by Steven Levy
Roberta talked to her mother about it: “He’s going to go someplace,” she said. “To really make it. Be something.”
Finally Ken told her, “We’re getting married, and that’s it.” She didn’t fight it. She was nineteen; he was a year younger.
Within a year, Roberta was pregnant, and Ken was pulling Ds and worrying about supporting a family. He knew from reading the want ads that there were a lot more jobs in computer programming than there were in physics, so he figured, just like it said on the matchbook covers, that he would find a career for himself in electronic data processing. Roberta’s dad cosigned a student loan for $1,500, tuition for a trade school called Control Data Institute.
The world Ken Williams was entering was nothing like the holy preserve of the MIT AI lab. His would-be colleagues in the business computing field had little of the hands-on hunger that drove the class of Altair graduates who hacked hardware. In the early 1970s the business computer field that Ken was entering was considered the creepiest in America. It was a joke, an occupation where meek little moles did things—who knows what those things were?—to the punch cards and whirring wheels of Hulking Giant computers. As far as the public was concerned, there wasn’t even much difference between the drones who mechanically punched the cards and hammered at the keyboards, and the skilled technicians who programmed the machines to put the cards in their places. They were all seen as the white-shirted, Coke-bottle-glasses moles in the computer room. Creatures of the disembodied age.
If Ken and Roberta had been part of a wide circle of friends, they might have had to confront that stereotype, which Ken did not resemble in the least. But Ken and Roberta did not bother to put down roots or establish close friendships. As a computer programmer, Ken was less a Richard Greenblatt or a Lee Felsenstein than he was Jonas Cord. Later, he would jauntily say, “I guess greed would summarize me better than anything. I always want more.”
Ken Williams was far from a dazzling programmer when he finished Control Data Institute, but he was certainly prepared to do anything required of him. And more. As much work as possible, to help him go as high as he could. Then take on another, more demanding job, whether or not he was qualified. Instead of cleanly breaking with the previous employer, Ken tried to keep on the payroll, in consultant mode.
He would claim to know computer languages and operating systems he knew nothing about, reading a book about the subject hours before a job interview and bullshitting his way into the position. “Well, we’re looking for a programmer in BAL,” they would tell him, referring to an esoteric computer language, and he would laugh almost derisively.
“BAL? I’ve been programming in BAL for three years!”
Then he would immediately rush out to get hold of some books, since he had never even heard of BAL. But by the time the job started he would have procured documentation, uniformly buried in dense, cheaply printed loose-leaf manuals, to fake expertise in the “BAL environment,” or at least buy time until he could get into the machine and divine the secrets of BAL.
No matter where he worked, in any number of nameless service companies in the yawning valley above Los Angeles, Ken Williams did not meet one person who deserved an iota of his respect. He would observe people who’d been programming computers for years and he would say to himself, “Give me a book and in two hours I’ll be doing what they’re doing.” And sure enough, stackloads of manuals and a few fourteen-hour days later, he would at least appear to be one hotshot programmer.
He’d come into the heavily air-conditioned computer sanctums at weird hours of the night to fix a bug, or get the computer back up when one of his programs accidentally fed on itself and tripped the millions of calculations up in such a fury of misunderstanding that nothing the regular crew could think of could revive the machine. But Ken, confident that the stupidity of his colleagues was dwarfed only by the astounding compliance of the Dumb Beast whom he could feed and befriend with his programming skills, would work three days straight, forgetting to even stop for a meal, until the Dumb Beast was back on the job. Ken Williams, hero of the day, tamer of the Dumb Beast, would go home, sleep for a day and a half, then return to work, ready for another marathon. Employers noticed, and rewarded him.
Ken was rising at quantum speed—Roberta figured they moved to various locations in the L.A. area about twelve times in that go-go decade, always making sure that they turned a profit on the house. They had no time for making friends. They felt like loners and misfits, usually the only white-collar family in a blue-collar neighborhood. The consolation was money. “Wouldn’t it be nice to make another two hundred dollars a week?” Roberta would ask, and Ken would get a new job or take on more consulting work . . . but even before Ken had settled into this new job, he and Roberta would be sitting in the tiny living room of whatever house they happened to be living in, and saying, “Wouldn’t it be nice to earn two hundred dollars more?” The pressure never stopped, especially since Ken Williams had idle dreams of fantastic sums of money, money enough to goof off with for the rest of his life—not only all the cash that he and Roberta could spend, but all that his kids could spend, too (Roberta was pregnant by then with the second Williams son, Chris). Wouldn’t it be nice, he thought, to retire at thirty?
By then something else was changing: his relationship with the Dumb Beast. When Ken had time, he would often pull out some of those dense, cheaply printed looseleaf manuals, trying to figure out what made the big Burroughs or IBM or Control Data machine really tick. As he gained proficiency in his profession, he began to respect it more; see how it could approach art. There were layers of expertise that were way beyond what Williams had previously come to assume. A programming pantheon did exist, almost like some sort of old-time philosophical brotherhood.
