Читать книгу Battle of the Titans: How the Fight to the Death Between Apple and Google is Transforming our Lives - Fred Vogelstein - Страница 7

The fifty-five miles from Campbell to San Francisco is one of the nicest commutes anywhere. The journey mostly zips along the Junipero Serra Freeway, a grand and remarkably empty highway that abuts the east side of the Santa Cruz Mountains. Known as 280 to locals, it is one of the best places in Silicon Valley to spot a start-up tycoon speed-testing his Ferrari and one of the worst places for cell phone reception. For Andy Grignon in his Porsche Carrera, therefore, it was the perfect place for him to be alone with his thoughts early on January 8, 2007.

This wasn’t Grignon’s typical route to work. He was a senior engineer at Apple in Cupertino, the town just west of Campbell. His morning drive typically covered seven miles and took exactly fifteen minutes. But today was different. He was going to watch his boss, Steve Jobs, make history at the Macworld trade show in San Francisco. Apple fans had for years begged Jobs to put a cell phone inside their iPods so they could stop carrying two devices in their pockets. Jobs was about to fulfill that wish. Grignon and some colleagues would spend the night at a nearby hotel, and at 10:00 a.m. the following day they—along with the rest of the world—would watch Jobs unveil the first iPhone.

Getting invited to one of Jobs’s famous product announcements was supposed to be a great honor. It anointed you as a player. Only a few dozen Apple employees, including top executives, got an invite. The rest of the spots were reserved for Apple’s board of directors, CEOs of partners—such as Eric Schmidt of Google and Stan Sigman at AT&T—and journalists from around the world. Grignon got an invite because he was the senior engineer for all the radios in the iPhone. This is a big job. Cell phones do innumerable useful things for us today, but at their most basic they are fancy two-way radios. Grignon was in charge of the equipment that allowed the phone to be a phone. If the phone didn’t make calls, connect with Bluetooth headsets, or connect to Wi-Fi setups, Grignon had to answer for it. As one of the iPhone’s earliest engineers, he’d dedicated two and a half years of his life—often seven days a week—to the project. Few deserved to be there more than he did.

But as Grignon drove north, he didn’t feel excited. He felt terrified. Most onstage product demonstrations in Silicon Valley are canned. The thinking goes, why let bad Internet or cell phone connections ruin an otherwise good presentation? Jobs’s presentations were live, however. It was one of the things that made his shows so captivating. But for those in the background, such as Grignon, few parts of the job caused more stress. Grignon couldn’t remember the last time a Jobs show of this magnitude had gone sideways. Part of what made Steve Jobs such a legend was that noticeable product-demo glitches almost never happened. But Grignon found it hard to recall the last time Jobs was so unprepared going into a show.

Grignon had been part of the iPhone launch-preparation team at Apple and later at the presentation site in San Francisco’s Moscone Center. But he had rarely seen Jobs make it all the way through his ninety-minute show without a glitch. Jobs had been rehearsing for five days, yet even on the last day of rehearsals the iPhone was still randomly dropping calls, losing the Internet connection, freezing, or just shutting down.

“At first it was just really cool to be at rehearsals at all—kind of like a cred badge. ‘Fuck yeah, I get to hang out with Steve,’” Grignon said. Like everything else that surrounded Jobs, the preparations were as secret as a U.S. missile attack on Afghanistan. Those who were truly in felt as if they were at the center of the universe. From Thursday through the end of the following week, Apple completely took over Moscone. Backstage it built an eight-by-eight-foot electronics lab to house and test the iPhones. Next to that it built a greenroom with a sofa for Jobs. Then it posted more than a dozen security guards twenty-four hours a day in front of those rooms and at doors throughout the building. No one got in or out without having his or her ID electronically checked and compared with a master list that Jobs had personally approved. More security checkpoints needed to be cleared once visitors got inside. The auditorium where Jobs was rehearsing was off-limits to all but a small group of executives. Jobs was so obsessed with leaks that he tried to have all the contractors Apple had hired for the announcement—from people manning booths and doing demos to those responsible for lighting and sound—sleep in the building the night before his presentation. Aides talked him out of it.

“It quickly got really uncomfortable,” Grignon said. “Very rarely did I see him become completely unglued. It happened. But mostly he just looked at you and very directly said in a very loud and stern voice, ‘You are fucking up my company,’ or, ‘If we fail, it will be because of you.’ He was just very intense. And you would always feel an inch tall [when he was done chewing you out].” Grignon said that you would always ask yourself two questions during one of these lectures: “‘Is it my shit that broke this time?’ and ‘Is it the nth time it broke or the first time?’—because that actually mattered. The nth time would frustrate him, but by then he might have figured out a way around it. But if it was the first time, it added a whole new level of instability to the program.” Grignon, like everyone else at rehearsals, knew that if those glitches showed up during the real presentation, Jobs would not be blaming himself for the problems, he would come after people like Grignon. “It felt like we’d gone through the demo a hundred times and that each time something went wrong,” Grignon said. “It wasn’t a good feeling.”

The iPhone didn’t work right for a good reason; it wasn’t close to being finished. Jobs was showing off a prototype. He just didn’t want the public to know that. But the list of things that still needed to be done before the iPhone could be sold was enormous. A production line had yet to be set up. Only about a hundred iPhones even existed, all of them of varying degrees of quality. Some had noticeable gaps between the screen and the plastic edge, others had scuff marks on the screen. Thus no one in the public was allowed to touch an iPhone after Jobs unveiled it, despite a day of press briefings and a whole exhibit set up for them in the convention center. The worry was that even the best prototypes wouldn’t stand close scrutiny, Grignon said. They’d look fine at a distance and for Jobs’s demo, but if you held one in your hand, “You would laugh and say, ‘Wow, this thing really looks unfinished.’”

