Chapter 4

What’s Next?

In the balmy autumn days that followed his departure from Apple, Steve and his renegades gathered at his house in bucolic Woodside, a horsey community west of I-280 and nestled in a valley whose western slope swoops up to the coastal range of low mountains that protect Silicon Valley from the Pacific. The house, which he’d bought in 1984, was perhaps the most extravagant, albeit typically peculiar, nod to his rock-star status. It had been built by another controversial innovator, Daniel C. Jackling. In the early 1900s, Jackling had pioneered open-pit mining, the highly efficient yet disastrously polluting method for reaping low-grade copper that is still used throughout the world. Like Steve, Jackling made a bundle developing his idea, and the Woodside house, designed in Spanish Colonial style, was his personal monument. The rambling 17,000-square-foot structure had fourteen bedrooms, custom-made wrought-iron lamps, and a pipe organ that had been expanded to seventy-one ear-shattering pipes. Charles Lindbergh and Lillian Gish had been feted there at parties that flowed forth from the enormous ballroom. The driveway leading to the mansion showed off its ample landscaped grounds, which had fallen into some disrepair. Some of Steve’s indulgences—a BMW motorcycle and a gray Porsche 911—were parked out front. On the inside it didn’t feel much like a home at all. Steve hadn’t gotten around to buying much furniture. Strewn about the house were a mattress, a lamp, some Ansel Adams prints. He’d bought a monstrously big house, but had settled in without bothering to make it a home.

The apostolic high-tech renegades from Apple—Rich Page, a hardware engineer and Apple fellow; Bud Tribble, a leading software programmer; George Crow, another hardware specialist; Dan’l Lewin, who had led Apple’s efforts to sell Macs to universities and colleges; and Susan Barnes, the Mac financial manager—converged here with their rock-star leader every few days to plot the next revolution. From the very start, the world was peeking in through the windows. For the Newsweek cover story on Steve’s departure from Apple, the magazine photographed the six of them outside his house, awkwardly sitting on the lawn in their “business” clothes. (Dan’l Lewin, the chiseled Princeton grad who would lead the sales team, was even wearing a tie.) “It’s hard to think that a $2 billion company with 4,300 employees couldn’t compete with six people in blue jeans,” Steve told the reporter. His humility, of course, was utterly disingenuous.

When he resigned, Steve had told the Apple board that his new company, dubbed NeXT, would not attack any important Apple markets. That was nonsense. His stated target—the higher education market—was in fact very important to Apple, and he had stolen away Lewin, the company’s key sales link to the academic world. But Steve had his eyes on even more than that one narrow slice of Apple’s potential market. As Steve saw things, he had created the first two signature moments of the personal computing era: the Apple II and the Mac. (There had, of course, been another signature moment: the release of the IBM PC in 1981. But Steve discounted that milestone, since he couldn’t imagine a world in which most people would choose to buy machines that were so much harder to use than the ones he created.) Now it was time for a third shift, and he naturally would be the one to drive the change. He would show those bureaucrats who had mismanaged Apple a thing or two about real leadership and innovation.

Steve believed that he now had everything he needed to succeed as a world-class CEO. He had been involved in every aspect of Apple’s business over the previous eight years. He was a quick study, a savant who could envision revolutionary products and inspire the close group of folks who designed and made them, and he was an instinctive marketer. In Steve’s eyes, no one could say the same for Sculley and his “market-driven” tactics. “I think the question everyone is asking about Apple is this,” Steve told me during one of our first interviews. “Does the environment to create the next Macintosh still exist at Apple? Would they know it if they found it?” His new company, called NeXT, would surely grow to be even bigger than Apple, for the sole reason that he had higher expectations for it. “The world doesn’t need another $100 million computing company,” he announced, scoffing at the thought that he might produce something so trivial.

He was convinced that he really was the only person who could create from scratch the amazing blockbuster products that would give rise to the industry’s next great company. So were his renegades. “I had had plenty of experience with the downside of Steve,” remembers Lewin, the last of the five Apple exiles to sign up. “I definitely thought about the risk of going to work for him and leaving my job at Apple. But I worried that if I didn’t go to NeXT, I would have always said, ‘Dammit, I should have gone along for the ride.’ ” Says another early staffer, who signed on in 1986: “You would have had to be an idiot not to believe that Steve was going to create the next big thing. Everyone believed that.”

Little did they know that in due time, NeXT would turn out to be the full, unfortunate blooming of Steve Jobs’s worst tendencies at Apple. Yes, Steve had been a product visionary and a great spokesman for the company and the industry he had helped create. But he was hardly poised to be a great chief executive. In many ways, he wasn’t even a grown-up yet.

At the very moment when Steve had convinced himself that he had won a richly deserved freedom from an oppressive, dull overseer, he was in fact slave to so much else: to his celebrity, to his unbalanced and obsessive desire for perfection in the most innocuous of details, to his managerial flightiness and imperiousness, to his shortcomings as an analyst of his own industry, to his burning need for revenge, and to his own blindness to these faults. He was immature and adolescent in so many ways—egocentric, unrealistically idealistic, and unable to manage the ups and downs of real relationships.

Steve was too self-centered to see how much of Apple’s success had depended upon a combination of perfect timing and the work of others. Nor did he recognize how much he had contributed to its many problems. He didn’t realize how little he had truly absorbed from his crash course in business. Steve had been the titular CEO of Apple for only a few brief months at its inception, before Mike Scott was hired, and actually knew quite little of the true demands of corporate leadership. He was smart enough to realize that a successful CEO must prioritize among his employees’ many projects and ideas, but it would take years for him to learn how to do so efficiently or without the ego that came with thinking that his own ideas were always best. Nor did he have any real knowledge of how to launch a company into a field crawling with competitors. And he was unaware of each and every one of these weaknesses.

One afternoon that fall, the wind kicked up outside the Woodside house during an early meeting. “The doors were slamming,” remembers Barnes, who served as NeXT’s CFO, “just opening and shutting, opening and shutting in the wind. It was driving Steve crazy. And I could see, there was a piece of him that wanted to turn on one of us and just take us out. But it was his house. I wasn’t in charge of facilities, like I was at Apple! So, hey dude, it’s your house, you’re the one who’s got the door-slamming problem, not me.” Jobs, it seemed to Barnes, had no clue about the hundreds of little things other people had been doing over the years to keep Apple afloat while he was dreaming up his big ideas. Now he had to learn. “When you’re the CEO and the funder,” she says, recalling that afternoon, “everything’s on your shoulders.”

UNLIKE IN 1975, when he and Woz had pioneered the personal computing industry, in 1986 Steve was trying to enter a hypercompetitive marketplace with such a wide range of offerings that any newcomer would be hard-pressed to offer anything truly unique. Computer technology, drafting on the amazing multiplicative power of Moore’s law, had made tremendous strides in a decade. Nineteen eighty-five was the year that semiconductor makers like Intel and NEC first boasted of cramming a million transistors on a single memory chip. (Of course, that pales next to today’s highest-capacity chips, which can hold 128 trillion discrete elements.) But that wasn’t the only technology that had improved so swiftly. Hard disk drives were finally cheap enough for consumers to afford: If you shopped around, you could find one that held 10 megabytes of digital storage for about $700. Back then that was enough space to hold all the essential software and applications you used for quick access and then some. (To give a sense of comparison, today, $700 will buy you 10 terabytes of storage, or roughly 100,000 times as much capacity—room enough to stockpile more than a thousand high-definition movies.)

As a result, the performance and capacity of reasonably priced microcomputers were improving by leaps and bounds, and would continue to do so in the foreseeable future. Steve understood this, and thought it was possible to find a perfect niche for his perfect new machine, right between the personal computer and a new category of desktop systems that had come to be called engineering workstations.

