ideas come from Joy, actually. Get him riffing on network "architectures" and he turns into a motormouth in a vain attempt to keep up with his racing brain, which finds revealing parallels to his work in just about everything he hears, sees, or reads. A man of prodigiously broad interests, he can talk knowledgeably about Meso-American art or cattle ranching or photography or stock market "quant" theories. During a discussion of high-tech corporate cultures, for example, he dances deftly among references to Jungian psychology, the plays of Eugene O'Neill, and the writings of G.I. Gurdjieff to make a startling point. The best companies, he contends, have a wild side they can draw on in times of competitive stress--employees free to pursue creative tangents. "Some companies are just too well planned," he says. "I don't expect any great innovation to come out of Compaq or Dell, for example. They're in a business where every little bit matters. They have to plan down to the penny, so they don't have a wild side. But what will they do when the technology takes an unexpected turn?"

While he swims with the geeks, Joy is shrewd; he's forever finding worldly, even humanistic, ways to solve knotty problems of computer and network design, and spotting business opportunities where others just see technical barriers. "Bill Joy's point of view is so important because he is one of the few who can really see the big picture and yet understand all the nuances," says John Seely Brown, director of Xerox Corp.'s Palo Alto Research Center. And the delectable irony is that this postmodern polymath treasures one virtue above all others: simplicity.


"Explaining the new way of doing computers and networks is difficult. It's like expecting people in a feudal society to know what it would be like to live in a capitalistic society. It's just too hard to imagine until you're in it." --Bill Joy

Joy's sometimes radical notions about where computers and networks are headed have steadily gained currency--so much so that, as with the theory of evolution, it's hard to imagine a time when they didn't seem plausible. This vision puts much more emphasis on how computers are networked than on the computers themselves, because, Joy argues, once computers are in a network, many assumptions about how they and their software are designed go out the window.

Just about everybody knows what owning a computer today entails. It sits on your desk like a monument to Progress. To use it, you have to conjure with the operating system, inscrutable software that lets you tap applications and that manages the peripherals attached to the machine--its keyboard, monitor, printer, disk drives, speakers, modem, network adapter card, and so on. Each computer is its own complete data-processing world.

The growing power and usefulness of computers have fueled Silicon Valley's spectacular growth. But for users, computers' growing complexity creates headaches: For one thing, you must frequently upgrade costly hardware and software just to keep pace. For another, operating systems have become dizzyingly complicated. Microsoft's upcoming Windows 2000, which replicates much of what Unix does and then some, will contain upward of 30 million lines of code, by some estimates, when it comes out this year. Operating systems this huge are impossible to debug fully, so computers crash a lot. Joy calls the spiraling complexity putting "Star Wars-scale software on the desktop."

Joy's obsession has been to devise ways for the network to absorb the complexity--so that, unbeknown to the user, computers and other devices on the net would recognize each other and serve up data and programs the user needs. "It was clear to Bill even way back in the 1970s that there was an alternative architecture for computing somewhere 'in the net,' " says Eric Schmidt, an old graduate school buddy, now CEO of Novell. For years Joy groped for a way to articulate his idea. Finally, in 1988, John Gage, another old Berkeley friend and director of Sun's science office, coined a Zen-like slogan that got the gist of Joy's vision: The network is the computer. (Translation for users: The network you're connected to, not the box on your desk, will do the job you want done.)

Indeed, the full implications of what might be possible under Joy's vision have become clear only in the past few years, as the boom in large corporate networks, online services like AOL, and, of course, the Internet has shown just how powerful networks can be. Surfing the World Wide Web is just the beginning. If Joy's vision holds sway, self-contained mainframes, servers, and PCs will give way to specialized computers and other smart gadgets that cooperate and share information. We'll still have lots of computers on our desks, but the network will be what's important: a meta-world in which much of our business activity and communication and daily routines take place.