The phone’s software was in even worse shape. A big chunk of the previous four months had been consumed figuring out why the iPhone’s processor and its cell radio wouldn’t reliably communicate. This huge problem was akin to a car with an engine that occasionally doesn’t respond to the accelerator, or wheels that occasionally don’t respond to the brake pedal. “It almost brought the iPhone program to a halt,” Grignon said. “We had never seen a problem this complicated.” This was ordinarily not a problem for phone makers, but Apple’s obsession with secrecy had kept Samsung, the manufacturer of the phone’s processor, and Infineon, the maker of the phone’s cell radio, from working together until Apple, in desperation, flew teams of engineers from each company to Cupertino to help fix the problem.

Jobs rarely backed himself into corners like this. He was well-known as a master taskmaster, seeming to always know just how hard he could push his staff so that they delivered the impossible. But he always had a backup, a Plan B, that he could go to if his timetable was off. Six months prior he’d shown off Apple’s upcoming operating system, Leopard. But that was after letting the date for the final unveiling slip.

But Jobs had no choice⁸ but to show off the iPhone. He had given this opening keynote at every Macworld since he’d returned as Apple’s CEO in 1997, and because he gave public presentations only once or twice a year, he had conditioned Apple fans to expect big things from them. He’d introduced iTunes here, the iMac that looked like a fancy desk lamp, the Safari web browser, the Mac mini, and the iPod shuffle.

It wasn’t just his own⁹ company that Jobs had to worry about disappointing this time. AT&T was expecting Jobs to unveil the iPhone at Macworld too. In exchange for being the exclusive carrier of the iPhone in the United States, AT&T had given Jobs total control of the design, manufacture, and marketing of the iPhone. It had never done anything like this before. If Jobs didn’t launch on time, AT&T could back out of its deal. It’s not hard to explain that a product called the iPhone that couldn’t make calls would sell poorly. Days before, Jobs had flown to Las Vegas to give AT&T’s top mobile executives a limited demo of the iPhone. But they were expecting a full show at Macworld.

Lastly, the iPhone was truly the only cool new thing Apple was working on. The iPhone had been such an all-encompassing project at Apple that this time there was no backup plan. “It was Apple TV or the iPhone,” Grignon said. “And if he had gone to Macworld with just Apple TV [an experimental product back then], the world would have said, ‘What the hell was that?’”

The iPhone’s problems were manifest. It could play a section of a song or a video, but it couldn’t play an entire clip without crashing. It worked fine if you sent an email and then surfed the web. If you did those things in reverse, however, it did not. Hours of trial and error had helped the iPhone team develop what engineers called “the golden path,” a specific set of tasks, performed in a specific way and in a specific order, that made the phone look as if it worked.

But even when Jobs stayed on the golden path, it required all manner of last-minute work-arounds to make the iPhone functional. On announcement day the software that ran Grignon’s radios still had bugs. So too did the software that managed the iPhone’s memory. And no one knew whether the extra electronics Jobs had required to be added to the demo units would make these problems worse.

Jobs had required the demo phones he would use onstage to have their screens mirrored on the big screen behind him. To show a gadget on a big screen, most companies just point a video camera connected to a projector at the gadget. That was unacceptable to Jobs. The audience would see his finger on the iPhone screen, which would mar the look of his presentation. Instead, he had Apple engineers spend weeks fitting extra circuit boards attached to video cables onto the backs of the iPhones he would have onstage. The video cables then connected to the projector showing the iPhone image on the screen. When Jobs touched the iPhone’s calendar app icon, for example, his finger wouldn’t appear, but the image on the big screen would respond. The effect was magical. People in the audience felt as if they were holding an iPhone in their own hands. But making the setup work flawlessly given the iPhone’s other major problems seemed hard to justify at the time. “It was all just so monkey-patched together with some of the ugliest hacks you could imagine,” Grignon said.

The software in the iPhone’s Wi-Fi radio was so unstable that Grignon and his team ultimately soldered antenna wires to the demo phones and ran them offstage along the wires to the projection setup. The iPhone would still connect wirelessly to the network, but the signal wouldn’t have to travel as far. Even then, Grignon and his team needed to make sure no one in the audience could get on the frequency they were using. “Even if the base station’s ID was hidden [and therefore not showing up when laptops scanned for Wi-Fi signals], you had five thousand nerds in the audience. They would have figured out how to hack into the signal.” The solution, Grignon said, was simply to tweak the AirPort software to think it was operating in Japan instead of the United States. Japanese Wi-Fi uses some frequencies that are not permitted in the U.S.

There was even less they could do to make sure the phone call Jobs planned to make from the stage went through. All Grignon and his team could do was make sure the signal was good and pray. They had AT&T bring in a portable cell tower so they knew reception would be strong. Then, with Jobs’s support, they preprogrammed the phone’s display to always show five bars of signal strength regardless of the true signal. The chances of the radio’s crashing during the few minutes that Jobs would use it to make a call were small, but the chances of its crashing at some point during the ninety-minute presentation were high. “If the radio crashed and restarted, as we suspected it might, we didn’t want people in the audience to see that. So we just hard-coded it to always show five bars,” Grignon said.

None of these kluges fixed the iPhone’s biggest problem: it often ran out of memory and had to be restarted if asked to do more than a handful of tasks at a time. Jobs had a number of demo units onstage with him to manage this problem. If memory ran low on one, he’d switch to another while the first was restarted. But given how many demos Jobs planned, Grignon worried that there were far too many potential points of failure. If disaster didn’t strike during one of the dozen demos, it was sure to happen during Jobs’s grand finale, when Jobs planned to show all the iPhone’s top features operating at the same time on the same phone. He’d play some music, take a call, put it on hold and take another call, find and email a photo to the second caller, look up something on the Internet for the first caller, and then return to his music. “Me and my guys were all so nervous about this. We only had 120 megabytes of memory in those phones, and because they weren’t finished, all these apps were still big and bloated,” Grignon said.