The “workstation” segment of microcomputers emerged in the early 1980s, around the time that Apple was working on the Lisa and IBM was gearing up to introduce its machines. Workstations were basically PCs beefed up with more memory, more data storage, faster processors, and, most visibly, gigantic, twenty-four-inch screens. They emerged mainly from computer science departments in academia and were designed to put as much raw processing power as possible in the hands of a single user—most likely an engineer or scientist whose institution could afford the machine, and who could write his or her own applications to perform heavy-duty calculations or mathematical models. Workstations had two other attributes that really made them stand apart. First, they were designed from the ground up to participate in networks with other workstations. Second, they employed the most advanced microcomputer software operating system of the time, which had been originally developed by computer scientists at AT&T’s Bell Laboratories, and then nurtured and improved by academic researchers and scientists in national laboratories. Called Unix, it was the operating system that enabled the first data “network of networks,” which later came to be called the Internet.

Sun Microsystems, a Silicon Valley workstation maker, got its start in 1982 making such machines for use on the Stanford University Network (hence its name). Sun set a record that still stands in the annals of American business for being the company that from a dead start reached the $1 billion sales mark faster than any other manufacturer—it took all of four years. In fact, Sun was nearing that heady milestone the year that Steve’s new venture opened for business. Sun was a no-nonsense company. Its powerful computers had no special flourishes other than their outstanding performance benchmarks. They delivered great bang for the buck, but their lack of aesthetics offended Steve; instead of seeing the utility of such computers, he saw only opportunity—of course, he assumed, the world would be partial to something easier to use and more attractive.

Meanwhile, personal computers sold by IBM, and by the growing number of “clone” manufacturers like Compaq and others, were perfectly serviceable machines for the thousands of businesses starting to make rudimentary computing a standard part of their office workflow. The workplace was a fast-growing market, with ferocious competitors serving business customers who focused on price, productivity, and return on investment. A startup out to make its mark would have to come up with a computer that stood out from the rest, that gave schools, businesses, or consumers something they really couldn’t find anywhere else.

Given this fierce competition, it’s easy to understand why Sculley and the Apple board sued Steve. With IBM and the makers of other MS-DOS-based clones dominating the market for PCs sold to corporations, Apple needed the school and university market more than ever. Workstations were quickly becoming the lab benches for many disciplines at research universities and in corporate R&D skunkworks. It was only natural that Apple would want to offer its own unique approach to these machines as well. Apple’s suit stalled Steve’s effort to move quickly, by making it difficult for NeXT to do basic things like arrange deals with suppliers, incorporate, hire employees, and so on.

But Apple withdrew its legal challenge in January 1986, in part because Sculley finally decided he didn’t have the stomach for the public relations fallout of a court suit against a popular public figure. In the meantime, Jobs had been able to use the fall of 1985 to study the education market. He, Lewin, and some of the other founders made several trips to universities to hear what professors and researchers really wanted. The founders would remember these treks as fondly as they would the gatherings at the Jackling house. Funded by Steve, who could still be a tightwad, the early employees operated on the cheap, with “that startup hustle,” as Steve put it. “We didn’t have a lot of money,” Bud Tribble told me. “All six of us would squeeze into a single rental car to go make our visits. We even shared hotel rooms. We developed a real pioneer spirit.” For a couple of months, NeXT had the feeling of a true startup. And the road crew actually learned something promising: academics truly did want all the power in those $20,000 workstations. But the crew also learned what the company’s challenge would be: academics absolutely could not spend more than $3,000 per machine. As he had done at Apple, Lewin created a consortium of schools to serve as consultants—and as pilot customers for the NeXT computer. It wasn’t just the allure of signing on with the great Steve Jobs that appealed to the university presidents; it was the fact that the great Steve Jobs had promised that he could indeed deliver the machine they craved for a mere $3,000. It was a promise he wouldn’t come close to keeping.

LATER IN HIS LIFE, Steve would become more adept at managing the press than any other businessman alive. But as he entered his thirties, his idea of good PR was to get attention of any kind. Launching NeXT, Jobs felt that some initial publicity would help attract the investors he needed to build this new and better version of Apple. So he opened the doors for two prestigious media outlets, Esquire magazine and PBS. The results were fascinating: a portrait of a young entrepreneur trying on the clothes of a seasoned businessman, and not quite filling them out.

Steve’s segment of a PBS show called The Entrepreneurs kicked off with an image of him pulling carrots from garden soil. He did occasionally garden, and he may have intended the image as confirmation of his counterculture roots, but the shot lent an almost laughably gauzy warmth to the introduction, setting the stage for a piece that revealed far more about him than he can have intended. The episode consisted mostly of film clips from the company’s first two off-site meetings—intense getaways that were one part group therapy brainstorming, and one part endurance test. Everyone involved in the production wanted to tell the tale of a heroic young entrepreneur, and the voiceover obligingly delivered, describing the show as a chance to see Jobs “at his lucid best, as a company builder and a motivator.” But the language didn’t match the actual footage from the retreats, which made clear just how hard it would be for Jobs to bring NeXT into focus.

The two off-sites were at Pebble Beach, California, the first in December 1985 and the second in March 1986. They were designed for Jobs and his small staff to define their grand project and to assign clear responsibility for the various strands of its development. Footage from the December meeting shows Jobs at the whiteboard trying to get the group to agree on their top priority: Was it more important to meet the $3,000 price target, create a machine loaded with great technology, or deliver that computer by the spring of 1987? As with any startup, different factions advanced their own agendas; Rich Page claimed that the company was pointless if its computer wasn’t a radical technological advance. Dan’l Lewin, the head of sales and marketing, explained that since schools purchased computers during the summer, missing the target date would mean missing a year of revenue. George Crow, another hardware whiz, argued that price was paramount. As always, Jobs was charismatic and confident and clearly aware of when the cameras were rolling. His sentiments were touching, his heart seemed in the right place, and his bold words were inspiring. “More important than building a product, we are in the process of architecting a company that will hopefully be much more incredible, the total will be much more incredible than the sum of its parts,” he said. “The cumulative effort of approximately twenty thousand decisions that we’re all going to make over the next two years are going to define what our company is. And one of the things that made Apple great was that, in the early days, it was built from the heart.” But not surprising for a CEO emphasizing the importance of “twenty thousand decisions,” he made little progress in steering the group toward consensus. His one clear conclusion—“The delivery date is a line in the sand”—registered as what it turned out to be: an unattainable fiat. His team seemed smart, passionate, and intelligent; but it also seemed young, naïve, unfocused, and in desperate need of a leader more decisive than Steve.

As they pontificated, deliberated, and pointed fingers, especially during one fraught March discussion on cost cutting, the group made apparent how absurd it was to think that they could actually deliver a great computer in fifteen months. For years, Jobs had been criticized by Scott, Sculley, Markkula, Woz, and others as being a divisive and impulsive manager who sowed chaos unnecessarily, delivered products late, gave unclear and shifting directions, and advanced his own ideas at the expense of the corporation. The squabbling was a clear harbinger that similar troubles awaited the NeXT crew.

For the Esquire piece, which was published in December 1986, Steve invited the writer Joe Nocera to spend a week at the company. Nocera (now an op-ed columnist at the New York Times) attended planning meetings and strategy sessions at the company’s new offices in the Stanford Research Park in Palo Alto (the same building I visited when I met Steve for the first time), where he spoke with a wide range of staffers. He dined with Jobs and visited him at home—activities that would be strictly off-limits to most journalists later in Jobs’s career. As always, Steve had a point he wanted to make, in this case that NeXT was “going to take the technology to the next level,” as he told Nocera. Getting there would mean re-creating the intensity and passion that he had loved during the development of the Mac. “I remember many late nights coming out of the Mac building, when I would have the most incredibly powerful feelings about my life,” Jobs said. “Just exhilarating feelings about my life. I feel some of that now with NeXT. I can’t explain it. I don’t really understand it. But I’m comfortable with it.”

