The Project Purple team was running into a different set of problems. In their desire to repurpose and yet maintain compatibility with traditional Macintosh hardware and software, Project Purple’s engineers were running into the bugaboos that Microsoft and the others had encountered with their tablet PCs: bulk, weight, battery life, and cost. Even a relatively small ten-inch screen would guzzle power and quickly exhaust the tablet’s rechargeable batteries when operating untethered. Wi-Fi technology, which was the best means for connecting a mobile computer to the Internet or to other computer networks, also sucked power, as did traditional PC microprocessors—even those tailored for laptops. The power demands of a tablet seemed like an intractable problem, given that existing batteries were big and heavy.

So, while the discrete technologies to build an iPad derived from Mac technology were coming together, the actual device you could make would be heavy and impractical, and would carry just about the same price tag as a conventional MacBook. Steve knew that would be a hard sell. Still, he didn’t shutter the operation. Until he had a plan B, he wouldn’t pull the plug on Purple.

Greg Christie and Bas Ording, meanwhile, had spent several months in 2004 putting together and playing with a rather funky, but working, prototype of a multi-touch screen. The pair projected the live video image of a computer screen on a touch-sensitive surface the size of a conference room table. Using two hands, you could “move” folders around, activate icons, shrink and enlarge documents, and “scroll” around the screen horizontally and vertically with somewhat intuitive dexterity. The multi-touch gestures they had contrived to do all of this were rudimentary at this point, but “Jumbotron,” as design chief Jony Ive eventually dubbed their prototype, was intriguing enough to offer a sense of how engaging it would be to control a touch-screen computer with your fingers. Ive, who had become a self-appointed scout for game-changing user interface technologies in Apple’s own labs, had been following Christie and Ording’s work all along and was mesmerized when he saw the Jumbotron demo in action. He wanted Steve to see this. He believed Apple could make multi-touch the basis of a new kind of device, and he believed it should be a tablet computer.

STEVE, TOO, HAD been thinking that Apple’s next step would probably involve some kind of fundamental reconfiguration of the traditional personal computer. He had always been leaning toward making a tablet. That’s why he gave the green light to Project Purple in the first place. But shortly after he returned from surgery, during one of their regular brainstorming walks around the Apple campus, Steve told Jony Ive that he was beginning to think differently. “Steve wanted to shelve the project,” Ive recalls. “I was so surprised because I was so excited about it. But one of the observations he made—and this is classically brilliant Steve—was that, ‘I don’t know that I can convince people that a tablet is a product category that has real value. But I know that I can convince people they need a better phone.’ ” This suggestion wasn’t made in glorious ignorance of the engineering it would require. He knew absolutely that building a phone was much, much harder than doing a tablet, because it had to be so small, and because it had to be a good phone and a good computer and a good music player. What he really wanted was to try to sell a whole new category of device. That, to him, was worth the risk.

When Steve finally checked out the Jumbotron multi-touch demonstration prototype by Greg Christie and Bas Ording, “he was completely underwhelmed,” says Ive. “He didn’t see that there was any value to the idea. And I felt really stupid because I had perceived it to be a very big thing. I said, ‘Well, for example, imagine the back of a digital camera. Why would it have a small screen and all of these buttons? Why couldn’t it be all display?’ That was the first application that I could think of on the spot, which is a great example of just how early this was. Still he was very, very dismissive. It was another example of one of those times when what he says and the way he says it is not personal. You could take it that way, but it wasn’t.”

After mulling over multi-touch for a few days, however, Steve changed his mind. Perhaps multi-touch really was the user-interface leap he had been looking for. He started to pick the brains of people he respected. He called Jony to talk about it further. He conferred with Steve Sakoman, another former Newton and Palm engineer who now worked for Avie Tevanian as the VP of software technology, and who had been pushing for Apple to make the move into phones. And he wanted to hear what the iPod guys thought about multi-touch, since they’d already built the two musicphone prototypes. He asked Tony Fadell to come check out the Jumbotron, since he had the hardware engineering expertise to judge what it might take to build such a technology into a much smaller device that could be mass-produced. Once he saw it, Fadell agreed that the technology was really interesting, but allowed that it wouldn’t be easy to shrink that demo the size of a Ping-Pong table down to something functional that could fit into a pocket-sized device. So Steve gave him exactly that challenge. “You’ve figured out how to blend music and a phone,” he told Fadell. “Now go figure out how to add this multi-touch interface to the screen of a phone. A really cool, really small, really thin phone.”

In hindsight, it’s clear that seeing Christie and Ording’s multitouch demo was an epiphany for Steve, one that was not all that different from his first visit to Xerox PARC twenty-five years earlier. Helping people interact more directly and intuitively with intelligent devices was the central factor in creating a new genre of smart mobile gadgets. The Mac had been a radical new conception of the user interface for a computer, and the iPod’s thumb-wheel had been a user interface breakthrough as well. Multi-touch had the same potential as the Mac’s GUI. But he’d have to move quickly.

Thanks to the iPod, Steve knew that his team could strike fast. And also because of the iPod experience, he knew that Apple could make mind-boggling quantities of whatever device it created. So he decided that Apple was going to create a cellphone. The company was going to put in the palm of your hand a gadget as slick and compact as an iPod, that could download or play music and even video streamed directly over a wireless network, that would be a great phone with amazing voice-mail and directory features, and that would be a computer as powerful as the engineering workstations he’d built at NeXT. Most people hated their cellphones, he liked to say. Apple would create one they would love.

All this decision making took place in late January 2005. It was hardly the only big thing going on Apple—after all, at MacWorld Steve had unveiled the Mac Mini computer, the iPod Shuffle, and a new suite of personal productivity applications called iWork, which he hoped would compete directly with Microsoft Office. But the cellphone project quickly became the main topic of discussion when he and Jony met, as they did almost every day now. They would have lunch together three or four times a week, and take long walks afterward kicking around ideas for solving such mundane-sounding problems as how to keep a touch screen from reacting to contact with your ear when you are talking on the phone, or which materials to use so that your screen wouldn’t get all scuffed when sitting in your pocket alongside keys and loose change. Steve would sometimes go back to Jony’s design lab and sit there for hours, watching designers tinker with prototypes, or else the two of them would stand together at the whiteboard, drawing and modifying each other’s design ideas. They were two kindred spirits, and Steve would now collaborate more closely with Jony than he ever had with Woz or Avie or Ruby or even Ed Catmull and John Lasseter.