The idea that one of the biggest moments of his career might implode made Grignon’s stomach hurt. At forty, Grignon looks like the kind of guy you’d want to drink with—and he is. When he moved from Campbell to Half Moon Bay in 2010, he quickly became friendly with the sommelier at the Ritz-Carlton Hotel. He even had a wine fridge in his office. But behind that gregarious exterior is a fierce intellect and an ultracompetitive streak. Once when trying to get to the bottom of a slew of software bugs in an iPhone subcontractor’s equipment, he turned the AC on high in the conference room he used to make the subcontractors uncomfortably cold. When that didn’t get them moving fast enough, he tried a more aggressive approach: he accused them of holding out on him and threw his laptop against the wall.

By 2007 he’d spent virtually his entire fifteen-year career at Apple or companies affiliated with it. While at the University of Iowa in 1993, he and his friend Jeremy Wyld—now cofounder with Grignon of Quake Labs—reprogrammed the Newton MessagePad to wirelessly connect to the Internet. That was quite a feat back then, and it helped them both get jobs at Apple right out of school. Wyld actually worked on the Newton team, and Grignon worked in Apple’s famous R & D lab—the Advanced Technology Group—on video conferencing technology. Even though the Newton did not succeed as a product, many still think of it as the first mainstream handheld computer. But by 2000 Grignon had found his way to Pixo, a company spun out of Apple that was building operating systems for cell phones and other small devices. When Pixo’s software found its way into the first iPod in 2002, Grignon found himself back at Apple again.

By then, thanks to his work at Pixo, he’d become well known for two other areas of expertise besides building video conferencing technology: computer radio transmitters (what we now call wireless) and the workings of software inside small handheld devices such as cell phones. Grignon works in an entirely different world from that inhabited by most software engineers in the Valley. Most rarely have to think about whether their code takes up too much space on a hard drive or overloads a chip’s abilities. Hardware on desktop and laptop computers is both powerful, modifiable, and cheap. Memory, hard drives, even processors, can be upgraded inexpensively, and computers are either connected to electric outlets or giant batteries. In Grignon’s world of embedded software, the hardware is fixed. Code that is too big won’t run. Meanwhile, a tiny battery—which might power a laptop for a couple of minutes—needs enough juice to last all day. When Jobs decided to build the iPhone at the end of 2004, Grignon had a perfect set of skills to become one of the early engineers on the project.

Now, in 2007, he was emotionally exhausted. He’d gained fifty pounds. He’d stressed his marriage. It had been a grueling two years. Apple had never built a phone before, and the iPhone team quickly discovered the process didn’t resemble building computers or iPods at all. “It was very dramatic,” Grignon said. “It had been drilled into everyone’s head that this was the next big thing to come out of Apple. So you put all these supersmart people with huge egos into very tight, confined quarters, with that kind of pressure, and crazy stuff starts to happen.”

The iPhone didn’t start out as Apple’s “next big thing.” Jobs had to be talked into building a phone. It had been a topic of conversation among his inner circle almost from the moment Apple launched the iPod in 2001. The conceptual reasoning was obvious: Why would consumers carry two or three devices for email, phone calls, and music when they could carry one?

But every time Jobs and his executives examined the idea in detail, it seemed like a suicide mission. Phone chips and bandwidth were too slow for anyone to want to surf the Internet and download music or video over a cell phone connection. Email was a fine function to add to a phone. But Apple couldn’t leverage all the work it had put into building a music player such as the iPod to do that. Research in Motion’s BlackBerry was fast locking up that market, anyway. Apple even considerd buying Motorola in 2003, but executives quickly concluded it would be too big an acquisition for the company then.

Worst of all¹⁰, if Apple wanted to make and sell a phone in the United States, it would be at the beck and call of the U.S. wireless carriers. Back then, phone manufacturers such as Motorola were the serfs of high tech in the United States. They depended on carriers’ marketing dollars to get consumers into stores, and then they depended on carriers to make the phones affordable by subsidizing their purchase price. That made manufacturers powerless to resist carriers’ meddling in how each phone should be built. Manufacturers occasionally pushed back against this dominance and were always met with the same response from the carriers: “You can build the phone your way, but we might not subsidize it, market it, or allow it on our network.” Manufacturers always caved in the face of this threat.

Jobs was personally offended¹¹ by this way of doing business and wanted no part of it. “We’re not the greatest at selling to the Fortune 500, and there are five hundred of them—five hundred CIOs [chief information officers] that are orifices you have to go through to get” that business. “In the cell phone business there are five. We don’t even like dealing with five hundred companies. We’d rather run an ad for millions and let everyone make up their own mind. You can imagine what we thought about dealing with five,” he said during an onstage interview at the All Things D conference in May 2003. Translation: I am not about to spend hundreds of millions of dollars to have a bunch of suits tell me how to build and sell my phone.

That sounded tough and principled. But by the end of 2003, as the iPod became Apple’s most important product since the Macintosh, it was also starting to look misguided. Cell phone makers were putting music-listening applications in their phones. And companies such as Amazon, Walmart, and Yahoo! were beginning to sell downloadable music. Executives such as iPod boss Tony Fadell worried that if consumers suddenly gave up their iPods for music phones, Apple’s business—only five years removed from its flirt with bankruptcy—would be crushed. “We didn’t really have a hit on our hands [with the iPod] until late 2003, early 2004, so we were saying maybe we don’t have the market domination—the retail channels—to expand the iPod’s business properly,” Fadell said.