For companies large and small, that will mean they'll no longer have to design and oversee--or own--so much of their infotech infrastructures; instead they'll be free to subscribe to "utility" services via the Internet. They'll unplug many of their mainframes and servers, and their main hardware purchasing decisions will be which PCs or terminals to put on people's desks and in their pockets.

For individuals, Joy's vision points to the advent of the "Personal Network." You'll start taking for granted being online during many of your waking hours, via all sorts of electronic devices on your desk, at home, in your car, and in your pocket. Through them, you'll be a simultaneous participant in a hierarchy of digital networks--the one at work that helps you do your job; the one at home that entertains you, lets you shop, helps run your household, and keeps you connected with family and friends; and the wireless networks standing by to help when you're on the go. Says Joy: "We'll stop thinking of our computers or even the network as a thing, and instead we'll view them as a collection of services and places to get things done. Your computers will be like your TV or your telephone--pieces of plastic and metal that let you transcend time and space and the physical world."

Best of all, you won't have to be a rocket scientist to take advantage of all this. Everything will just work.


"When Bill was a little kid, he wanted to know everything about everything way before he should've even known he wanted to know. We answered him when we could. And when we couldn't, we would just give him a book." --William Joy, Bill's father

It's hardly a surprise that as a schoolboy, Joy was a classic Eisenhower-era brainiac--and then some. Born in Michigan in 1954 and the eldest of three children, he started reading at 3, entered kindergarten at 4, took on advanced math assignments at 5, and soon started skipping grades. "We always worried about his social development because he was younger than the other kids," says his dad, a schoolteacher who later became a business professor and stockbroker. "But he seemed happy, and he just flew through school." At 13, Joy memorized the periodic table of the elements in a single evening. When he graduated at 15, he was voted North Farmington High's Most Studious Student. "In other words," he says, "I was a no-date nerd."

He discovered computing at the University of Michigan. His employer after class was a professor who was experimenting with one of the first "parallel" supercomputers, which used several microprocessors arrayed in a kind of tight network. When the time came for Joy to go to graduate school, three computer science meccas wanted to snap him up: Stanford, Caltech, and the University of California at Berkeley.

Joy's choice was an early sign that this nerd was unusually canny. He picked Berkeley, not because it was cheaper or because it had been a hotbed of anti-Vietnam protests (the year was 1975, but Joy was apolitical). "I went to Berkeley because it had the worst computer facilities of the three," he says. Huh? "I figured it would force me to be more ingenious."

The techie world first became aware of Joy soon after he got to Berkeley. He and some colleagues in the computer science department poked under the hood of a Digital Equipment computer and repaired some bugs in its Unix operating system. They compiled their fixes on a computer tape and started offering copies to other universities that owned similar machines, charging $50 a copy to cover their costs.

The fledgling venture was oddly in sync with the computer subculture that was aborning. (The only "personal computer" then available was a mail-order kit called the Altair; kids who loved computers usually cadged time on university and business minicomputers and mainframes.) In 1976, Berkeley bought another Digital Equipment machine, a new model called a VAX. Soon Joy and his buddies worked up their own version of Unix for it too. It was more elaborate than their first offering, so Joy raised the price to $300.

The VAX quickly became a hot seller to universities and research labs, even though each system cost $200,000 or more. One way to save money was to buy a stripped-down VAX without any software or disk drives, and add cheaper versions from third parties. Joy, who was making a name for himself as the department's cleverest entrepreneur, wrote a memo that circulated widely among academic computer centers called "How to Buy a VAX," relating how much money he'd saved doing just that.

Orders for Berkeley Unix came rolling in by the hundreds. Joy realized he was really on to something when, in 1978, his Unix team beat out some of Digital's own programmers for a contract from the Pentagon's Defense Advanced Research Projects Agency (DARPA). The assignment was to develop software to enable VAXes to hook into something DARPA called the Internet. It gave Joy the incentive to learn about and perfect the networking software protocols he had always wanted to put into Unix.

