But the iPod had done more than simply create a huge secondary stream of revenue. It had solidified Apple’s foundation and expanded its potential. Tim Cook now managed an intricate supply chain that fed a global manufacturing network capable of churning out tens of millions of iPods a month. Jony Ive had responded to this higher metabolism and greater manufacturing scale by experimenting with new metals, alloys, durable plastics, and super-hard glass that could be sculpted into devices as small as an iPod Mini and as big as a 32-inch computer screen. The executive team was starting to feel that the company would succeed with whatever it took on. “One of the things I’ve always felt,” Steve told me, “is that if you’re going to be creative, it’s like jumping up in the air; you want to make damn sure the ground is going to be there when you get back.” The ground under Steve had never been this solid. The time was right to jump into something radically new that completely changed the game. Steve just didn’t know which direction to leap. Resolving the dilemma would turn out to be the biggest decision of his professional life.

APPLE DID NOT have a formal research and development unit per se. Steve didn’t like the idea of relegating all forward-looking tinkering to a separate area that somehow wasn’t beholden to the people leading his most important product development efforts. Instead, research projects flowered in pockets all around the company, many of them without Steve’s blessing or even awareness. They’d come to Steve’s attention only if one of his key managers decided that the project or technology showed real potential. In that case, Steve would check it out, and the information he’d glean would go into the learning machine that was his brain. Sometimes that’s where it would sit, and nothing would happen. Sometimes, on the other hand, he’d concoct a way to combine it with something else he’d seen, or perhaps to twist it in a way to benefit an entirely different project altogether. This was one of his great talents, the ability to synthesize separate developments and technologies into something previously unimaginable. It’s a talent that he would call on to decide what came next.

Two projects had been launched with the intention of exploring the possibilities for creating a new kind of cellphone. Steve himself had asked the folks who developed Apple’s Airport Wi-Fi networking product line to do some early research on cellular phone technology. This decision made some on his team just shake their heads—Wi-Fi data-networking technology has very little to do with the cellular radio technology behind wireless phone networks. But there was another, much more immediate project in the works. Beginning in the fall of 2003, several members of Steve’s executive team, including Eddy Cue, the mastermind behind the iTunes Music Store, had been engrossed in finding a way to build iTunes-compatible music players and iTunes Music Store accessibility right into cellphone handsets.

“Everybody carried two devices. A cellphone and an iPod,” Cue recalls, patting both front pockets of his jeans. “We knew you could add iTunes to a phone and it would be almost like an iPod. It was mostly a software problem. We looked around at the industry, and in early 2004 we settled on working with Motorola, which at the time completely dominated the handset business with its RAZR flip phone. Everybody had one.” Motorola had been a key supplier to Apple for decades. Its microprocessors powered all of Apple’s computers up until the mid-1990s, and after that it was part of a consortium with IBM that designed the PowerPC chips that would be CPUs in Macs up until 2006. Motorola promised Apple that it would create a new line of phones, called the ROKR, expressly as a vehicle for iTunes.

The ROKR project was controversial from the start, for one simple reason: most people at Apple didn’t like the idea of collaborating with other companies. The iPod hardware team, especially, led by Tony Fadell, couldn’t stomach the notion of ceding the development of what they had started to call “musicphones” to the traditional handset industry. And the more Motorola showed them of its plans for the ROKR, the more certain they became that licensing their precious iPod and iTunes software had been a mistake. While Motorola had certainly built sleek and beautiful phones in the past, the company seemed hopeless when it came to designing software that could replicate the simplicity of Apple’s iPods. To Apple’s whiz kids, Motorola’s approach seemed all but inept. The Illinois company assigned separate teams of programmers to build different software components, like a directory of contacts, text messaging, and a crude Internet browser that could only display stripped-down mobile versions of websites. Nothing about these features was as intuitive as the iPod screen interface, and trying to combine the efforts of disparate, disjointed teams led to a hopeless muddle. Fadell became so exasperated with Motorola that he decided to develop his own prototypes for an Apple cellphone, the first featuring music and the second focusing on video and photos.

Ironically, two other projects that started out having nothing to do with cellphones would come to have the greatest impact on Steve’s decision about what Apple would pursue next. One of these was called Project Purple. It was a skunkworks effort Steve had ordered up to devise a new approach to what was proving to be an elusive “form factor” for personal computing: an ultralight, portable device that resembled a tablet or a clipboard, with an interactive touch screen. The concept had thwarted Microsoft’s best researchers and engineers for years, but Steve believed that his guys could make headway where others had failed. There simply had to be a more direct and intuitive way for users to interact with a computer than a keyboard and a mouse. Preferably it would be something he could use anywhere, even when sitting on the toilet.

The other effort was something that developed far from Steve’s purview. In 2002, Apple researchers Greg Christie and Bas Ording started looking into a user-interface technology that had been stuck in the mud for years. Christie and Ording decided to reconsider the possibilities of a touch-screen monitor, which allowed people to use a fingertip to activate an icon or button displayed on a video screen. Initially developed by IBM in the 1960s, touch screens had not followed a path anyone would call revolutionary. In 1972, Control Data sold a touch-screen mainframe terminal, called the Plato IV. In 1977, CERN, the European high-energy physics research consortium, built one to control particle accelerators. In the 1980s, Hewlett-Packard became the first big manufacturer to offer a touch-screen monitor as an accessory for some of its early desktop PCs—but most software available at the time couldn’t use it. Rudimentary touch screens would become the interface of choice for ATMs, airline check-in kiosks, and cash registers, but they didn’t seem to hold out much promise for personal computing.

In the early 1990s, a handful of startup entrepreneurs, along with researchers in the R&D labs of several computer makers, hit upon the idea that they might be able to reconfigure touch-screen technology into something they dubbed “pen computing.” Their idea was that users would mimic the actions of a mouse by working directly on the screen of a portable computer with a special stylus. They believed that drawing or writing directly on a screen was so natural and familiar that it would be the best way for people to interact with their computers. This was the nascent technology that John Sculley had counted on to make the Apple Newton MessagePad the next big wave in personal computing when it was introduced in 1993. The Newton failed, of course, partly because its handwriting recognition was embarrassingly inaccurate. Microsoft tried for two decades to make something of pen computing in tablet versions of the PC, but to no avail. The only somewhat successful stab at the genre was Palm’s Pilot personal digital assistant (PDA). But the small device was never intended to be a full-featured computer, and its success was fleeting.