It’s hard to imagine¹² a time when the iPod wasn’t an iconic product, selling more than 50 million units a year; but back then Apple had sold only 1.3 million devices in two years and was still having trouble getting retailers such as Best Buy to carry it. “So we were thinking, ‘How do we get above the noise? How do we make sure that we are at least competitive so that anyone who is carrying a cell phone can get iTunes music?’ Because if we lost iTunes, we would have lost the whole formula,” Fadell said.

Publicly, Jobs continued his¹³ harangue against the carriers. At the D conference in 2004, Stewart Alsop, Jr., the venture capitalist and former journalist, actually begged Jobs to make a smartphone that would improve on the popular Treo. “Is there any way you can get over your feelings about the carriers?” Alsop asked, offering to connect Jobs with Verizon CEO Ivan Seidenberg, who was also in the audience. Not a chance, Jobs said. “We’ve visited with the handset manufacturers and even talked to the Treo guys. They tell us horror stories.” But privately, Jobs was thinking hard about the content of Alsop’s pitch.

Jobs’s first answer to the growing competition wasn’t the iPhone, but something much more modest—a music phone called the Rokr, to be built in partnership with Motorola and Cingular, the big wireless carrier that would, via two mergers, become AT&T. The deal, agreed to in early 2004, seemed like the best of all worlds for Apple. It would license its iTunes software to Motorola to be put on Motorola’s supersuccessful Razr cell phone, and Motorola would handle the rest. Apple would get a license fee for letting Motorola use the software, and Jobs wouldn’t have to deal with the wireless carriers. iTunes would help Motorola sell more phones, get Cingular more wireless customers, and enable Apple to compete with the music phones it feared. “We thought that if consumers chose to get a music phone instead of an iPod, at least they would be using iTunes,” Fadell said.

Instead, the Rokr was an embarrassment. When Jobs unveiled it nearly eighteen months later in September 2005, it was not capable of over-the-air music downloads, the device’s main selling point. It was big and chunky—nothing like the sleek Razr that Motorola had made famous. And its music capacity was artificially limited to a hundred songs.

The tension between the partners¹⁴, especially Apple and Motorola, was obvious quickly after Jobs was done demoing the device onstage at Moscone Center in San Francisco. Jobs had released the first iPod nano at the same time, and when a reporter asked Motorola CEO Ed Zander a few weeks later if he felt upstaged by the other products Jobs had unveiled, his answer was succinct: “Screw the nano.” Wired magazine soon put a story of the fiasco on its cover under the headline YOU CALL THIS THE PHONE OF THE FUTURE?

Jobs successfully pinned the Rokr screwup¹⁵ on Motorola, but the fiasco was mostly Apple’s fault. Yes, Motorola had produced an ugly phone, and it continued to produce phones that didn’t sell well for the next four years until Zander resigned. But the Rokr project’s real problem was that Jobs’s reason for the deal evaporated almost as soon as it was signed, Fadell said. The deal was designed as a defensive maneuver, a hedge against companies’ trying to build music phones without having to deal with the carriers themselves. But with each passing month in 2004 it became clearer that the last thing Apple needed to do with iTunes and the iPod was to play defense. It didn’t need the Rokr to help it more broadly distribute iTunes. It just needed to hang on as iPod sales took off like a rocket ship. In the summer 2003 Apple was selling only about three hundred thousand iPods a quarter. At the beginning of 2004 it was selling only eight hundred thousand a quarter. But by summer 2004 sales exploded. It sold 2 million during the quarter that ended September 30, 2004, and another 4.5 million in the final quarter of the year. By the time ugly Rokr prototypes showed up in the fall of 2004, many Apple executives saw clearly that they were on the wrong path, and by year-end Jobs had all but abandoned the project. He was still driving the iTunes team to deliver the software that would go in the Rokr, but he was listening more carefully to executives who thought the Rokr project had been folly from the start.

It wasn’t just the iPod’s success in 2004 that diluted Apple’s enthusiasm for the Rokr. By the end of the year, building its own phone no longer seemed like such a bad idea. By then it looked like most homes and cell phones would soon have Wi-Fi, which would provide high, reliable bandwidth over the homeowner’s DSL or cable connection. And outside-the-home cell phone bandwidth looked like it would soon be fast enough to stream video and run a fully functioning Internet browser. Phone processor chips were finally fast enough to run cool-looking phone software. Most important, doing business with the carriers was starting to seem less onerous. By the fall of 2004, Sprint was beginning to sell its wireless bandwidth wholesale. That meant that by buying and reselling Sprint bandwidth, Apple could become its own wireless carrier—an MVNO, short for “mobile virtual network operator.” Now Apple could build a phone and barely have to deal with the carriers at all. Disney, on whose board¹⁶ Jobs sat, was already in discussions with Sprint about just such a deal to provide its own wireless service. Jobs was asking a lot of questions about whether Apple should pursue one as well.

Cingular executives involved in the Rokr project such as Jim Ryan watched Jobs’s interest in an MVNO with Sprint grow, and it terrified them. They worried that if Apple became a wireless carrier, it would cut prices to win customers and crush profits in the industry as other carriers cut prices to compete. So while they had access to Jobs and his team, they gently lobbied him to cut a deal with them instead. If Jobs would agree to an exclusive deal with Cingular, they said, they would be willing to throw out the rule book on carrier–manufacturer relations and give Jobs the control he needed to build a revolutionary device.