Ken had gotten a taste of this more exotic realm when he fast-talked his way into a job as systems programmer for Bekins Moving and Storage. Bekins was switching then from a Burroughs computer to a bigger and slightly more interactive IBM machine. Ken baldly fabricated a career history of IBM wizardry for himself, and landed the job.
At Bekins, Ken Williams became hooked on pure programming. His task was installing a heavy-duty telecommunications system on the IBM that would allow one computer to support eight or nine hundred users in the field across the country, and the problems and complications were beyond anything he’d confronted so far. He would experiment with three or four languages that had nothing to do with his job, fascinated with the techniques and mind-frames required with each language. There was a whole world inside this computer . . . a way of thinking. And maybe for the first time Ken Williams was being drawn to the process of computing more than to the goal of completing a task. In other words, hacking.
As a consequence of his sustained interest, Ken remained at Bekins longer than at most of his other employers: a year and a half. It was time well spent, since his next job presented him with an even greater challenge, as well as contacts and ideas which would soon enable him to act out his wildest fantasies.
• • • • • • • •
The company was called Informatics. It was one of a number of firms that sprang up in the mid-sixties to take advantage of a gap in the mainframe computer software field. More and more big companies and government agencies were getting computers, and almost none of the software that the behemoth computer companies supplied could artfully execute the tasks the computers were supposed to perform. So each company had to hire its own programming staff, or rely on highly paid consultants who invariably would disappear just when the system crashed and valuable data came out looking like Russian. A new team of programmers or consultants would then come out to untangle the mess, and the process would repeat itself: starting from scratch, the new team would have to reinvent the wheel.
Informatics and companies like it were set up to sell software that made the Hulking Giants a little more comprehensible. The idea was to invent the wheel once and for all, slam a patent on it, and sell it like crazy. Their progra
mmers would toil away at the assembly level and finally come up with a system that would allow low-level programmers, or even in some cases nonprogrammers, to perform simple computer tasks. After all, these commercial systems all did pretty much the same thing—you had something coming in from a clerk or a branch office on paper which got keypunched and entered into a system which modified some preexisting file. Informatics came up with a pre-programmed system called Mark 4. Sometime in the seventies it became the largest selling mainframe computer software product of all time, approaching at one point $100 million in yearly revenue.
In the late seventies, one of the managers in charge of Informatics’ new products was Dick Sunderland, a former FORTRAN programmer who was climbing the corporate ladder after reluctantly foregoing a late-in-life stab at law school. In place of the law, Sunderland had determined to pursue a romance with a bright and holy concept of management. To be a leader of men, a deft builder of competent, well-meshed employee teams, a persuasive promoter, and a constructive manipulator . . . this was what Dick Sunderland aspired toward.
A small, chalk-complexioned man with hooded eyes and a contemplative drawl, Sunderland considered himself a natural manager. He had always been interested in the advertising, selling, promoting of things. Psychology fascinated him. And he was especially enamored of the idea of choosing the right people to work together so that their joint output dwarfed the measly sum of their individual inputs.
Dick was trying to do that at Informatics with his new product team. He already had one genuine wizard on the staff, a lean, quiet man in his forties named Jay Sullivan. Jay was a former jazz pianist who had come to Informatics from a more mundane job in his native Chicago. He later explained why: “Systems software [at Informatics] was much more interesting. You didn’t have to worry about mundane things like applications or payrolls. It was much more real programming to me; you dealt more in the essence of what programming was about. The actual techniques of programming are more important than the specifics of the job at a specific time.” In other words, he could hack there.
In his programming, Sullivan worked like a vacationer who, having planned his trip carefully, educating himself on the subtle characteristics of the local scenery, followed the itinerary with enhanced consciousness. Yet he still retained the curiosity to stray from the plan if circumstances seemed to call for it, and derived pleasure from the careful exploration that such a fork in his path would involve, not to mention the sense of accomplishment when the detour proved successful.
As with many hackers, Sullivan’s immersion in programming had taken its social toll. Sullivan later explained that with computers “you can create your own universe, and you can do whatever you want within that. You don’t have to deal with people.” So while he was a master in his work, Sullivan had the infuriating kind of programmer personality that led him to get on splendidly with computers but not pay much attention to the niceties of human interaction. He would casually insult Dick, and nonchalantly go about his business, doing brilliant things with the operating system, but often would see his innovations die because he was not adept at politicking, a process necessary at the large company. Dick Sunderland had forced himself to be patient with Sullivan, and eventually they had arrived at a seller-inventor relationship which produced two lucrative improvements to the Mark 4 line.
Dick was looking for more master programmers, calling recruiters and making it quite clear that he was looking for cream-of-the-crop people, nothing less. One recruiter mentioned Ken Williams to him. “This kid’s a genius type,” the recruiter said.
Sunderland called in Ken for an interview and made sure that his true genius, Jay Sullivan, would be there to test the mettle of this Williams person. Dick never before had seen anyone stand toe-to-toe with Jay Sullivan, and was curious to see what might come of the interview.
