The room was tense when Schmidt rose to respond. He defused it with humor, said Lack, referring to himself “as the skunk at my garden party. I can feel in this room, shall we say, a certain indifference towards my contribution to all of our work together, and I feel sorry about that, because I think there are great contributions to be made working together.” Schmidt acknowledged that Google and the Internet can negatively affect newspapers and other media businesses, but ended by urging them to talk and search for ways to work together. Sulzberger said he had “no recollection of the specific incident,” adding, “You can certainly check with Eric on this.” Eric Schmidt confirmed Lack’s account.
“I admired Eric for the way he handled himself,” Lack said. “There was no armor to him, no bluff, no bravado.”
By 2004, relations between Schmidt and the founders were harmonious. The founders are happy with Schmidt, said one longtime Google executive who did not want to be quoted, because “Eric does everything they don’t want to do.” Bill Campbell sees it from another angle. He lavished praise on Page and Brin for their entrepreneurial brilliance and inquisitiveness. But he added, “Here’s the part you don’t see: Let’s assume they had ten ideas they thought were great. Let’s assume they applied six of them. That gauge of what you can apply and what you can’t is where Eric comes in big time. These guys decide this is what they want to do, and Eric will say, ‘This is worth fighting for. This is a really important thing. Let’s go do that. Let’s pull that, it will take us a little off track.’ What Eric has, and the founders are the first to say, is judgment, judgment, judgment. He knows when to take their initiatives and drive them to a conclusion, or to talk them out of it.”
SOMETIMES ENGINEERS CAN BE CLEAR about the wrong thing. By relying so heavily on algorithms and science, the Google founders—and Schmidt—have sometimes been clueless about right side of the brain issues, as they were with their original approach to Gmail, or book search, or their clumsy dealings with traditional media companies. Google collects an enormous amount of data about the people who use it. It asks users to trust them with private information, much as a credit card company asks users to trust it won’t share card numbers. The difference is that Google’s business model is based on selling advertising. And the data Google collects—the amount of time users spend with an ad or reading something, what they click on, what they search for, what they seem to like or dislike—is invaluable to the advertiser. Although Google does not hand over the data to an advertiser, it does use the data to help advertisers target customers. As Winograd points out, Google is really saying, “‘We’re smart guys. We have integrity. Trust us.’ They see things not from an institutional, political point of view but from this personal and engineering point of view: ‘We would never do that sort of thing.’ They believe that in their hearts.” Winograd believes them too. But the engineer’s passion, he said, drives them to also believe that they are “smart enough to make sure that it won’t happen by accident.” With the air of an empathetic but rigorous professor grading a smart but innocent student, Winograd arches a huge white eyebrow and concludes that this entails “a certain amount of technical arrogance—‘The system cannot fail, cannot fail.” But the system can fail, he added, because it is managed by fallible human beings, not machines.
Google, at least abstractly, is aware of this danger. Their IPO filing acknowledged that “privacy concerns” could sabotage the trust the company requires from users. In disclosing to investors the various ways in which Google could fail, they write: “Concerns about our collection, use or sharing of personal information or other privacy-related matters, even if unfounded, could damage our reputation and operating results. Recently, several groups have raised privacy concerns in connection with our Gmail free email service.... The concerns relate principally to the fact that Gmail uses computers to match advertisements to the content of a user’s e-mail message.”
If users lost trust in Google, believed their private data was being exploited and shared with advertisers (or governments), the company regularly judged one of the world’s most trusted brands would commit suicide. Do Google’s engineers, in their gut, believe they could lose the user trust they have earned? Unclear. What is clear is that there is often a fine line between certitude and hubris.
CHAPTER SEVEN
The New Evil Empire?