Ryan, who has never talked publicly about those days until now, said the experience taxed every ounce of his negotiating skills. He’d been assembling complex carrier deals for nearly a decade and was known in the industry as one of the early thinkers about the future of wireless. He’d grown Cingular’s wireless data business from almost nothing to $4 billion in revenue in three years. But Apple and Jobs had little experience negotiating with carriers, making it much harder for Ryan to predict how they would respond to his various offers. “Jobs hated the idea of a deal with us at first. Hated it,” Ryan said. “He was thinking that he didn’t want a carrier like us anywhere near his brand. What he hadn’t thought through was the reality of just how damn hard it is to deliver mobile service.” Throughout 2004, during the dozens of hours he and his team spent in meetings with Apple executives in Cupertino, Ryan kept reminding Jobs and other Apple executives that if Apple became a carrier itself, it would get stuck with all the hassles of running an inherently unpredictable asset—a cell phone network. A deal with Cingular would insulate Apple from all that. “Funny as it sounds, that was one of our big selling points to them,” Ryan said. “Every time the phone drops a call, you blame the carrier. Every time something good happens, you thank Apple.”

Cingular wasn’t just playing defense¹⁷. Executives such as Ryan thought partnering with the inventor of the iPod would transform the way customers thought about their own company. Apple’s explosive success with the iPod in 2004 and 2005—it sold 8.2 million iPods in 2004 and another 32 million in 2005—had taken Jobs’s status as a business and cultural icon to unparalleled heights. The likely torrent of new customers who would come to Cingular if it were the carrier for a phone as revolutionary as the iPod had been made them salivate.

Another Cingular executive who worked on the deal but who would not be named put it this way to me when I was working on a story for Wired in 2008: “Jobs was cool. He was hip. There were studies done in colleges that asked, ‘What is the one thing you can’t live without?’ For twenty years it was beer. Now it was the iPod. Things like that made us say this guy has got something. That probably gave us that much more energy to make sure this deal happened.”

While Cingular was lobbying Jobs from the outside, a handful of Apple executives, such as Mike Bell and Steve Sakoman, were pushing Jobs to sign off on building a phone from the inside. “We were spending all this time putting iPod features in Motorola phones. That just seemed ass-backwards to me,” said Bell, who now is cohead of Intel’s mobile-device effort. He told Jobs that the cell phone itself was on the verge of becoming the most important consumer electronics device of all time, that no one was good at making them, and that, therefore, “if we [Apple] just took the iPod-user experience and some of the other stuff we were working on, we could own the market.”

Bell was a perfect executive to be making this pitch. He’d been at Apple fifteen years and had helped build some of the products, such as the iMac, that enabled Apple to avoid bankruptcy in 1997. Most important, because he ran not only a chunk of the Mac software division but the software group responsible for Apple’s AirPort Wi-Fi devices, he knew more about the wireless industry than most other senior executives inside Apple. He doesn’t claim credit for being the father of the iPhone. He ultimately didn’t run or even work on the project. Fadell ran it, before Scott Forstall took it over. But even today most say Bell was an important catalyst.

“So I argued with Steve for a couple of months and finally sent him an email on November seventh, 2004,” Bell said. “I said, ‘Steve, I know you don’t want to do a phone, but here’s why we should do it: [Design director Jony Ive] has some really cool designs for future iPods that no one has seen. We ought to take one of those, put some Apple software around it, and make a phone out of it ourselves instead of putting our stuff on other people’s phones.’ He calls me back about an hour later and we talk for two hours, and he finally says, ‘Okay, I think we should go do it.’ So Steve and I and Jony [Ive] and Sakoman had lunch three or four days later and kicked off the iPhone project.”

It wasn’t just Bell’s persistence and Ive’s designs that helped convince Jobs. Sakoman came to lunch having already done some early engineering work about what it might take to build a phone. He’d been at Palm until 2003, where, among other things, he helped build the software that went inside Treo smartphones. And as vice president of software technology at Apple, he had become the executive most familiar with the software inside the iPod. If Apple was going to make a smartphone, the iPod was a logical place to start. That’s what consumers were expecting Apple to do. So by the time Sakoman arrived for lunch, he and his team had already figured out a way to put a Wi-Fi chip inside an iPod and get it to connect to the Internet.

They’d even begun working on new software for the music player—a version of Linux—so that it could handle the increased demands of being a phone and an Internet browser. Linux, the open-source software made famous by Linus Torvalds in the 1990s, had not supplanted Microsoft Windows as many geeks predicted it would. But by then it had become the software of choice for less powerful and sophisticated electronics. Sakoman briefed Jobs on his team’s progress and later that afternoon told his team, “You better start figuring this out because this [phone project] is going ahead.”

Bell says one reason why he remembers the meeting is that he’d never seen anyone eat the way Jobs did that day. “You know how you remember certain things because of their bizarreness? So we’re meeting outside at the Apple cafeteria, and when Steve walks out, on his tray is a glass bowl full of avocado halves. Not one or two, but, like, fifteen covered in salad dressing. So I remember sitting there with Jony and Sakoman and watching Steve mow through a mound of avocados. I guess, having read Walter Isaacson’s biography [of Jobs], it was one of those food phases he was in to cure his cancer, but at the time I had no idea what was up.”

The final deal between Apple and AT&T, which acquired Cingular in 2006, took more than a year to hammer out. But it would prove easy compared to what Apple went through just to build the device. Many executives and engineers, riding high from their success with the iPod, assumed it would be just like building a small Macintosh. Instead, Apple designed and built not one iPhone but three entirely different devices in those two years. One executive on the project thinks Apple made six fully working prototypes just of the device it ultimately sold—each with its own set of hardware, software, and design tweaks. Many on the team were so burned-out, they left the company shortly after the first phone hit store shelves. “It was like the first moon mission,” said Fadell, who was one of the key executives on the project, and who left Apple to start his own company, Nest, in 2010. “I’m used to a certain level of unknowns in a project, but there were so many new things here that it was just staggering.”