Dick and Jay were talking about a problem in implementing a new, user-friendly language that Informatics was working on when Ken showed up, wearing slacks and a sport shirt which fit so badly that it was obvious T-shirts were his norm. The discussion had been fairly technical, focusing on the problem that to make a language a nonprogrammer would understand—a language like English—one would have to avoid any kind of ambiguous words or acronyms.
Suddenly Jay Sullivan turned to Ken and said, “What do you think of the word ‘any’?”
Without hesitation, Ken correctly asserted that it was a very valuable word, but an ambiguous word nonetheless . . . and then extemporaneously tossed off ideas about how that word might be handled.
It seemed to Dick that he was witnessing a classic battle—the cheeky Pomona Kid versus venerable Chicago Slim. While Ken had a charismatic quality to him, and obviously knew computers, Dick still had his money on Jay. Jay did not let him down. After Ken stopped, Jay, speaking quietly and methodically, “sliced Ken up with a razor blade,” Dick later recalled, enumerating the errors and incompleteness of Ken’s thoughts. Yet it was impressive to Dick—and even to Jay—that this college dropout could even think such thoughts. What’s more, rather than being dissuaded by Jay’s broadside, Ken came right back. Dick watched the two pick up threads of each other’s ideas and weave them into more refined concepts. This was synergy, the manager’s holy grail, Dick decided to hire Ken Williams.
Dick put Ken under Jay’s supervision, and the two of them would chatter about programming arcana for hours. For Ken it was an education: he was learning the psychology of computerdom in a way he never had. Of course, one part of the job that Ken Williams did not like was having a boss; Ken in this regard was a typical antibureaucratic hacker. So he came to dislike Dick, with all his schedules and fixation on managerial details—obstacles to the free flow of information.
Ken and Jay would be talking about the intricacies of some aspect of programming language—like trying to figure out, when somebody says “List by customer,” what that really means. Does it mean “SORT by customer,” or perhaps “List ALL customers”? Or maybe “List ANY customers”? (That word again.) The computer had to be programmed so it wouldn’t screw up on any of those interpretations. At the very least it should know when to ask users to clarify their meaning. This took a language of considerable flexibility and elegance, and though Ken and his new guru Jay might not have said it out loud, a task of that sort goes a bit beyond technology and into primal linguistics. After all, once you get waist-deep into a discussion about the meaning of the word “any,” it’s only a short step to thinking philosophically about existence itself.
Somewhere in the midst of one of these conversations Dick would come in, eager to witness some synergy among his troops. “We’d try to supersubset it so that a two-year-old would understand, ask Dick’s opinion, he’d give it, and we’d chase him out of the room,” Ken later recalled. “Dick never understood what we were putting up. He was obviously out of his league.”
At those times Ken might have felt superior to Dick, but in retrospect he had to admit that Dick was smart enough to recognize talent. Ken realized that he was one of the weaker members of a superteam of programmers who were doing great stuff for Informatics. Sometimes Ken figured that Dick must have gotten lucky, accidentally corralling five of the most creative people around for his new products team. Either that or he was the best manager in the world, or at least the best talent evaluator.
Ken, always needing more money, began moonlighting. Sunderland was refusing his constant requests for raises, and when Ken suggested that he might like to head a programming group, Dick, a little astounded perhaps at the chutzpah of this brilliant but scattershot kid, flatly denied the request. “You have no talent for management,” said Dick, and Ken Williams never forgot that. Ken was regularly going home to Roberta and complaining about Dick—how mean he was, how strict, how he had no understanding of people and their problems—but it was less a dissatisfaction with his boss than his desire for more money, money for a bigger house, a faster car, a CB radio, a motorcycle, a hot tub, more electronic gadgets, that led h
im to double and even triple up on work, often phasing into a no-sleep mode. Eventually the outside work got to be more than the inside work, and he left Informatics in 1979, becoming an independent consultant.
First there was a guy with a scheme to do tax returns for big companies like General Motors and Shell, and then there was some work with Warner Brothers, programming a system for the record company to keep artists’ royalties straight. There was a bookkeeping system he constructed for Security Pacific Banks, something about foreign tax plans. Ken was becoming a finance guru; the thirty thousand a year he was pulling down looked to be only the beginning, if Ken kept hustling.
He and Roberta began weaving a little fantasy. At night—the nights Ken wasn’t out consulting for someone—they would sit in the hot tub and talk about splitting the Simi Valley suburban trap and moving to the woods. Where they would go water skiing, snow skiing . . . just goof off. Of course there weren’t nearly as many hours in a day to make money to turn that kind of trick, no matter how many companies Ken set up tax programs for. So the fantasy was just that, a fantasy.
Until Ken’s little brother Larry got an Apple Computer.
Larry brought it over to Ken’s office one day. To Ken, who had been dealing with telecommunications networks that handled two thousand people all at once, who had invented entire computer languages with mainframe wizards the likes of Jay Sullivan, the idea of this sleek, beige machine being a computer seemed in one sense ludicrous. “It was a toy compared to the computers I’d been using,” he later explained. “A piece of junk, a primeval machine.”