Steve’s strong feelings about Apple rippled through the story, so much so that Nocera called Jobs’s assertion that he had put Apple behind him “wishful thinking.” “Apple,” Jobs admitted, “is like an intense love affair with a girl you really, really like, and then she decides to drop you and go out with someone who’s not so neat.” The story even dipped into Steve’s relationship with his girlfriend at the time, Tina Redse, describing how Steve wrote her a long note apologizing for working late one night. Nocera found his single-mindedness lonely. Jobs, who at one point in the article failed to remember if he had curtains in his house, refused to acknowledge feeling any kind of wistfulness or dissatisfaction.

“That impression of eternal youth,” Nocera wrote, “is reinforced by some guileless, almost childlike traits: By the way, for instance, he can’t resist showing off his brutal, withering intelligence whenever he’s around someone he doesn’t think measures up. Or by his almost willful lack of tact. Or by his inability to hide his boredom when he is forced to endure something that doesn’t interest him, like a sixth grader who can’t wait for class to end.” Looking back, it’s clear that Nocera had landed on something few people, including Jobs, wanted to see—the fact that the Steve Jobs of 1986 was too raw, too self-centered, and too immature to successfully pull off the balancing act required of a big-time CEO.

About the time Nocera had started his reporting, Steve hired a new PR agency, Allison Thomas Associates. He had gotten to know Thomas when she was steering a state commission on industrial innovation, a project that solidified California’s support for high-tech companies and led to corporations getting a tax credit for donating computers to schools, among other initiatives. Steve wanted to reposition his image, to move past all the stories about his erratic behavior. Thomas, who became close to Steve over the years, found a way to discuss the problem without setting him off: What could be done about “the other Steve,” the one who came off seeming arrogant and mean? It was an ingenious approach that helped the two work closely for several years. But in the end “the other Steve” won out: Steve badgered Thomas incessantly, urging her to cut off all communications with any reporters who criticized him. She would quit in 1993, a few weeks after Jobs paged her three times during the inauguration of President Bill Clinton, which she was attending in Washington, D.C.

AFTER ONE OF NeXT’s early board meetings, Steve pulled aside Susan Barnes, his CFO. “When my life is over,” he told her, “people will give me credit for all the creative stuff. But no one will know I actually know how to run a business.”

As Steve started NeXT, it was true that he did know certain key things about running a computer business. He was a strong, if somewhat confounding motivator, and a restless innovator. He had shown himself to be a good negotiator with parts suppliers, often getting Apple better prices in its early days than its volume had really justified. He could synthesize big ideas, and he could see how different technologies could be combined into something that added up to a whole lot more. “He knew about inventory terms, he understood the mechanics of capital investment, he knew cash flow,” says Barnes. “He did understand this, and starting Apple had taught him things you can try to teach an MBA. But he actually knew them. [They were] survival skills.”

Steve craved recognition for this, and spoke often about how well he was going to manage NeXT, and how much he had learned from the mistakes Apple had made during its years of unfettered growth. “This is really the third time around for me and a number of other people at NeXT,” he told me. “When we were at Apple, we spent half the time fixing things that were breaking, whether it be an employee stock ownership plan, or a parts numbering system, or a way of manufacturing a product. At NeXT we have the benefit of having the experience of growing a company from zero to a couple of billion dollars before, and we could anticipate some of the more sophisticated problems that we didn’t anticipate the first or second time around. It gives us a certain level of confidence which enables us to take more risks. We’re working much smarter. We’re thinking things through more, which results in more getting done with less work.”

It sounded good. But much of it was chutzpah and self-delusion. When he started Apple, he had not presumed that he knew how to run a business—he was willing to rely, at least for a while, on his mentors and bosses. Now, however, he acted as if he knew everything, from payroll and engineering to marketing and manufacturing. He was out to do absolutely every little thing right this time. You could see it in his body language. Whenever someone nattered on about a subject Steve believed he knew well—knew better than anyone else, in his opinion—he would look away, tap his feet, shift restlessly in his seat, and behave like a teenager undergoing physical torment until he could finally break in and say his piece. And of course all this was done in a way that was obvious to everyone else in the meeting.

Steve’s overbearing need to weigh in on everything—to get those “twenty thousand decisions” exactly right—slowed everyone down. This micromanagement was the primary example of the fact that Steve did not know how to prioritize in any kind of holistic way at this stage of his career. Remember how he wanted the group at the first Pebble Beach offsite to decide on NeXT’s top priority: a great machine, on-time delivery, or a price tag under $3,000? It was the wrong question. NeXT absolutely needed to do all three things. But Steve couldn’t keep his company focused on what mattered when he couldn’t focus himself efficiently.

Steve was unable to effectively manage all the cash that he had been able to raise. NeXT was bankrolled by $12 million that Steve put in in two stages, as well as by investments of $660,000 each from Carnegie Mellon and Stanford, and $20 million from H. Ross Perot, the idiosyncratic businessman who offered to back NeXT after seeing the episode of The Entrepreneurs. (“I found myself finishing their sentences,” Perot raved to Newsweek.) The investments gave the young, productless company an exorbitant valuation of $126 million in 1987. (Two years later Canon, the Japanese camera and printer maker, would kick in $100 million more, raising the overall valuation of the company to $600 million.) Steve touted the investments as proof of concept. The Carnegie Mellon and Stanford money showed that the schools were anxiously awaiting his computer. Perot’s endorsement just underscored the size of the potential market, and was evidence that the most innovative businesspeople understood Steve’s greatness, potential, and maturity. Perot swore that he’d keep a close eye on his investment: “This is going to be hell on the oyster,” he said in the Newsweek article covering the deal, equating himself in his folksy way to the sand that irritates the oyster to create the pearl inside. But in truth Perot was hands-off with the man he viewed as a young genius. Years before, he had decided against investing early in Microsoft, missing out on billions of dollars as the company’s stock soared, and this time he was determined to roll the dice on one of those brilliant techies from the West Coast. Steve promised to be a careful steward of the cash. In the Entrepreneurs video, he repeatedly urged his staff to conserve resources, to the point of complaining about the hotel room rates they were getting. Despite having seen him throw money around at Apple, Barnes was initially hopeful that Steve might change his ways. “I thought he’d be better when it was his own money,” she remembers. “Boy, was I wrong.”

Most great Silicon Valley startups start out lean and simple. The advantage they have over established companies is the focus they can bring to a single product or idea. Unencumbered by bureaucracy or a heritage of products to protect, a small group of talented folks is free to attack a concept with speed and smarts. Eagerly working hundred-hour weeks, the employees want little more from the “company” than that it pay the bills and get out of their way. They know that if they execute their idea so successfully that their enterprise grows big, at some point they’ll have to deal with the rigors and strains of a corporation. But generally that’s a worry that’s tackled later. At the beginning, corporate trappings can just get in the way, and distract from the all-consuming job of creating an object of desire.

As he had explained to Nocera, Steve enjoyed the spirit of a startup. But his definition of lean and mean had been changed by his experience at Apple. “Living on the cheap was difficult for him after he’d lived the high life there,” says Barnes. Jobs had enjoyed the benefits of Apple’s resources and size, of its manufacturing prowess and rich marketing budget. Despite what he said about wanting to repeat the experience of the Apple II and the Mac, what Steve really wanted at NeXT was the garage spirit of a startup meshed with the safety, status, and perks of the Fortune 500. It wasn’t a combination he could pull off.