Jobs wanted the iPhone to run a modified version of OS X, the software that comes with every Mac. But no one had ever put¹⁸ a gigantic program like OS X on a phone chip before. The software would have to be a tenth the size, and even then there wasn’t a phone chip being made in 2005 that could run it fast enough and with a long enough battery life. The chips that run Apple laptops were never considered because they generated too much heat and would suck a phone battery dry in minutes. Millions of lines of code would have to be stripped out or rewritten, and until 2006 engineers would have to simulate chip speed and battery drain because actual chips weren’t available until then. “Initially we just worked on Gumstix boards [cheap circuit boards hobbyists buy],” said Nitin Ganatra, one of the early software engineers. “We started with the Mac address book—a list of names—and to see if we could make it scroll [on a screen] at between thirty to sixty frames a second. We just wanted to figure out if there was any way to make this [OS X on a phone chip] work—whether we were even in the right ballpark. We wanted to know if we could push bits fast enough to get that iPhone look and feel. If we couldn’t get it to work on a Gumstix board, we knew we might have a problem.”

No one had ever put a capacitive multitouch screen in a mainstream consumer product before either. Capacitive touch technology—which creates “a touch” when a finger or other conductive item completes a circuit on the device—had been around since the 1960s. Elevator buttons in office buildings and screens on ATMs often used it. And research into multitouch technologies had been around since the 1980s. Trackpads on laptops were probably the most sophisticated use of this technology because they could recognize the difference between one- and two-finger inputs. But it was also well known that to build the multitouch screen Apple put on the iPhone and produce it in volume was a challenge few had the money or guts to take on. The next steps—to embed the technology invisibly in a piece of glass, to make it smart enough to display a virtual keyboard with auto-correct, and to make it sophisticated enough to reliably manipulate content such as photos or web pages on that screen—made it hugely expensive even to produce a working prototype. Few production lines even had experience manufacturing multitouch screens. There were touchscreens in consumer electronics, but over the years these had typically been pressure-sensitive touchscreen devices on which users pushed on-screen buttons with a finger or a stylus. The PalmPilot and its successors such as the Palm Treo were popular implementations of this technology. Even if multitouch iPhone screens had been easy to make, it wasn’t at all clear to Apple’s executive team that the features they enabled, such as onscreen keyboards and “tap to zoom,” were enhancements that consumers wanted.

As early as 2003 a handful of Apple engineers, who had done cutting-edge academic work with touch interfaces, had figured out how to put multitouch technology in a tablet. But the project was mothballed. “The story was that Steve wanted a device that he could use to read email while on the toilet. That was the extent of the product spec,” said Josh Strickon, one of the earliest engineers on that project. “But you couldn’t build a device with enough battery life to take out of the house, and you couldn’t get a chip with enough graphics capability to make it useful. We spent a lot of time trying to figure out just what to do.” Before joining Apple in 2003, Strickon had been a student at MIT for a decade, getting his B.A., master’s, and Ph.D. in engineering. He was a huge proponent of touchscreen technology, having built a multitouch device for his master’s thesis. But he said given the lack of consensus at Apple about what to do with the prototypes he and his fellow engineers developed, he left the company in 2004 thinking it wasn’t going to do anything with multitouch.

Tim Bucher, one of Apple’s top executives at the time and the company’s biggest multitouch proponent, said part of the problem was that the prototypes they were building used software, OS X, that was designed to be used with a mouse, not a finger. “We were using ten- or twelve-inch screens with Mac mini–like guts … and then you would launch these demos that would do the different multitouch gestures. One demo was a keyboard application that would rise from the bottom—very much what ended up shipping in the iPhone two years later. But it wasn’t very pretty. It was very much wires, chewing gum, and bailing wire. It left too much to the imagination.” Bucher, who has never before talked publicly about his work at Apple, had hoped to keep pushing the effort forward, but he lost a political battle with other top executives and left Apple in early 2005.

Few even thought about making touchscreen technology the centerpiece of a new kind of phone until Jobs started pushing the idea in mid-2005. “He said, ‘Tony, come over here. Here’s something we’re working on. What do you think? Do you think we could make a phone out of this?’” Fadell said. “So we sat there and played with the demo (he was showing me) for a while. It was huge. It filled the room. There was a projector mounted on the ceiling and it would project the Mac screen onto this surface that was maybe three or four feet square. Then you could touch the Mac screen and move things around and draw on it. I knew about it [the touchscreen prototype], but I didn’t know about it in detail because it was a Mac thing [Fadell ran the iPod division]. So we all sat down and had a serious discussion about it—about what could be done.”

Fadell had serious doubts about whether such an enormous prototype could be shrunk so much. But he also knew better than to answer no to Steve Jobs. He was one of Apple’s superstars, and he didn’t get there by being timid about thorny technological problems. He’d joined Apple in 2001 as a consultant to help build the first iPod. By 2005, with iPod sales exploding, he had become, at thirty-six, arguably the single most important line executive at the company.

“I understood how it could be done,” Fadell said. “But it’s one thing to think that, and another to take a room full of special, one-off gear and make a million phone-sized versions of that in a cost-effective, reliable manner.” The to-do list was exhausting just to think about. “You had to go to LCD vendors [companies that make the screens that go in computer monitors and TVs] who knew how to embed technology like this in glass; you had to find time on their line; and then you had to come up with compensation and calibrating algorithms to keep the pixel electronics [in the LCD] from generating all kinds of noise in the touchscreen [sitting on top of it.] It was a whole project just to make the touchscreen device. We tried two or three ways of actually making the touchscreen until we could make one in enough volume that would work.”

