It was a fascinating admission. Steve was never much for looking back at his own mistakes, and yet during this very public conversation with a friend whom everyone but Jobs now acknowledged as the leader of the computer industry, he was downright contrite. Later in the conversation, he even pulled out a story he’d ripped from the pages of Newsweek to make sure that Bill wasn’t offended by the author’s claim that Steve was no longer his friend. “I tore this out and I was going to call you before I knew we were getting together,” he said, brandishing the page like a trial attorney. “This is not true at all, and I have no idea where they got that.”
Steve became most engaged when we started discussing whether the PC industry would ever again produce a breakthrough machine like the Mac. That was the kind of product that most interested Steve, of course. At every stage of his life, he always wanted to create devices that would completely reset the industry. “Fundamentally,” he explained, “the PC industry is taking the existing and repackaging it or making it run faster. I think that’s much more valuable than I used to. But I also think that what’s the real trick, and the real necessity to keep our industry healthy, is to balance that incremental improvement with some big steps. I worry about the big steps, and where they’re going to come from.” Later he added, “The standard bearer needs a kick in the ass every once in a while. [Besides,] it’s great for the creator of the deviant innovation. If they’re right, there’s a big pot of gold there, and the ability to make a contribution to the world.”
Bill wasn’t obsessed with the revolutionary. He knew that there was a place for breakthrough technologies, and that the nature of the tech business—indeed, human nature itself—guaranteed that such milestones would arise. But over the course of the interview he made clear that what was closer to his heart was the pain that such disruptions caused the corporate customers of his software. “All I want is a car that will run on the current streets,” he explained. “I’m on this evolutionary path.” The huge investments corporate America had started to make in personal computers and in the critical applications it used to run operations “make for some very unusual dynamics,” he said. “In an Egghead Software store five years from now you’re not going to find business software for six different types of desktop computers. Personally, I would be stunned if you would find software for more than one overwhelmingly successful type of computer, and maybe a couple of others. More than three would be shocking.”
When Steve had left Apple in 1985, the primary competition in the computer hardware business had been framed as a battle to design the best machine; whoever did that, it was assumed, would win the most customers. But six years later that wasn’t the game at all, a fact that Steve was only slowly coming to understand, in light of his difficulties with the NeXT computer. The game was now all about serving corporate customers with their millions of machines. Those companies were increasingly reliant on their PCs, which ran custom-built applications that helped them execute complicated, data-intensive operations. They needed these applications to work with every new unit. The cost of re-creating their data to fit, say, a NeXT computer that didn’t work with the Windows operating system would have been exorbitant, not just in the financial cost of reprogramming but in the opportunity cost lost to all the time required by a retrofit. It wasn’t the bells and whistles that excited these customers; in fact, they found bells and whistles kind of scary. Nope, what they needed was more power, more speed, and above all else, reliability.
Very few people writing about this new industry in the mainstream press truly understood how personal computers had already begun to revert to institutional machines. This was mainly because it was easier for most journalists of the early 1990s to envision and get personally excited about the potential of educational software, or of managing their personal finances, or organizing their recipes in the “digital” kitchen, or imagining how amateur architects could design funky homes right on their home computers. Who wouldn’t be excited about more power in the hands of people, the computer as an extension of the brain, a “bicycle for the mind,” as Steve put it? This was the story of computing that got all the ink, and it was a story no one unfurled as well as Steve.
Bill Gates wasn’t swayed by that romance. He saw it as a naïve fantasy that missed the point of the much more sophisticated things PCs could do for people in the enterprise. A consumer market can be an enormously profitable one—put simply, there are so many more people than businesses that if you sell them the right product you can mint money. But the personal computers of that time still didn’t have enough power at a low enough price to excite the vast majority of consumers, or to change their lives in any meaningful way. The business market, however, was a different beast. The potential volume of sales represented by all those corporate desktops, in all those thousands of companies big and small, became the target of Bill Gates’s strategic brilliance and focus. Those companies paid good prices for the reliability and consistency that Windows PCs could deliver. They welcomed incremental improvement, and Bill knew how to give it to them. Steve paid lip service to it, but his heart wasn’t in it. He thrilled only to the concept of how a dramatically better computer could unlock even more potential for its user.
This fundamental difference between the two coparents of the PC was made utterly clear by the interview. What wasn’t made clear, and what Bill didn’t even come close to revealing, was how his deep understanding of the computing needs of businesses would transform the computer business itself over the next several years, further sidelining anyone who, like Steve, chose to focus on the aesthetics and thrills of personal computers. Even though nobody recognized it at the time, Bill was about to take the personal right out of personal computing. Ironically, in so doing, he would leave an opening for Steve to fill—eventually.
THE DECADE OF the 1990s was about to become the Age of Microsoft, a period of time when a single company dictated the direction of the entire computing industry. Microsoft did have a key partner in Intel, whose chips powered almost every machine running the Windows operating system. But the combination of Windows and a growing suite of office productivity applications gave Microsoft an entree to corporations that Intel could never match. While the steadily increasing power and speed of Intel’s chips set a rhythm of inexorable advancements for technology, Windows and Microsoft’s other software shaped the look and feel of corporate computing. By attending to every need of both the Fortune 500 and small businesses, Bill Gates was becoming technology’s king. Intel CEO Andy Grove was, somewhat to his dismay, relegated to the avuncular role of the “elder statesman.”
Together, Gates and Grove had exploited something that Steve had ignored. Looking over the horizon, they could see that the architecture of PCs would improve so much in performance as to subsume almost every aspect of computing. In the past, high-end business machines were based on proprietary designs that didn’t benefit from the economies of scale of standardized parts. Gates and Grove knew that eventually—and it wasn’t going to take very long at all—the expensive, customized guts of engineering workstations would become juiced-up PC circuit boards, and that the same evolution would ultimately subsume business minicomputers, mainframes, and even supercomputers, those rare and superexpensive machines used for everything from modeling weather patterns to controlling nuclear devices. (For example, IBM’s Watson, the machine that in 2011 beat Jeopardy! phenomenon Ken Jennings, is one such computer based on a PC-like architecture.) As a result, pretty much every computer that companies relied on to manage their most critical operations would adopt the internal electronic architecture of a PC writ large. All were much, much cheaper and easier to program and operate than unwieldy mainframes, because they were built out of the very same semiconductor components as PCs, and usually used a variation of the Windows operating system software. Thus they benefited from the ever-improving economics of scale afforded by the combination of Moore’s law and by the breathtaking growth of the PC market itself.