The first sign of his extravagance came early, when Steve paid Paul Rand $100,000 for his beautifully designed NeXT logo. Choosing Rand to design the logo was an indication of Steve’s ambition: Rand’s most famous logo is the one still used by IBM. His prestige was such that Steve agreed to Rand’s stringent terms that all he’d get for his hundred grand was a single draft—take it or leave it. Fortunately, Jobs loved everything about it. He loved the logo, and he loved the way it was presented in a classy booklet explaining in detail how Rand had arrived at his notable design, including the philosophical rationale for the lowercase e and the four vivid colors set on black. The day the members of Team NeXT were given copies of the manifesto, Lewin found himself reflecting on when he had first met Steve in 1977 as a Sony salesman based in an office near Apple’s headquarters on Stevens Creek Boulevard. He recalled the way Steve had lovingly fondled the Sony sales materials, making note of the fine paper stock and professional design: “Steve was a freak about Sony, right? Why did people spend fifteen percent more for a Sony product? Steve would walk into our office, and look at the paper and feel the paper that Sony printed their brochures on. It wasn’t the products, it was the tactile feel, the surface and the presentation that mattered to him.” But NeXT was a startup, not a mature, successful company like Sony with billions in revenues, for whom such a pamphlet would be pocket change.

Extravagant expenditures soon became standard operating procedure at NeXT, especially when it came to the company’s headquarters. The Palo Alto offices featured expensive, custom-designed furnishings, Ansel Adams prints, and a kitchen with granite countertops. And when NeXT moved into bigger offices in Redwood City in 1989, no expense was spared. The lobby featured long, lush leather couches imported from Italy. The crowning touch was a floating staircase designed by world-famous architect I. M. Pei, who designed the glass pyramid entrance to the Louvre that opened that year. The staircase was a ravishing predecessor to the showy stairways that now grace some of Apple’s retail stores.

Steve’s spendthrift ways extended throughout the company. “Our information system,” he told me proudly in 1989, “is designed for a company with $1 billion in annual sales.” (NeXT’s 1989 sales would top out at just a few million dollars, leaving the company at least one hundred times short of that $1 billion mark.) But he justified his spending by explaining that he was creating the infrastructure of a Fortune 500 company from the ground up. Unlike Apple, he told me, “We were able to make the investment up front to do it right the first time. Let’s get the best people we can find, and let’s brainstorm and strategize, but let’s just do it once. And let’s have it be good enough to last for a number of years. It will take a little more startup expense, but it will pay many, many times over in the coming years.”

The centerpiece of Steve’s spending was a state-of-the-art manufacturing facility to churn out NeXT computers—a factory designed to be the envy of the world. The plant, fifteen miles across the bay from Redwood City, in Fremont, was small, but it was a marvel. Steve took me on a tour of the place, just before it went into production in 1989. The factory was nearly empty; Steve explained that the place had been designed to operate with few people. He took great pride in every detail, pointing out the robots and machines that had been repainted in the tones of gray that he had specified. The production area was on a single floor about the size of a large restaurant. With no one around on that quiet day, it seemed like something of a Potemkin factory—an empty shell for show—but Steve claimed it had the capacity to produce up to 600 machines a day, which was the equivalent of, yes, $1 billion worth of hardware in a year.

The place had been laid out by an army of manufacturing system engineers—for a while, there were more PhDs working for NeXT’s manufacturing division than for its software arm. It would be flexible, capable of steadily serving a just-in-time manufacturing scheme. The robots would handle almost everything that required great dexterity, including some of the assembly tasks that Woz and Jobs had performed themselves when making the Apple 1: they placed the chips on the circuit boards, soldered everything into place, and tested and measured to make sure everything was right. A human would step in to do one final check, and would handle the final assembly and pop boards into their appropriate slots inside the magnesium cube.

Steve was right—the place was indeed a paragon. This was at a time when Japanese manufacturers had chased most American companies out of the semiconductor fabrication business and were held up as object lessons for automakers in Detroit. He hoped that his pristine factory would give the world glittery proof that American high-tech manufacturers could still excel. More important, he felt that the seeming perfection of the place and his obsessive focus on its details sent a message to employees: if you aim for perfection in everything you do, you’ll achieve greater results than you could ever imagine.

It was a lovely principle. But it didn’t come close to justifying spending outrageous sums on a state-of-the-art factory to build computers for which there was not yet any demand. Steve could easily have outsourced manufacturing; by the late 1980s the computer industry had grown to include a host of contract manufacturers right there in Silicon Valley that could build a highly demanding product like the NeXT computer. The cost would have been far less. For all its beauty, from the lush landscaping out front to the meticulously crafted wheeled tables on which the computer components rolled through the assembly process, the NeXT factory turned out to be a sinkhole. Forget producing 600 computers a day: the factory never produced more than 600 machines in a single month.

IN THEORY, THERE’S nothing wrong with a state-of-the-art factory, a beautiful office for your employees, or a fancy logo. It’s just that in decision after decision, Steve failed to account for the trade-offs that accompanied his fanciful choices. Steve couldn’t distinguish between the extraneous and the critical. As CEO of a fledgling company, that was his key responsibility. At NeXT, he utterly failed to do this.

Steve decided early on, for instance, that the NeXT computer should have an optical disk drive for storing information, rather than a standard hard drive. The optical disk drive had two great advantages: its disks could hold up to two hundred times as much information as the standard hard drive of the time, and they were removable. Steve heavily promoted the idea that regular folks could essentially carry around their life in data, moving from one computer to another armed with their own personal optical disk. It seemed that he wanted to enable the utopian idea of a mobile population carrying its key information with it. (Today, of course, we can access much more data from our smartphones or tablets, but the data resides in the so-called “cloud.”) However, the optical option had many problems, primarily that its drives retrieved information from the disks very slowly. Steve had chosen a vision—the potential of abundant storage—over the customers’ real need—the convenience of data speedily available. When the NeXT computer did finally hit retail stores in late 1989, competitors like Sun happily cast it as a slowpoke compared with their hard-drive-based machines.

Many of the features of the NeXT computer seemed intended primarily to dazzle. Like a standard-issue PC, the NeXT computer consisted of four devices: a keyboard, a mouse, a cube containing the computer, and the monitor. Its designer was Hartmut Esslinger, the German industrial aesthete who had worked with Steve on the first Mac. Esslinger was another expensive choice, a world-class designer who was just as uncompromising as Steve. He ordered up a true cube, with sharp right angles as opposed to the infinitesimal curves found on the edges of the machines from other manufacturers, including Apple. Those curves on the conventional machines weren’t so much an aesthetic choice; they were a concession to manufacturing realities. Creating a perfect cube with true sharp angles required expensive custom molds, which could only come from a specialty metals shop in Chicago. Esslinger and Jobs also insisted that the case be made from magnesium, which is far more expensive than plastic. Using magnesium was a choice, like Jobs’s selection of cast aluminum for the Apple III’s case eight years earlier, that had a significant downside. Magnesium had certain advantages over plastic, but it was much harder to machine perfectly, leading to more flaws in the manufacturing process.

Designing a computer laden with details like these made building it for $3,000 absolutely impossible. The flourishes just added up too fast. “The business plan,” says Lewin, “called for a cube whose material cost was fifty dollars, without the motherboard. Steve went off on this fantasy of wanting the paint job to be of the same quality as some titanium tone arm he’d seen on a four-thousand-dollar turntable. So he sends three people off to General Motors to learn how to do paint that way—Perot had been on the board there, and GM knew how to paint metal better than anyone in the world. And so we figured out how to do that. But that cube that was supposed to cost fifty dollars all-in? The paint job alone cost fifty dollars. It was really fantasyland.”

Even more damaging were some of Steve’s aesthetic fiats about the inside of the machine. One in particular stands out. In a typical production sequence, engineers are told the specifications a computer must achieve; they design circuitry to meet those demands; and only then do they wrestle with the question of exactly what size and shape the computer’s circuit board must be. Steve reversed the process at NeXT. He told George Crow and his hardware engineers that the circuit board for the NeXT computer would have to be a square that fit exactly into the magnesium cube. A square was an odd configuration for the engineers. Insisting on the exact shape of the board, Steve severely limited the engineers’ ability to create something inexpensive that met the computer’s specifications. He added an unnecessary level of complexity, meaning yet more money spent for more engineers working more hours to accommodate a design that contributed nothing meaningful to the final product.