Shrinking OS X and building a multitouch screen, while innovative and difficult, were at least within the skills Apple had already mastered as a corporation. No one was better equipped to rethink OS X’s design. Apple knew LCD manufacturers because it put an LCD in every laptop and iPod. The peculiarities of mobile phone physics, on the other hand, were an entirely new field, and it took those working on the iPhone into 2006 to realize how little they knew.

To ensure the iPhone’s¹⁹ tiny antenna could do its job effectively, Apple spent millions buying and assembling special robot-equipped testing rooms. To make sure the iPhone didn’t generate too much radiation, Apple built models of human heads—complete with goo to simulate brain density—and measured the effects. To predict the iPhone’s performance on a network, Apple engineers bought nearly a dozen server-size radio-frequency simulators for millions of dollars apiece. One senior executive believes Apple spent more than $150 million building the first iPhone.

The first iPhone prototype was not ambitious. Jobs hoped that he would be able to develop a touchscreen iPhone running OS X. But in 2005 he had no idea how long that would take. So Apple’s first iPhone looked very much like the joke slide Jobs had put up when introducing the real iPhone—an iPod with an old fashioned rotary dial on it. The prototype was an iPod with a phone radio that used the iPod click wheel as a dialer. It grew out of the work Steve Sakoman had used to pitch Jobs on a phone project in the first place. “It was an easy way to get to market, but it was not cool like the devices we have today,” Grignon said. He worked for Sakoman at the time and is one of the names on the click wheel dialer patent.

The second iPhone prototype in early 2006 was much closer to what Jobs would ultimately unveil. It incorporated a touchscreen and OS X, but it was made entirely of brushed aluminum. Jobs and Ive were exceedingly proud of it. But since neither of them were experts in the physics of radio waves, they hadn’t realized they’d created a beautiful brick. Radio waves don’t travel through metal well. “I and Ruben Caballero [Apple’s antenna expert] had to go up to the boardroom and explain to Steve and Ive that you cannot put radio waves through metal,” said Phil Kearney, one of Bell’s deputies, who left in 2008. “And it was not an easy explanation. Most of the designers are artists. The last science class they took was in eighth grade. But they have a lot of power at Apple. So they ask, ‘Why can’t we just make a little seam for the radio waves to escape through?’ And you have to explain to them why you just can’t.”

Jon Rubinstein, Apple’s top hardware executive then and known to many as the Podfather for driving the creation and development in the iPod, said there were even long discussions about how big the phone would be. “I was actually pushing to do two sizes—to have a regular iPhone and an iPhone mini like we had with the iPod. I thought one could be a smartphone and one could be a dumber phone. But we never got a lot of traction on the small one, and in order to do one of these projects you really need to put all your wood behind one arrow.”

It all made the iPhone²⁰ project so complex that it occasionally threatened to derail the entire corporation. Many of the top engineers in the company were being sucked into the project, forcing slowdowns in the timetables of other projects. Had the iPhone been a dud or not gotten off the ground at all, Apple would have had no other big products ready to announce for a long time. Worse, its top engineers, frustrated by the failure, would have left Apple for other jobs, according to 2012 testimony by Scott Forstall, one of Apple’s top executives on the project and Apple’s head of iOS software until October 2012. He testified during the Apple v. Samsung patent trial.

Even Apple’s experience designing screens for iPods didn’t help the company design the iPhone screen. After much debate, Jobs decided the iPhone screen needed to be made of hard Plexiglas. He and his executives thought a glass screen would shatter when dropped—until Jobs saw how scratched a plastic prototype had gotten when he carried it around in his pocket with his keys. “Jobs goes, ‘Look at this. Look at this. What’s with the screen?’” said an executive who witnessed the exchange. “And the guy [a midlevel executive] takes the prototype and says, ‘Well, Steve, we have a glass prototype, but it fails the one-meter drop test one hundred out of one hundred times, and blah blah blah …’ Jobs cuts him off and says, ‘I just want to know if you are going to make the fucking thing work.’”

There was a good reason the executive argued with Jobs. This was September 2006. The iPhone would be unveiled in four months. And Jobs wanted to rethink the phone’s most prominent component.

Through his friend²¹ John Seely Brown, Jobs reached out to Wendell Weeks, the CEO of glassmaker Corning in upstate New York, invited him to Cupertino, and told him he needed the hardest glass ever made for the screen of the iPhone. Weeks told him about a process developed for fighter-jet cockpits in the 1960s. But Weeks said the Defense Department never ended up using the material, known as gorilla glass, so it had never found a market. He said Corning had stopped making it decades ago. Jobs wanted Weeks to start production immediately, convincing Weeks that he could in fact get Jobs the glass he needed in six months. Weeks told Jobs’s biographer Walter Isaacson that he remains amazed at what Jobs convinced him to do. Corning took a factory in Harrodsburg, Kentucky, that had been making LCD displays and converted it, getting Jobs the glass he needed on time. “We produced glass that had never been made. We put our best scientists and engineers on it and we just made it work,” Weeks said.

“I still remember PC Magazine doing a screen durability test once the phone came out in July 2007,” said Bob Borchers, Apple’s then head of iPhone marketing. “They put it in a bag of coins and shook it up. They put keys in the bag and shook it up. They dropped it a few times on a carpet. And then they went out on the street and dropped it on the concrete three times. It survived all of that. We all laughed, looked at each other, and said, ‘Right, we knew that.’”