(2004-2005)
I
n Edgar Allan Poe’s story
The Purloined Letter,
an incriminating letter disappears from the private residence of the French queen. The Parisian police prefect takes on the case, but even after an extensive search, he cannot find the letter. And though he manages to narrow the search to a chief suspect, a government minister, he lacks evidence to arrest him. The prefect decides to consult the noted amateur detective C. Auguste Dupin. He explains that each night for three months, he has slipped into the minister’s home to assiduously search for the letter, removing cushions, the bottoms and tops of bedposts, the floorboards, the bindings of books—without success. The prefect is agitated; the suspect is a mere poet, he says, and he cannot believe such a “fool” could outwit him. Dupin, however, disagrees; he thinks the prefect and his detectives are the foolish ones, limited by their experiences, their routines, and “their own ideas of ingenuity.” They could not comprehend the acumen and cunning of a mind schooled not just as a poet but as a mathematician who follows his own “mathematical reasoning.”
Months go by, and the prefect returns, still unable to prove the minister’s guilt and ready to sign over the reward. Dupin, after persuading the prefect to sign a check, pulls the letter from his desk drawer. He explains that he cracked the case by climbing inside the supple mind of the suspect and imagining what he would do to conceal the letter. He imagined that the minister tricked the police by not attempting to conceal the letter. Rather, to avoid detection the letter was soiled, slightly torn, and crumpled in a card rack lying in plain sight in the middle of the minister’s room. Dupin found the letter where it had always been: under the nose of the prefect and his detectives.
Until 2004, most traditional media executives treated Google the way the prefect treated evidence: they failed to see the digital threat right under their noses. But soon after the IPO, their heightened awareness was captured in an eight-minute Flash-based movie that virally spread across the Internet. Called
Epic
2014, it was a faux documentary by two young journalists, Matt Thompson and Robin Sloan. With a voice-of-God narrator, it recounted how year by year a new media giant, Googlezon (the merged Google and Amazon), acquires or murders media companies, including the New York Times Company. By 2014, this Orwellian colossus employs its algorithms and computers to snare advertising and customize packages of news for individuals, whose wants are revealed by the cookies Googlezon gathers to track the behavior of its users.
Not surprisingly, this depiction jarred Googlers. When Sheryl Sandberg joined the company in late 2001, she believed she had a public mission, a mission parallel to the one she felt as a ranking member of the Clinton administration. Yet to her shock, not long after the IPO, she first heard Google referred to as the “evil empire.” She was attending a Google conference—“I was standing there with our partners and they said, ‘How do we sustain ourselves against the power of—’ I thought they were going to say Microsoft. Instead they said, ‘Google.’”
The hostility, said Eric Schmidt, “did not begin until Google went public and people realized how much money we were making.” The reaction had more to do with fear than envy. It took Microsoft fifteen years to exceed one billion dollars in revenues; it took Google just six years. The evidence was now visible that Google was attracting more Internet advertising than anyone else, and these dollars were being siphoned from traditional media. This was perceived as a threat to most traditional media companies, and perhaps none more so than the advertising industry. Google was able to sell advertising with just a few search words, and without charging the same 2 to 5 percent fee extracted by the media buying agencies. The buy was better targeted. And for advertisers it was more efficient, for Google only charged the advertiser when the consumer actually clicked on an ad. “There’s that same ‘think big’ attitude about markets and opportunities,” Steven I. Lurie, a former Microsoft executive who had friends at Google, told the
New York Times
at the time of the IPO. “Maybe you can call it arrogance, but there’s that same sense that they can do anything and get into any area and dominate.”
IT WAS IN THE CONTEXT of this growing backlash that the fight with book publishers, begun a few years earlier, started to come to a head. Like the Googlezon film, the uproar over digitizing books seemed to surprise Googlers. In their assessment, by scanning books and making them part of search, they were performing an ambitious and noble public service—they thought of the effort as their “moon shot”—and they assumed that they could do this without seeking permission of the copyright holders. Google knew that only about a third of the more than twenty million books ever published were no longer protected by copyright. But the mission was to scan
all
books. With books under copyright, Google said it would merely show “snippets,” which it claimed was permissible under the fair use clause of copyright law. Google did not precisely define the maximum number of words in a “snippet.” Nor does the law, but the rule of thumb is that fair use involves only enough text to briefly explain a book or briefly quote from an article or song.