Again and again, Steve made choices that seemed justifiable in isolation but that damaged the company’s critical mission. Steve did a poor job of evaluating these ideas against one another. He couldn’t accept that it was impossible for him to have everything exactly the way he wanted it.

In part, this was because he believed his own press. He was a genius, according to the media and his investors. Ross Perot frothingly described Jobs as “a 33-year-old with 50 years’ worth of business experience.” Little did he know how wrong he was. President Ronald Reagan’s secretary of commerce, Malcolm Baldrige, called Jobs for advice. Editors of the most important publications in the land kept sending their reporters to the West Coast to find out what Steve was thinking about all kinds of subjects, not just computing and technology. (I once tracked Steve down for such an assignment, and listened to him confidently opine on industrial policy, competition with Russia, the drug war, and General Manuel Noriega of Panama.) The fascination with his new company, so out of proportion for a startup with no product entering a highly competitive industry, confirmed his own sense that he was destined to do great things. That sense of genius and destiny made it harder for Steve to sideline any of his own ideas. He acted as if each detail he advanced could make the difference between creating a breakthrough product and putting out the kind of dreck he thought was offered by other manufacturers. Years later, Perot admitted that he had been snowed. “One of the biggest mistakes I ever made was to give those young people all that money,” he said.

Also, Steve could not resist pursuing anything that would show up Apple. Since Apple had a logo that had become iconic, Steve needed one with the same potential and a great pedigree. Apple had a state-of-the-art factory, so Steve’s tiny company built an outrageously expensive factory that could handle as much volume as Apple needed. His obsession with Apple seemed to ooze out of his pores, despite the silence he’d imposed on his handlers. The first time that John Huey, then the editor of Fortune magazine, went to visit NeXT, Huey was waiting in the lobby when Steve returned from a lunch date with other visitors. Not recognizing Huey, Jobs sat down on another of those expensive lobby couches and spent fifteen minutes flipping through a set of magazines, excoriating Apple’s “stupid” advertising created by whatever “bozos” they had running the show over there now.

Some writers have tried to cast Steve’s obsessiveness, and his hunger for the spotlight and success, as a Freudian attempt to bring down the birth parents who “rejected” him by letting him be adopted. It always struck me, however, that at his childish worst Steve was really nothing more than a spoiled brat. Brilliant, precocious, and meticulous, he had always gotten his way with his parents, and had brayed like an injured donkey when things didn’t turn out as he planned. As a grown-up he could behave exactly the same way, sometimes exploding in a temper tantrum. At NeXT there was no one to keep that side of him in check. While more grounded and cooler-headed folks like Lewin and Barnes would disagree with him and weigh in with advice, he ignored them with impunity and, often, scorn. Talking about the days after the historic introduction of the Mac, Steve had told Joe Nocera, “I think I know what it must be like to watch the birth of your child.” Unfortunately for the team at NeXT, in many ways Steve himself was still the child, rather than the more mature and supportive parent.

STEVE’S ARBITRARY DECISIONS dumbfounded those under him at NeXT, and his micromanagement gave them no peace. He assumed they would work nights and weekends. He wouldn’t hesitate to call them at home on Sundays or holidays if he’d discovered some “urgent” problem. And yet hardware and software engineers still could not resist working for Steve Jobs.

Steve understood their sensibility. Engineers, at heart, are problem solvers. They thrive on digging their way out of sinkholes, especially the gnarly kind with no clear path forward. Steve challenged them in ways they had never imagined. No one else in the computer business had such radical goals and expectations; no one else seemed to care so much about their work. The idea of creating a computer that could transform the very process of education was cool; but to his incredibly talented programmers and gearheads, the idea of creating this particular computer for this particular boss was irresistible.

As the years went on, it became apparent that Steve’s goals for the NeXT computer went way beyond serving the university market. Lewin and his salespeople were courting customers in all kinds of businesses, thinking that the NeXT computer could transform the corporate workplace, blending the computing power to do 3-D modeling and to interpret copious data with the ability to connect with others easily on corporate networks. A machine like that, made available not just to the denizens of the ivory tower but also to the quants of Wall Street and the merchants of Main Street, truly would be revolutionary. So even as the company drifted from month to month, year to year without delivering a final product to the market, many of the engineers continued to do great work and viewed their jobs as both a noble mission and a labor of love. Engineers ruled the roost at NeXT. They had their own special wing at headquarters, equipped with a grand piano and locks that kept out all other employees. And indeed, Steve’s amazing collection of geeks at NeXT produced some genuinely great work.

Richard Crandall, a physics professor from Reed College, became the company’s chief scientist and was given enormous latitude to see just how far computing could expand the scope of high-level teaching in fields such as computational science. His work at NeXT carried over into decades of advanced research on cryptography; he later became the head of Apple’s Advanced Computation Group. Michael Hawley, fresh out of the Massachusetts Institute of Technology, worked with a group of folks to create the world’s first digitized library, which included the complete works of Shakespeare and the Oxford Dictionary of Quotations. And when it finally did appear, the NeXT computer would have easy multitasking, easy ways to attach documents to email, and an intuitive user interface to facilitate the networking it made possible.

Most important, Jobs convinced Avie Tevanian, a young software whiz from Carnegie Mellon University, to come to NeXT rather than join Microsoft. At CMU, Tevanian had worked on Mach, a supercharged version of Unix, the powerful operating system for workstations. At NeXT he became Bud Tribble’s key developer on the computer’s operating system, called NeXTSTEP. For years, Tevanian kept a calculator window open on his computer that tallied daily the total value of the stock options he gave up when he turned down Microsoft. But he loved the work, in part because Jobs recognized his genius and handed him enormous responsibility as soon as he walked in the door.

Steve said many times that the difference between NeXT and the manufacturers of traditional workstations was that he cared more about software than they did. The NeXTSTEP operating system that Tribble and Tevanian developed truly was elegant; in typical Jobs fashion, it put a gorgeous, approachable face on an operating system that previously only engineers had been able to decipher. And Steve recognized that a technique called object-oriented programming (OOP for short) had great potential to help developers slash the amount of time required to create applications. One OOP toolkit that Tevanian’s team created, called WebObjects, eventually became a profitable product for NeXT; after the rise of the Internet, it proved to be a great help to companies looking to quickly build Web-based services.

As much as he depended on Tribble’s and Tevanian’s skills, Steve could not resist managing them ferociously. “Early on,” remembers Tribble’s wife, Susan Barnes, “Bud would complain to me about the fact that Steve kept pushing to see what he was working on in action, on a screen. ‘Steve can yell that the sun shouldn’t rise in the east,’ Bud would tell me, ‘but it’s going to rise in the east, and it’s going to take time to get this software to the point where you can see something visually on the screen. I know he’s a visual learner, I know he can see that way, and I know it’s frustrating for him to look at lines of code. But that’s life!’ ”

“The company was so small,” says Tevanian, who looks like a professional soccer player, with his dark curly hair, deep-set eyes, and athletic frame, “that everybody knew everyone else. I’d be working late at night and Steve would come by and I’d show him what I’m working on, and then he’d yell and scream at me, tell me how terrible it was and all that kind of stuff. But in the end, there was a bunch of stuff that I knew that he didn’t know. He knew that he didn’t know it, so we developed this mutual respect where I could tolerate some of his criticisms because he would also actually listen to me when I had something to say. We made it work.”

EARLY ON AT NeXT, Steve said the most important thing he could do was “architect a great company.” This potentially noble sentiment became a half-baked and confused endeavor, and yet another distraction. Sometimes Steve’s good intentions could lead to a deep intellectual self-deception, in which trivial issues loomed larger than life and fundamental realities were swept under the rug.

