Alissa Lee encountered this upside-down approach. In the first five years of Google’s life, one or both founders insisted on interviewing each applicant. Brin was introduced to a Harvard Law graduate, Alissa Lee, by David Drummond, who was the company’s outside counsel in the late nineties and became Google’s corporate counsel in 2002. Lee was a contracts lawyer, and in the course of her interview, Drummond remembers, Brin said, “‘I really need to see how you will practice law. I need you to draw me a contract. Don’t spend a lot of time on it. Draft it and send it to me and to David so we can review your work.’” And then came the Google test: “‘I need the contract to be for me to sell my soul to the devil.’” Brin remembered his request and recalled: “I just figured that if I’m interviewing an attorney I should validate their work product.”

Lee remembered repeating the question, not sure she had heard it correctly. Brin told her he wanted the contract e-mailed to him in the next thirty minutes. “Amid the surreal oddity of it all,” she recalled, “I had forgotten to ask him all sorts of lawyerly questions, like what sort of protections he needed, what conditions he wanted to attach, and what he wanted in return for his soul. But then I realized that I had missed the point. He was looking for someone who could embrace a curveball, even relish it, and thrive in the process of tackling something unexpected. I’m not sure he actually looked at what I sent him, but something in my crazy sale agreement or in my response must have satisfied him.”

“She was a clear hire,” said Drummond. Today Lee is Google’s associate general counsel.

John Doerr encountered a similar upside-down approach when Page asked him what he thought of Google’s buying a Boeing 767.

“I think that’s a terrible idea,” said Doerr.

“Why?”

“For the ethos and egalitarian nature you want to have in the company,” Doerr answered, “you’re never going to get away from the public perception of two Silicon Valley entrepreneurs owning a personal 767.”

“Look at the numbers,” Page said, showing Doerr a sheet of paper revealing that for seven million dollars they could purchase the 767, and for another ten million dollars they could install improved engines and a new interior. “A totally upgraded 767,” Doerr realized, “cost less than a G-5.” And it could fly longer distances and accomodate thirty-five people, transporting engineers and the founders or Schmidt around the world to visit Google sites. “They went ahead and did it.” They later purchased an additional plane, a Boeing 757.

WITH INTENSE PRODDING from the founders, the Google engineering shop was innovating at a furious pace. Among the new projects developed at this time were Desktop Search, Froogle (later changed to Google Product Search), Google Maps, Google Print (renamed Google Book Search), Google Docs, to allow users to create and edit documents, presentations, and spreadsheets, and Pyra Labs, a site to facilitate the creation of blogs. The founders were particularly enthusiastic about the idea for a new e-mail system. Unlike providers such as AOL, Google’s e-mail would be free, and would allow its users to easily search their own e-mail archives for content and contact names. And while Yahoo’s free e-mail offered its users 4 megabytes of storage, Google’s Gmail would provide 1 gigabyte, 250 times as much. To the engineers, it seemed clear that this was enough storage that a user would never have to delete e-mails. In the interest of efficiency, the first version of Gmail did not include a delete button. This had an unforeseen effect: Users feared that Google would peek at e-mails. And Paul Buchheit’s e-mail scanning software-the same program that had grown into AdSense-only fanned this fear. For Google, it was a way to make money from e-mail by placing ads when certain keywords were typed. But critics said it was an invasion of privacy, that Big Brother was watching everything. Google’s engineers failed to absorb the lesson of Microsoft’s Passport program. Introduced in 1999, the program stored personal information and allowed access via a log-in name and password. Its release triggered a storm of protests, complaints that Microsoft would access this personal data for its own business reasons. Perhaps a reason Google failed to absorb the lessons of Passport was because Google believed its intentions were noble and that Microsoft’s were probably not.

Terry Winograd, the Stanford professor, was a consultant on Gmail and described a “huge debate” over the program. “We said, ‘People want to delete things. There should be a delete.’ Larry, among others, said, ‘We want them to start thinking differently.’” Page said that because Google was offering so much storage, users could keep everything, and went on to argue: “If you delete stuff, you might later on decide you want it. Plus, you spend time thinking about whether I should delete this or not.” The “engineering optimization side,” said Winograd, claimed this was an inefficient use of users’ time.

The Electronic Privacy Information Center, a public interest research center in Washington that focuses on privacy and civil liberties issues, demanded that Gmail be shut down, declaring that it was “an unprecedented invasion into the sanctity of private communications.” Of course, a computer, not a human, was scanning the e-mail, as most e-mail providers do to prevent spam. At first, the founders and Schmidt tried to defend the no-delete button and the advertising feature of Gmail, believing the small tempest would pass. It did not, and Google was forced to add a delete button. For Winograd this was an early sign of troubles to come. He has enormous respect for his former students (and gratitude for the Google stock grants that made him a rich man), but what he saw in the Gmail debate was that Google relied so much on science, on data and mathematical algorithms, that it was insensitive to legitimate privacy fears-and, later, to fears they would dominate the search market. Winograd describes his two former students as impatient: “Larry and Sergey believe that if you try to get everybody on board, it will prevent things from happening. If you just do it, others will come around to realize they were attached to old ways that were not as good.” The attitude, he said, “is a form of arrogance: ‘We know better.’ The idea that somebody at Google could know better than the consumer what’s good for the consumer is not forbidden.”

Only die-hard Google bashers, however, would deny the idealism that drives many of the decisions Brin and Page have made at Google. In the wake of 9/11, Krishna Bharat, an Indian-born engineer who joined Google in 1999 and today has the title principal scientist, was moved by the awful events of that day to ponder its lessons. One lesson, he believed, is that Americans were largely ignorant of other peoples and creeds, including radical Islam. There was too little international news in print or on television. Bharat said he wanted to “broaden horizons, allowing people to see other perspectives, to see what the Arab street is saying today. It is hard for the New York Times to do justice to that.” Devoting his 20 percent time to this project, Bharat devised a program that would be known as Google News, initially offering free access to almost five thousand worldwide news links. The placement and selection of stories is made, Google announced, by “computer algorithms, without human intervention.” Google News would be ad-free, meaning Google would lose money from this effort. Like digitized books, Google News was advanced as a promotional and sales vehicle. It would, Google said, broaden newspaper readership, and allow newspapers to sell advertising once a user clicked on the newspaper’s link. “We send traffic to newspapers,” Bharat said. However, newspapers didn’t all jump up and down with glee.