Given the success of Berkeley Unix, it was inevitable that the computer industry would come a-knocking at Joy's door. It finally happened in 1982, when Vinod Khosla, a newly minted Stanford MBA, brought his classmate Scott McNealy and a Stanford electronic engineering whiz named Andreas Bechtolsheim to Joy's crowded office. Khosla wanted to start a company to commercialize one of Bechtolsheim's creations, a cheap but powerful desktop computer called a Stanford University Network, or S.U.N., workstation. They needed a Unix software hotshot and had been referred to Joy by just about everyone they asked. "You should have seen Andy and Bill when they first met," recalls McNealy. "Vinod and I just sat in the corner and read magazines while they did their geek bonding routine. It was a Vulcan mind meld."

To Joy, Sun represented a fresh opportunity to zag instead of zig. Not only was he frustrated with limitations imposed by the university on his budding Unix business (he couldn't get enough office space), but he hadn't made much progress on his Ph.D., even though he'd been at it for more than six years. It didn't hurt that the workstation Bechtolsheim wanted to design was Bill Joy's dream machine: a powerful desktop computer made to be networked. Within a few weeks, he signed on.

Sun was an almost instant success, even though its workstations faced a crowded field. The computers weren't the fanciest or the fastest, but they were cheap, well built, and reliable, thanks in large part to Joy's Unix software. (Later Sun would rename it Solaris.) Within six years, Sun's revenues passed the $1-billion-a-year mark, a record pace for a Silicon Valley startup that stands to this day.

Success brought wealth Joy hadn't bargained on. After Sun went public in 1986, his stake was worth more than $10 million, a sum that "seemed like more money than I would ever be able to spend," he recalls.

Being possibly the smartest geek in Silicon Valley did not, however, cause him to respond wisely. The shares burned a hole in his pocket; he thought about seeing his father's clients lose a bundle in the tumultuous stock markets of the previous decade; so he sold most of his stock within a couple of years. That decision cost Joy literally hundreds of millions of dollars in forgone profits: Co-founder McNealy's Sun holdings are now worth nearly $1 billion, but Joy's aren't significant enough to be reported in Sun's public filings. That's not to say he is destitute. A close friend is John Doerr, a partner with Kleiner Perkins Caulfield & Byers, Silicon Valley's most famous venture capital firm, which bankrolled Sun. In recent years, Doerr signed up Joy as a limited partner in some of Kleiner's most lucrative funds. Asked to estimate his net worth, all Joy will allow is that "it's a lot less than most people guess, but plenty for me."


"Bill always leads with the power of his ideas, not the power of his position." --John Doerr

Through the late '80s, Joy took on more and more responsibility at Sun, working as chief scientist and director of technology. He remembers having to worry about meeting budgets, approving or killing research projects, hiring and firing, refereeing turf wars, and, of course, warming a chair in lots of meetings. Sure, he was able to help Sun establish some truly significant networking and Unix software standards for the entire computer industry, and he got to stick his fingers in chip designs. But the job began to chafe. Things just weren't right.

He thought moving from Silicon Valley to San Francisco might help. He bought an apartment there and got into the cosmopolitan scene, even putting up money for a friend to open a gallery that sold primitive art. But the move just created a longer commute for a man who doesn't like to drive. After the October 1989 earthquake wrecked San Francisco's freeways, it was even worse.

The quake jarred Joy into thinking it was time to zag again. He fantasized about leaving Silicon Valley altogether, moving to a town where the scale was small, taking a demotion and minimizing his direct reports, keeping his own hours and following his whims. "I concluded that you're best off being away from the action sometimes. It was sort of like: 'Do you rule in hell or serve in heaven?' " he recalls.