He did try to be a good boss. For example, Steve hosted annual “family picnics” for his employees in Menlo Park. They were kid-oriented Saturday affairs, featuring clowns, volleyball, burgers and hot dogs, and even hokey events like sack races. At his invitation, I attended one in 1989 with my daughter, Greta, who was five years old at the time. Steve, who was barefoot, sat with me on a hay bale and chatted for an hour or so while Greta wandered off to watch the Pickle Family Circus, a Bay Area comedic troupe of acrobats and jugglers that Steve had hired. NeXT staffers would come up from time to time, thanking him for throwing the bash. We talked about his business a bit, but mostly Steve rattled on about how important families were to NeXT, and about how many families there were over at Pixar, the small graphics computing outfit he’d acquired from George Lucas. Some of it was hot air, but some of it was a reflection of the fact that Steve really was wrestling with the issue of paternal responsibility. Down deep, he ached for a family of his own. He was spending more and more time with his daughter Lisa, a reconciliation process that was never entirely successful, but that would eventually lead to her living with him during her high school years. I had the feeling that he looked at those picnics as evidence that he could in fact be a good father, if not to his daughter, then at least to his employees. “I think he looked around those gatherings and thought, ‘Oh God, I’m not just carrying all these employees, I’m carrying their families, too,” says Barnes. “It added to the pressure he felt.”

Steve’s budding paternalism carried over into his efforts to develop friendships with some of his closest executives. When Tribble and Barnes had their first child, Steve snuck into the hospital after hours to visit. “Steve so much wanted to be a father figure,” remembers Jon Rubinstein, who joined NeXT in 1991 and eventually replaced Rich Page as the lead hardware engineer. “He’s just a year older than I am. But he had this father-figure thing going that was very funny because, you know, he thought he knew more about life than anyone else around him. He always wanted to know about my personal life.”

But when he tried to intellectualize or institutionalize his paternalistic feelings, he often did so in shallow, poorly designed ways. As part of “architecting a great company” Steve tried to implement an idealistic social experiment he called the “Open Corporation.” Salaries were set by category, so that everyone with a certain job title would be paid the same amount. And every employee’s salary information would be available to everyone else. It was, Steve once claimed to me, an example of his commitment to treating everyone at the company fairly. Then he launched into the “heartfelt” soliloquy he’d prepared for that particular moment:

“It’s people who make our factory work. It’s people who write the software, who design the machines. We’re not going to have to out-scale our competitors, we have to out-think them. Every time we hire somebody, we put a brick into building our future.

“Hiring the right people is only the beginning—you also have to build an open corporation. Think of it this way: If you look at your own body, your cells are specialized, but every single one of them has the master plan for the whole body. We think NeXT will be the best possible company if every single person working here understands the whole basic master plan and can use that as a yardstick to make decisions. Sure, there is some risk with giving everybody access to all the corporate information, and potentially some loss. But what you gain vastly surpasses what you lose.

“The most visible sign of the open corporation at NeXT is our policy of allowing everybody to know what salary everybody else is making. There’s a list in the finance department, and anyone can go look at it. Why? In a typical company, a typical manager might spend three hours a week on compensation issues. Most of those three hours a week is spent defusing false rumors and talking in caged terms about relative compensation. In our company, the manager still spends those three hours, but we spend them defending in a very open way the decisions we made and explaining why we made them, and coaching the people that work for us about what it will take for them to achieve those levels of compensation. So we tend to look at those three hours as an educational opportunity.”

Talking about the Open Corporation gave Steve a way to cast a sheen of moral exceptionalism on NeXT. But his actions soon contradicted his words. By the time he was telling me this, the practice had already been exposed inside the company for the twaddle it was. That’s because Steve was always hell-bent on hiring the very best people in the world, especially engineers. “In most businesses, the difference between average and good is at best 2 to 1,” Steve once told me. “Like, if you go to New York and you get the best cabdriver in the city, you might get there thirty percent faster than with an average taxicab driver. A 2 to 1 gain would be pretty big. In software, it’s at least 25 to 1. The difference between the average programmer and a great one is at least that. We have gone to exceptional lengths to hire the best people in the world. And when you’re in a field where the dynamic range is 25 to 1, boy, does it pay off.”

The hiring process at NeXT was rigorous, with multiple interviews. In many cases, even one interviewer’s “No” could blackball a candidate. And there were candidates aplenty, vying for the chance to work with Steve. But of course, even at NeXT it wasn’t possible to hire the best of the best without strong financial incentives. So Steve started making exceptions for certain hires. Some folks got extraordinary signing bonuses. Others were simply granted higher salaries than their category would mandate. And when these backdoor deals started to make their way onto that list in the finance department, well, all of a sudden that list became a lot harder to find.

Not only was the Open Corporation logistically and managerially unrealistic; it was emotionally out of synch with the reality of a Steve Jobs workplace. He would repeatedly undermine the vision of harmony, peace, and equality he had promised to foster with his irascible temper and anger and his penchant for using passive-aggressive methods to drive his people harder and harder. Steve was as erratic and verbally abusive at NeXT as he was anywhere else during his career. Moreover, he was an equal-opportunity abuser, yelling not only at his engineers but also at his executive team and his own personal administrative assistants on a regular basis.

His inner circle came to understand the pattern of his anger, but that didn’t make it any easier. Tevanian did his best to protect his software engineers from the wrath of Jobs, by making sure they were away from the office when he informed Steve of a slip on schedule, or when a user interface feature he had ordered up turned out to be unworkable. Barnes, who had become familiar with Steve’s unpredictable anger while at Apple, had clear strategies for herself and her employees. “If he’d get mad and start screaming, I’d hang up the phone. He is the only person I knew that you could hang up the phone on, and then pick it up and call him back and he’d be calmer. I mean, if you hung up the phone on me, I would kill you. But with him, if yelling isn’t getting him what he wants, disengage. Leave the room and he will come back nicer, in a different way. I understood that this was something he could turn on and off, and that he would use if it worked.” As for her staffers, she routinely told them to mentally plug their ears and try to “listen through the yelling.” Explains Barnes: “You had to get through the yelling to the reason for the yelling—that was the important part, something you could try to fix.”

THE SENSE OF urgency around the company ratcheted up as Jobs pressed everyone to prepare for the October 22, 1988, debut of the NeXT computer. Steve always relished putting on a show to unveil his digital creations, but he hadn’t performed onstage since pulling the Macintosh out of the bag, like a rabbit out of a hat, back in 1984. Steve believed that these magic-act announcements not only were good salesmanship but also helped galvanize employees and energize a company that was weary after its Sisyphean struggle to ready the product for launch. His performances would grow more and more elaborate over the years, his stagecraft would show increasing sophistication, and the amount of groundwork involved would increase correspondingly, as well as the stress for anyone involved with staging the event. It was exhausting work, and afterward anyone who could do so would immediately head off on vacation.

Introducing the NeXT computer called for more sleight of hand than ever. The operating system, which was at least a year away from being released, was buggy. The optical storage drive ran too sluggishly for a demo. There were no apps written by outside software developers. With the possible exception of the iPhone nearly twenty years later, Steve would never unveil a product that was less ready for prime time. But he couldn’t wait any longer. Steve needed the event to be a success. The halo of being “Steve Jobs’s next great company” was wearing off; even potential like Steve’s comes with an expiration date.

More than three thousand guests packed Davies Symphony Hall, the sleek modern home of the San Francisco Symphony Orchestra. Security was tight, and dozens of self-proclaimed VIPs were bluntly turned away. Inside, an exhibit of photography by folk rocker Graham Nash graced the curvilinear vestibules, hinting at the possibility of the presence of some real celebrities on the program.

Once audience members stepped into the concert hall they could see a giant video screen serving as a backdrop for the darkened stage. A tall table on the left held a large vase bursting with white French tulips and an array of remote controls. On the right side of the stage, shrouded in black velvet, was what appeared to be a phalanx of computer monitors on an elliptical table. Behind the desk chair facing them stood a pillar about four feet tall with another black velvet mantle draped over it. Befitting the venue, chamber music wafted through the sound system as the crowd settled in. It was a Tuesday morning, yet most of those in attendance were dressed as if they had arrived for a night at the symphony. (I even wore a suit.) That’s just the way Steve wanted it.

