Brin went to Stanford on a National Science Foundation scholarship, where he decided to focus on the study of data mining. (In a double whammy, to itself if not to them, MIT rejected him as well as Page.) There were eight comprehensive tests he needed to pass for his doctorate, and he aced seven of them soon after he arrived. “The one I thought I was best at, I didn’t pass,” he recalled. “I went to the prof and debated the answers. I wound up talking him into it. So I passed all eight.”133 That left him free to dabble in whatever courses he wanted and to indulge his quirky athletic interests in acrobatics, trapeze, sailing, gymnastics, and swimming. He could walk on his hands and, so he claimed, once considered running away and joining the circus. He was also an avid rollerblader, and was often seen zipping through the halls.
A few weeks after Page arrived at Stanford, he and Brin moved with the rest of the Computer Science Department into the new Gates Computer Science Building.VIII Annoyed by the uninspired numbering system for offices that the architect had provided, Brin devised a new system, which was adopted, that conveyed better the location of each room and the distance between them. “It was very intuitive, if I may say so,” he said.134 Page was assigned to a room with three other graduate students, and Brin made that his base as well. There were hanging plants with a computer-controlled watering system, a piano connected to a computer, an assortment of electronic toys, and sleeping pads for naps and all-nighters.
The inseparable duo became linked, in CamelCase fashion, as LarryAndSergey, and when engaged in argument or banter they were like two swords sharpening each other. Tamara Munzner, the only woman in the group, had a phrase for it: “goofy smart,” she called them, especially when they took to debating absurd concepts, such as whether it was possible to construct something the size of a building using only lima beans. “They were fun guys to share an office with,” she said. “We all kept crazy hours. I remember once at three in the morning on a Saturday night, the office was full.”135 The duo was notable not only for their brilliance but for their boldness. “They didn’t have this false respect for authority,” according to Professor Rajeev Motwani, one of their advisors. “They were challenging me all the time. They had no compunction in saying to me, ‘You’re full of crap!’ ”136
Like many great partners in innovation, LarryAndSergey had complementary personalities. Page was not a social animal; he could make eye contact with a screen more easily than with a stranger. A chronic vocal cord problem, stemming from a viral infection, meant that he spoke in a whispery and raspy voice, and he had a disconcerting (although in many ways admirable) habit of simply not talking at times, which made his utterances, when they did occur, all the more memorable. He could be impressively detached but was sometimes intensely engaging. His smile was quick and real, his face expressive, and he listened with a focus that could be both flattering and unnerving. Intellectually rigorous, he could find logical flaws in the most mundane comments and effortlessly steer a shallow conversation into a deep discussion.
Brin, for his part, could be charmingly brash. He would barge into offices without knocking, blurt out ideas and requests, and engage on any subject. Page was more reflective and reserved. Whereas Brin was satisfied knowing that something worked, Page would ruminate about why it worked. The intense and talkative Brin dominated a room, but Page’s quiet comments at the end of a discussion made people lean forward and listen. “I was probably a little bit more shy than Sergey, although he’s shy in some ways,” Page observed. “We had a great partnership, because I maybe thought more broadly and had different skills. I’m trained as a computer engineer. I’m more knowledgeable about the hardware. He has more of a mathematical background.”137
What particularly amazed Page was how smart Brin was. “I mean, he was just unusually smart, even for somebody at the Computer Science Department.” In addition, Brin’s outgoing personality helped him bring people together. When Page arrived at Stanford, he was given a desk in an open room known as the bullpen with the other new graduate students. “Sergey was pretty social,” Page said. “He would meet all the students and come hang out in the bullpen with us.” Brin even had a knack for befriending the professors. “Sergey had this way of walking into professors’ offices and hanging out with them, which was kind of unusual for a grad student. I think they tolerated it because he was so smart and knowledgeable. He could contribute on all sorts of random things.”138
Page joined the Human-Computer Interaction Group, which explored ways to enhance the symbiosis between humans and machines. It was the field that had been pioneered by Licklider and Engelbart, and it had been the subject of his favorite course at Michigan. He became an adherent of the concept of user-centered design, which insisted that software and computer interfaces must be intuitive and that the user was always right. He had gone to Stanford knowing that he wanted as his advisor Terry Winograd, a joyful, Einstein-haired professor. Winograd had studied artificial intelligence but, after reflecting on the essence of human cognition, changed his focus, as Engelbart had, to how machines could augment and amplify (rather than replicate and replace) human thinking. “I shifted my view away from what would be thought of as artificial intelligence to the broader question, ‘How do you want to interact with a computer?’ ” Winograd explained.139
The field of human-computer interactions and interface design, despite its noble heritage from Licklider, was still considered a rather soft discipline, looked down upon by hardnosed computer scientists as something usually taught by mere psychology professors, which Licklider and Judith Olson had once been. “For people studying Turing machines or whatever, dealing with human responses was considered very touchy-feely, almost like you’re stuck in the humanities,” according to Page. Winograd helped make the field more reputable. “Terry had a hard computer science background from his time working on artificial intelligence, but he was also interested in human-computer interaction, a field that nobody much was working on and I think didn’t get enough respect.” One of Page’s favorite courses was Film Craft in User Interface Design. “It showed how the language and techniques of film can actually be applied to computer interface designs,” he said.140
Brin’s academic focus was on data mining. With Professor Motwani, he started a group called Mining Data at Stanford, or MIDAS. Among the papers they produced (along with another graduate student, Craig Silverstein, who would become the first hire when they founded Google) were two on market basket analysis, a technique that assesses to what extent a consumer who buys items A and B is more or less likely also to buy items C and D.141 From that Brin became interested in ways to analyze patterns from the data trove on the Web.
With Winograd’s help, Page began casting around for a dissertation topic. He considered close to a dozen ideas, including one on how to design self-driving cars, as Google would later do. Eventually he homed in on studying how to assess the relative importance of different sites on the Web. His method came from growing up in an academic environment. One criterion that determines the value of a scholarly paper is how many other researchers cite it in their notes and bibliography. By the same theory, one way to determine the value of a Web page was to look at how many other Web pages linked to it.