Trading the office adjacent to the CEO's for life in a plush ski resort may sound like a no-brainer. What's more, McNealy didn't veto the idea. All the same, there was substantial career risk. Joy could easily lose touch with the marketplace and weaken his clout with Sun's senior management. His little research team could become ineffectual: Everybody in Silicon Valley knew the cautionary tale of Xerox PARC, which in the late '70s couldn't get its parent back East to cash in on amazing inventions like the graphical user interface and the Ethernet.

Even so, change was what Joy needed, and he's been on a creative roll ever since. If anything, his influence over the company is stronger: He's Sun's shaman on the mountaintop. The best example of his new fecundity is Java, the Internet-friendly programming language. While Joy didn't write Java, it was his idea in the mid-1980s to assign crack programmers to design a concise language--then called Oak--as part of a failed effort to pioneer interactive TV. Joy reasoned that such a language could make it easier to network consumer-electronics equipment and portable gadgets that couldn't pack the power of a PC.

Oak was ahead of its time; interactive TV failed to materialize, and Oak's demoralized authors moved on to other projects. But when the World Wide Web began to blossom in 1994, Joy realized Oak was the perfect technology to make the Internet even more interactive. It could be used to write small, efficient programs that could be sent 'just in time' via the Web to run on just about any kind of computer. Today, Web retailers like Levi-Strauss use Java applications to take orders.

Out in Aspen, Joy had time to map out technical and marketing strategies for Oak, plans that were so well articulated that McNealy couldn't say no. Everybody in the computer industry knows the rest of the story. Java caught on like wildfire, until even Microsoft felt obliged to adopt it. Sun set up a subsidiary to promote the language; IBM assigned more than 1,000 programmers to work with it; Kleiner Perkins started a $100 million fund to back startups willing to write corporate Java applications; universities offered Java programming classes because students demanded them; and Joy became a sought-after speaker on the computer industry conference circuit.

"Java shows how even a little Bill can take you a long way, because he finds the shortest distance between a question and an insight," says Eric Schmidt, who recently asked Joy to join the Novell board. "If you can get Bill to give you that insight, you can save months or even years of trying to get something to work."

Joy's latest insights have to do with the problems he encountered with his home theater system, and his wish to connect a lot more than just computers. The upshot is the technology he named Jini (pronounced "genie").

Jini makes real the possibility that just about anything electronic--printers, TVs, CD players, digital cameras, you name it--can connect via the Net. So can wireless devices like cell phones and PDAs. So can appliances, if equipped with Java chips--simple, low-cost processors already being sold by Sun. Jini consists of simple programs written in Java that enable any gizmo connected to a network to announce its capabilities and help other devices use them. A "Jini-enabled" photo printer, for example, would tell other machines what sorts of digital pictures it stands ready to receive and print; a Jini-enabled light switch hooked to a household network might announce, in effect, "I'm the living room lights! I'm OFF right now! Here's how to turn me ON!"

Sun has persuaded dozens of companies to support Jini, including Sony, Philips, and other consumer electronics powers, as well as disk-drive suppliers and computer equipment makers like Xerox. The first Jini-enabled printers, disk drives, and other computer devices are slated to appear this year. But some of the most fascinating advances will come as Jini enters the wireless world. Say you carry a Jini-enabled PDA when you go to the airport. As you enter the terminal, it might alert wireless systems there that you've arrived, confirm your seat assignment, and receive and display your gate and departure information. Automatically.

Joy is proud of Jini because, more than any other technology he has helped foster, it simplifies the network for ordinary people. He's proud of Jini for managerial reasons too. It shows the folks back at headquarters that Sun's "wild side" really does deliver the kind of innovation that can keep the company growing. In the overcaffeinated Generation-X crazy world of Silicon Valley, what does Sun's wild side actually consist of? The team Joy assigned to develop Jini was a small group of fortysomething men and women in Massachusetts. Says Joy, "It's nice to show the world that just because you're an aging baby-boomer and work outside Silicon Valley doesn't mean you're over the hill."

Right on.