The show was so extravagant a success that it really could have been considered NeXT’s first major product. The crowd went silent as soon as Steve, clean-shaven and with his hair neatly trimmed, stepped into the spotlight. He was wearing a dapper, dark Italian suit, a blindingly white shirt, and a burgundy-and-black crosshatched tie. Pausing to soak it all in, he smiled with pursed lips, trying hard not to break into a toothy grin; the applause went on and on.

“It’s great to be back,” he said, after the clapping stopped completely. And then, pressing his hands together in a prayerful gesture, he launched into his comeback pitch. It would last two and a half hours. He had spent months polishing his remarks, which were given as more of a business school lecture than a sales spiel. Steve laid out the new taxonomy of the computer industry—a version of that taxonomy that made his new machine seem like the natural next grand step, of course. He did so using presentation slides that had been meticulously put together by hand, because no computer application yet existed to automate the process. Work on the slides had gone on and on; after days of trying to find the exact shade of green for one slide, Steve finally found a tone he liked and kept muttering, “Great green! Great green!” The phrase became a mantra for the beleaguered marketing team.

He explained how what he was now calling a “personal workstation” fit the needs of sophisticated computer users much better than the workstations that Sun and Apollo were selling for tens of thousands of dollars. While at Apple, he admitted, he had overlooked the significance of linking personal computers such as the Macintosh into networks. The NeXT computer was designed from the ground up to be connected to a network.

Computer scientists already knew this history, of course, but the broader public that was so fascinated with Jobs didn’t. Steve had always been able to describe the potential of obscure yet real technologies with such aplomb that he created something akin to lust in his audience. He had absolute self-confidence that he could sell people a sense of discovery in the form of technological products they previously didn’t even know they wanted, a confidence that was usually justified. When he held up the NeXT computer’s innards and described it as “the most beautiful printed circuit board I’ve ever seen in my life,” the audience gasped and then broke out into applause, despite the fact that at any distance over a few feet every circuit board looks pretty much the same. The audience even clapped when he described the Cube’s ten-foot power cord. On this day, the crowd would follow wherever Steve would lead. When he called big universities “Fortune 500 companies disguised by another name,” they even seemed to believe that this was true.

The tricky part of the show came when he had to explain that this radically new computer would have to make do with black-and-white and grayscale graphics, a cost-cutting decision (it saved NeXT $750 per machine) that had become unavoidable as Steve’s persnickety meddling had delayed the machine and driven up its cost. No matter. Steve simply presented the screen as a magnificent design element. He bragged about the subtle shades of gray in a way that almost demeaned a color screen as an unnecessary extravagance. As the demo went on, Steve’s claims became more grandiose, as if these NeXT machines might revolutionize the academics of not just science but the arts as well. Given all this potential, he suggested, it was remarkable that the NeXT computer would cost only $6,500; that its printer was priced at a mere $2,000; and that customers who wanted a conventional hard drive to augment the machine’s storage capacity would pay just $2,000 more. Still, he couldn’t completely conceal the reality that a fully functioning NeXT computer system would cost well over ten grand—some seven thousand dollars more than it was supposed to have.

Steve knew he had to end the show with something that would obscure this unfortunate detail, something that would bring the concert hall crowd to its feet. And that something was music. For the previous six weeks, he had pushed Tevanian, who had been with the company just a few months, to build a music synthesizer software application that could show off the Cube as a more multitalented computer than anything else around. Developing that music synthesizer capability was a tricky bit of programming, and one night, after weeks of effort, Tevanian finally figured out how to make it work. “Suddenly, this circuit board was producing sound! I thought, This is amazing!” Tevanian remembers. “It’s eleven o’clock at night, though, and there is nobody there to show it to. So I run to the building next door, and lo and behold, Steve is still working. I said, ‘Steve, I’ve got to show you something.’ So we run back over to the engineering lab and I show him. And he just starts swearing at me. ‘Why did you show this to me? I can’t believe you did this!’ he yells. And I say, ‘Steve, you don’t understand, it works!’ And he says, ‘I don’t care, because it sounds horrible. I don’t ever want to see anything like this again.’

“I learned a lot from that interaction,” Tevanian adds. “Most people who work in a Steve Jobs organization end up quitting or being fired when that happens, but I just put my head down and thought, Okay, so there is a bar that you have to exceed before you can show it to him. I can show it to other people, but not to him.”

At the very moment when the audience might have been expected to grow restless, Steve unveiled Tevanian’s trick. He had the computer play gamelan music from Indonesia in a percussive demonstration of polyphonic sound generation. The audience sat rapt as amplified, synthesized music swelled to fill the hall. Steve, like a parent trying not to seem overly proud, broke into a taut smile. Nobody had heard anything quite like this come out of a computer. But that was just the warm-up act. Steve next invited Dan Kobialka, the concertmaster for the San Francisco Symphony, to perform a duet alongside the Cube on his violin. The selection was an excerpt from Bach’s Violin Concerto in A Minor. Steve backed away and spotlights illuminated the two performers; for more than five minutes, Davies Symphony Hall seemed as intimate as a living room. When Kobialka lifted his bow at the end, the standing ovation was spontaneous. A third spotlight trained on Steve, who was holding a single rose as he bowed to the adoring throng.

Reactions to the premiere were as over-the-top as the event itself. Industry mavens like Stewart Alsop, Dick Shaffer, and Michael Murphy forecast that the machine would make NeXT a $200- to $300-million-a-year company by 1990. Shaffer called himself a “convert.” Even leaving the pundits aside—after all, they were in the business of grandiose predictions—the more sober press accepted Steve’s claims for his new machine. I called it “dazzling” myself in my front-page article for the Wall Street Journal, and even described it as “relatively inexpensive.” As Steve had intended, I was comparing the machine to existing workstations, and not to the original price point he’d promised the universities that were supposed to be his key customers.

The truth that all of us missed was that this was a machine that had virtually no chance to succeed in the marketplace. Steve’s mismanagement meant that the NeXT computer was only somewhat less expensive than most workstations, with just a few marginal improvements that didn’t offset its many shortcomings. The principles on which NeXT had been based were in tatters, the goals of those long-gone offsite meetings trashed. Jobs had been told emphatically from the start that the machine should cost no more than $3,000; more recently, his collegiate advisers had told him that it should probably be priced at half that. Colleges were not about to spend $10,000 on a fully tricked-out NeXT computer system, versus $2,500 for a Mac or $5,000 for a low-end Sun workstation. The game was already over, but few of us knew it.

STEVE’S MOST IMPORTANT and direct competitors were not fooled by the glitzy debut. The folks at Sun Microsystems laughed off the introduction. CEO Scott McNealy, a brash Detroiter who played hockey in his spare time, thought that Steve’s fancy fonts and magnesium case were wasted on the hard-core buyers of workstations. “We give them what they want,” he told me, “and they don’t really care how pretty the icons are.”

If Steve had started NeXT with a clear mind and even an ounce of humility, Sun is the company he would have acknowledged as his most dangerous competitor—and potentially his best role model. McNealy, one of four cofounders who started the company in 1982, had become CEO in 1984. He was only three months older than Steve but seemed far more seasoned. His father had once been the CEO of American Motors, the now-defunct automaker that is remembered primarily for offbeat car models like the Nash Rambler and the AMC Pacer. At night, as a child, Scott would pore through his father’s briefcase when he wasn’t looking. As an adolescent, and later as a college student, he became what he called a “factory rat,” spending time on the plant floor learning firsthand about the complexities of auto manufacturing, and about the dynamics of corporations that manage large numbers of people. A child of privilege, he went to Detroit’s most prestigious private schools, and then on to Harvard for a business degree, finishing up with a Stanford MBA.