Google believed it had provided protection to authors and publishers. In its contracts with libraries, Google said that if, within three years of the digital transfer of material, “Google decides not to use that content” (a particular book) because of a copyright dispute, the library would destroy the digital copy. They believed authors and publishers would see Google Books as a wonderful way to promote authors and their works, and to bring back books no longer in print. Google had earlier launched a Partners Program, signing up publishers who agreed to allow snippets to be shown for certain books, along with a link to an online bookseller. But publishers did not agree to allow all books to become part of search. The gulf between Google and the publishers and authors was vast. Google wanted to push the envelope of copyright, expanding the definition of fair use to allow more extensive quotations from books. It stressed the rights of search users, echoing the views of Web pioneers like Kevin Kelly, the “senior maverick” at
Wired
magazine, who said that in return for government copyright protection, authors and publishers had a “copyduty” to “allow that work to be searched.” Google was offering to pay the cost of moving and scanning the books; what publisher—or library or university or author—could refuse that offer?
One clue of Google’s fundamental attitude toward books—and fundamental innocence of the publishing process—is a conversation I had with Brin while reporting this book. It was the second of our three interviews and upon entering the small conference room down the hall from the second floor glassed office he shares with Page, Brin playfully ribbed me for writing this book. “People don’t buy books,” he said. “You might as well put it online.” He meant: You might as well publish it for free.
“You might make more money if you put it online,” he said. “More people will read it and get excited about it.”
There’s little evidence that such a free book succeeds, I said. Stephen King tried it, and gave up the effort because he thought it was doomed.
“I guess that’s true,” he acknowledged a little sheepishly.
Following Google’s business model, would he expect authors to generate their income by selling advertising in their books? If there was no advance from a publisher, who would pay to cover the writer’s travel expenses? (I made thirteen week-long roundtrips to Google from New York, rented a car, stayed at hotels, and paid for dinner interviews most nights.) With no publisher, who would edit and then copyedit the book, and how would they get paid for their work? Who would pay lawyers to vet it? Who would hire people to market the book so that all those potential online readers could discover it? The usually voluble Brin grew quiet, ready to change the subject.
But our rhetorical go-round hinted at something fundamentally true about Brin and Page and the dynamic company they have forged. Their starting predicate is that the old ways of traditional media are usually inefficient, and scream to be changed. This is a reason Google fundamentally misread the reaction of publishers and authors. While Google did reach various agreements with a variety of libraries, including Harvard, Stanford, the University of Michigan, Oxford University, the Library of Congress, and the New York Public Library, publishers did not like the idea of not getting paid for the use of their books. The Association of American Publishers denounced Google’s plan as an invitation to piracy, for the books stored on servers would be vulnerable to hackers. Publishers claimed they could be hit by the same thunderbolt that struck the music industry: free downloads.
Richard Sarnoff, the chairman of the Association of American Publishers and the executive vice president of Random House, said, “Google went to libraries and said we will digitize all your books and just use snippets of copyrighted books. They said it would be good for libraries and for users. This is true. But we have laws in this country which govern what we can do and not do. Like copyright, which prevents people from copying things for their own commercial use. And this is for Google’s commercial use, for search.” The publishers demanded that Google seek their permission before digitizing any book that was still protected by copyright. “The Internet is a grazing medium,” Sarnoff said. “Books tend to be a longer term experience.” Grazing can be a great way to promote a book, he said. “But we want to be extremely careful to make sure discovery does not become consumption.” To illustrate his fear of piracy, he pulled out his iPhone and said that the small device can hold fifty thousand books, all easily down-loadable. This, he noted, is the approximate capacity of a midsize bookstore. In October 2005, the publishers announced that they had filed a lawsuit.