On top of all that, Jobs’s obsession with secrecy meant that despite being exhausted from working eighty hours a week, the few hundred engineers and designers working on the project couldn’t talk about the project to anyone else. If Apple found out you’d told a friend in a bar, or even your spouse, you could be fired. Before a manager could ask you to join the project, you had to sign a nondisclosure agreement in his office. Then, after he told you what the project was, you had to sign another document confirming that you had indeed signed the NDA and would tell no one. “We put a sign on over the front door of the iPhone building that said FIGHT CLUB because the first rule of fight club is you don’t talk about fight club,” Forstall would explain in his court testimony. “Steve didn’t want²² to hire anyone from outside of Apple to work on the software, but he said I could hire anyone in the company I wanted,” Forstall said. “So I’d bring recruits into my office. Sit them down and tell them, ‘You are a superstar at Apple. Whatever you are doing now, you’ll do fine. But I have another project that I want you to consider. I can’t tell you what it is. All I can say is that you will have to give up untold nights and weekends and that you will work harder than you have ever worked in your life.”

“My favorite part,” said one of the early iPhone engineers, “was what all the vendors said the day after the unveiling.” Big companies such as Marvell Electronics, which made the Wi-Fi radio chip, and CSR, which provided the Bluetooth radio chip, hadn’t been told they were going to be in a new phone. They thought they were going to be in a new iPod. “We actually had fake schematics and fake industrial designs,” the engineer said. Grignon said that Apple even went as far as to impersonate employees of another company when they traveled, especially to Cingular (and, later, AT&T) in Texas. “The whole thing was you didn’t want the receptionist or whoever happens to be walking by to see all [preprinted Apple] badges lying out.”

On the other hand, Jobs wanted a handful of the top engineers on the iPhone project to use iPhone prototypes as their permanent phones. “It wasn’t ‘Carry an iPhone—and a Treo,’” Grignon said. “It was ‘Carry an iPhone and live on it,’ because that’s how we found bugs. If you can’t make a phone call because of a bug, you are going to be extra-motivated to start yelling to get that fixed. But it made for some awkward times where, if you were, say, in a club or an airport, you could spot an iPhone user a mile away because they were the person hunched over with their arms around their phone doing something mysterious. Snorting a line of coke—or using an iPhone?”

One of the most obvious manifestations of Jobs’s obsession with secrecy was the growth of lockdown areas all over campus—places that those not working on the iPhone could no longer go. “Each building is split in half, and there is this corridor that runs through the middle of them with common areas, and after one weekend they just put doors around the common areas so that if you were not on the project, and you were used to using that space, it was now off-limits,” Grignon said. “Steve loved this stuff. He loved to set up division. But it was a big ‘fuck you’ to the people who couldn’t get in. Everyone knows who the rock stars are in a company, and when you start to see them all slowly get plucked out of your area and put in a big room behind glass doors that you don’t have access to, it feels bad.”

Even people within the iPhone project itself couldn’t talk to one another. Engineers designing the iPhone’s electronics weren’t allowed to see the software it would run. When they needed software to test the electronics, they were given proxy code, not the real thing. If you were working on the software, you used a simulator to test hardware performance.

And no one outside Jobs’s inner circle was allowed into chief designer Jony Ive’s wing on the first floor of Building 2. The security surrounding Ive’s prototypes was so tight that employees believed the badge reader called security if you tried to badge in and weren’t authorized. “It was weird, because it wasn’t like you could avoid going by it. It was right off the lobby, behind a big metal door. Every now and then you’d see the door open and you’d try to look in and see, but you never tried to do more than that,” said an engineer whose first job out of college was working on the iPhone. Forstall said during his testimony that some labs required you to “badge in” four times.

The four months leading up to announcement day were particularly rough, Grignon said. Screaming matches broke out routinely in the hallways. Engineers, frazzled from all-night coding sessions, quit, only to rejoin days later after catching up on their sleep. Forstall’s chief of staff, Kim Vorath, slammed the door to her office so hard that the handle bent and locked her in; it took colleagues more than an hour and some well-placed whacks with an aluminum bat to free her. “We were all standing there watching it,” Grignon said. “Part of it was funny. But it was also one of those moments where you step back and realize how fucked-up it all is.”

To Grignon’s amazement²³ and to that of many others in the audience, Jobs’s iPhone demo on January 9, 2007, was flawless. He started the show saying, “This is a day I have been waiting for two and a half years.” Then he regaled the audience with a myriad of tales about why consumers hated their cell phones. Then he solved all their problems—definitively. Virtually everyone in the audience had been expecting Jobs to announce a phone, yet they were still in awe.

He used the iPhone to play some music and watch a movie clip to show off the phone’s beautiful screen. He made a phone call to show off the phone’s reinvented address book and voice mail. He sent an email and a text, showing how easy it was to type on the phone’s touchscreen keyboard. He scrolled through a bunch of photos, showing how simple pinches and spreads of two fingers could make the pictures bigger or smaller. He navigated Amazon’s and The New York Times’ websites to show that the iPhone’s Internet browser was as good as the one on his computer. He found a Starbucks with Google Maps—and called the number from the stage—to show how it was impossible to get lost with an iPhone.

By the end, Grignon wasn’t just happy, he was drunk. He’d brought a flask of Scotch to calm his nerves. “And so there we were in the fifth row or something—engineers, managers, all of us—doing shots of Scotch after every segment of the demo. There were about five or six of us, and after each piece of the demo, the person who was responsible for that portion did a shot. When the finale came—and it worked along with everything before it, we all just drained the flask. It was the best demo any of us had ever seen. And the rest of the day turned out to be just a shit show for the entire iPhone team. We just spent the entire rest of the day drinking in the city. It was just a mess, but it was great.”