McNealy couldn’t have been more different from Steve Jobs. Aside from his formal education, he was a rabid jock. He liked country music and heavy metal, not Dylan and the Beatles. McNealy was an irrepressible practical joker who was prone to shoot from the lip, but he managed his company with a maturity that Jobs only pretended to have. Sun had hit $1 billion in sales within four years. McNealy did it by smartly targeting a customer base that had money to spend—corporate R&D departments, the U.S. military, and the National Laboratories, a less glamorous but much more affluent set of customers than the universities Steve went after. Sun next went after Wall Street, which was just beginning to discover the power of using computers to identify quick trading opportunities. These customers didn’t much care what the computers looked like, as long as they had big screens and could handle multiple computing threads simultaneously.

Sun succeeded by identifying the market’s real need, by delivering just that product, and by keeping its machines reasonably affordable. NeXT failed at all of that. In fact, NeXT didn’t actually sell its first computer until almost a year after that splashy debut in Davies Hall—four full years after Steve had started the company. McNealy was focused, budget conscious, and opportunistic. Steve’s goals were muddled, and he was a spendthrift who was slow off the mark. McNealy had thrown in with his cofounders at Sun to sell a lot of machines, serve his customers, and make a lot of money. Steve had founded NeXT because he was furious at John Sculley and Apple, because he desperately needed a second act, and because he thought it was his responsibility—and birthright—to keep astounding the world. There had been a market to be attacked when Steve founded NeXT; McNealy’s success proved it. But Steve was still young and immature, and didn’t think there was anyone else in the computer business who really mattered. He was looking in the mirror while McNealy had been looking out the window to learn what the world really needed.

Another industry leader was equally unimpressed by NeXT. Bill Gates refused to develop software for the NeXT computer, despite Steve’s repeated, if confused, efforts to lure him in with promises that Microsoft would profit as much with NeXT as it had with the Mac (for years, Microsoft had been the Mac’s leading applications developer). When Bill first visited Steve in Palo Alto to see what Jobs was putting together at NeXT, Steve left him stewing in the lobby for half an hour before coming to get him. It was a spiteful beginning to what would turn out to be a nonexistent relationship between Microsoft and NeXT. Gates rebuffed Steve again and again and again, with venom. “Develop for it?” he told InfoWorld. “I’ll piss on it.” Microsoft software was already on its way to defining the industry standard in nearly all aspects of computing, so Gates’s reluctance to support NeXT with custom versions of its application software effectively marginalized the company.

Gates didn’t let up after the Davies Hall showcase. “In the grand scope of things,” he said, “most of these features are truly trivial.” A year later, he said of the NeXT computer, “If you want black, I’ll get you a can of paint.” To this day Gates remembers the moment he definitively told Jobs that Microsoft absolutely would not develop software for NeXT. “He wasn’t livid,” Gates told me recently. “He was deflated. He was at a loss for words, which wasn’t typical. He knew what I was saying might be right. And it wasn’t a particularly pretty picture in terms of what it meant for big black cubes changing the world.”

TOWARD THE END of the show at Davies Hall, Steve Jobs revealed what should have been the biggest news to come out of the event: computer industry colossus IBM had decided to license the NeXTSTEP operating system for use on a line of its own engineering workstations. The vision of the world’s biggest computer company running Steve’s revolutionary operating system seemed a great endorsement.

The basic agreement between the companies was that IBM would license the right to use the NeXTSTEP operating system as a graphical interface in return for $60 million—a pittance for IBM, but critical operating capital for NeXT, which was burning through its investors’ cash as the years dragged on. Many people believed that the arrangement might have broader implications, an impression that Steve did nothing to dispel. IBM had an existing deal with Microsoft to jointly develop a new operating system for future PCs called OS/2. By announcing the deal with NeXT, IBM seemed to be indicating that it was not comfortable with Microsoft as its sole key partner. (Indeed, within a few months of the NeXT announcement, it became clear that IBM and Microsoft were having serious issues working together.) The tantalizing possibility was that Steve’s operating system might eventually power not only workstations but also the personal computers of Apple’s most feared competitor. If that ever happened, Steve’s comeback would be complete.

But Steve never seemed to quite know how to play his cards with IBM. He displayed an unsettled and juvenile mix of hubris and uncertainty. Jobs could be bold and strong, as when he secured IBM’s promised investment before any Big Blue exec had even laid eyes on the NeXTSTEP operating system. But he could also be just plain rude: Walking in late to one meeting with a suite of IBMers who had flown out to NeXT, Steve interrupted the proceedings with the dismissive pronouncement: “Your UI [user interface] sucks.” Dan’l Lewin, who would meet with Steve to carefully plot their strategy before every IBM meeting, never knew what to expect from his boss. Sometimes Steve would completely undermine the groundwork the two had carefully laid. “I’d sit there and literally kick him under the table,” Lewin recalls. “There was one meeting, for example, where he went in and actually told them, ‘I really don’t understand why you guys would want to help us.’ ”

Psychologically and emotionally, signing on with IBM was every bit as complicated for Steve as begging Bill Gates to support his computer. Steve had always envisioned NeXTSTEP as the backbone of his own spectacular computer. He wanted to be the hero, not a secondary partner to a more powerful computer company. If IBM had exploited his operating system, and sold a lot of computers running their version of NeXTSTEP, the glory would have been theirs, not his.

It shouldn’t have been surprising, then, that Steve failed to make this important relationship work. He killed the IBM deal by failing to follow through as a good business partner. IBM’s Bill Lowe, a veteran who had been instrumental in launching the PC back in 1981, had initiated the deal. But Lowe retired in 1990, and James Cannavino replaced him. Cannavino logically assumed IBM could use NeXT’s 2.0 version of NeXTSTEP on its machines. But Steve, who hadn’t even met Cannavino, held up IBM for more money, leading to another round of protracted negotiations. He overplayed his hand. Cannavino stopped taking Steve’s calls and just abandoned the project, although there was never any real announcement that it was over. It was a minor disappointment for IBM, ending its “Plan B” fantasy of creating a real alternative to Microsoft’s new Windows graphical operating system for PCs. But it was a fatal blow to NeXT, ending its last real chance to achieve the kind of scale that would have turned it into, as Steve had said in 1985, “the world’s next great computer company.”

Lewin quit NeXT in frustration months before the IBM deal dissolved. He was the first of the five original employees to leave NeXT. “We owned the world when we were at NeXT. And it failed because of Steve,” Lewin remembers. Steve invited Lewin to lunch two weeks after his resignation. “Well, now that you’re leaving, what do you really think?” Steve asked him.

“You’re going to go through every penny, the way you’re headed,” said the now former head of sales. At the time NeXT still had some $120 million in cash. “This company is just not going to happen. You may own 51 percent of it, or 58, whatever it is, but more than half the company worked for me. And I’ve been fighting with you because I believe in what I know we need to do to run this business. If you want to succeed, you need to listen to your people. Otherwise you’re doomed.”

A few months later, George Crow, another cofounder, left, tired of bearing the brunt of Steve’s furor for the delays in hardware development. Susan Barnes left around the same time in 1991. Managing Steve’s finances when he had little fiscal discipline and no checks on his spendthrift ways had grown wearying. “It was classic: his visionary optimism versus my reality,” says Barnes. “He always felt that we were going to turn the next corner. And I would always tell him that there was nothing in the business model to indicate that that was so.”

When Barnes resigned, Steve immediately and without warning cut off her phone and email access. A year later, Rich Page quit the company, and Barnes’s husband, Bud Tribble, picked up the phone to ask Scott McNealy at Sun if he could use a highly seasoned software engineer. A few days later Tribble went to work for the company that was everything NeXT should have been. Just six years after those heady brainstorming days at the old Woodside house, the renegades had all departed, leaving their rock star behind.