Crystal Lu Jan 12 2007 Features
By Crystal Lu
The clock is just turning 12 at Google. On the verdant green lawn outside the company’s biggest cafeteria, early birds are already having their lunch at white tables and chairs. A pizza stand near the entrance to the cafeteria displays steamy pies just out of the oven. A few steps away is an Asian section serving stewed chicken thighs and chow-mein. There are also an Italian section, a Mexican section, an Indian section, a make-your-own-sandwich bar, a salad bar, a soda machine, a transparent-door refrigerator full of Odwalla juices and a counter for making coffee and tea. Everything is free. It’s part of the Google legend to break the there-is-no-free-lunch convention and offer all employees three complimentary meals a day. There are 10 cafeterias on the Google “campus”.
They don’t call it a campus for nothing. Most of the Googlers carrying their trays could easily pass for graduate students. Almost everyone is in jeans, except for a few men in baggy, knee-length shorts and a few women in denim skirts or casual-looking floral skirts.
One man stands apart. He’s wearing a pale blue button-down shirt and a black pair of slacks. He’s got a schoolboy hairstyle---slightly longer than a crew-cut---that makes his whitish blond hair appear more sun-bleached than aging gray. He’s clearly older than the others, but his smooth face and the curious, inquisitive look in his large blue eyes make him blend in with a youthful spirit.
Mike Burrows is 43 years old, and although he’s been in the United States for the past 18 years he still talks with the crisp accent of his native England. Mike explains he owns more than a dozen blue shirts and five pairs of identical trousers, all of them made out of washable, non-iron fabric. He dresses the same every day. “I strongly value function over form,” he says.
Mike is having a salad made of shredded carrots, bite-sized pineapple, and raisins. The “function” of the salad is to give him more nutrients and less fat than pizza or French fries would.
One of the highest-ranking computer scientists at Google, Mike works on Google’s distributed system, a non-centralized network consisting of numerous computers that can communicate with one another and that appear to users as parts of a single storehouse of shared hardware, software, and data. In the circle of high-tech experts, Mike has a special place for his role in inventing Alta Vista, the first multilingual search engine. According to The Google Story, a Google-authorized bestseller, Larry Page at the beginning stage of creating the Google search engine borrowed Alta Vista’s way of displaying links that could instantly take users to other websites.
Mike is not just another hip, smart young Googler. He’s one of the pioneers of the information age. His invention of Alta Vista helped open up an entire new route for the information highway that is still far from fully explored. His work history, intertwined with the development of the high-tech industry over the past two decades, is distinctly a tale of scientific genius.
The story of the Internet begins six years before Mike was born---with Sputnik.
When the Soviet Union launched the first artificial earth satellite in 1957, the United States responded by forming the Advanced Research Projects Agency within the Department of Defense. To build a military research network that could survive a nuclear strike, ARPA contracted a technology-consulting firm, Bolt, Beranek, & Newman, to work on ARPANET, which linked computers at UCLA, Stanford Research Institute, UC Santa Barbara and the University of Utah. The system was completed in 1969. Three years later, Ray Tomlinson of BBN created the first e-mail program. In 1974, Vint Cerf and Bob Kahn first coined the term “Internet” in their paper on Transmission Control Protocol. Nine years after that, the Internet Activities Board was established. In 1990, a British scientist Tim Berners-Lee designed a virtual infrastructure called the World Wide Web to transmit hypertext documents internationally, and a program called a “browser” to view them. A global information space was forming. But the general public was still unaware of the Internet.
“Scientists were using the Internet long before it became popular,” says Mike, “The public just didn’t seem interested until somebody put some pretty pictures on line.”
Marc Andreesen was the one who first displayed pictures on the Internet. He was a student and part-time assistant at the National Center for Supercomputing Applications at the University of Illinois in 1992, when Internet browsers were not user-friendly and only used by academics and engineers who had access to certain expensive machines named Unix. Andreesen decided to develop a browser that was easier to use and graphically richer. He recruited a fellow NCSA employee, Eric Bina, to work with him on the project. In 1993, they created a new browser, Mosaic, which allowed pictures and text to appear on the same web page. The original version of Mosaic was designed for Unix, but PC and Mac versions soon followed. Mosaic’s success made the front page of the New York Times business section in December 1993. Then Andreesen graduated and moved to Silicon Valley, where he started a company in 1994 to produce an improved version of Mosaic, which became a household name: Netscape.
The popularization of the Internet generated a huge demand for new tools of navigation to help users find information. A new search engine appeared almost every week. WebCrawler, Lycos, Magellan, Infoseek, Excite, HotBot, and Inktomi all rose and fell. Each seemed like a small advance but each had similar problems: the results they delivered were often completely irrelevant. Then along came Yahoo!, which succeeded in bringing more relevant web pages to users. But strictly speaking Yahoo! wasn’t really a search engine because it didn’t operate automatically. The company employed a team of editors to select websites for an alphabetized directory. The manual work it required prevented it from keeping pace with the fast growth of the Internet.
Something new and innovative was needed. And then a little-known computer scientist at Digital Equipment Corporation came up with a bright idea.
When Mike Burrows walks, he strides. His two arms take turns moving forward as high as those of a marching soldier. At 5’8”, he doesn’t have particularly long legs, but each step looks almost twice the length of an average one.
Mike walks about 45 minutes from his house in Palo Alto to his Google office in Mountain View every weekday morning and another 45 minutes home in the evening. It provides the best combination of exercise, economy and environmental concern. “Biking could be some exercise, too, but not as much as walking,” he says.
In the mild morning sun, through quiet residential neighborhoods, along shady avenues, Mike walks in his everyday pale blue shirt, black trousers and black soft-leather shoes. The walk is as functional to him as his clothing, the style of which came from Manchester Grammar School.
Mike first attended the grammar school at age 11. Since then he has kept the same haircut and the same kind of clothes as his grammar school uniform, except for losing the striped tie.
Mike knew he wanted to be a scientist early on and focused on it at school. “I strongly prefer the British educational system,” he says. “We begin to specialize at age 14. That means after age 14 I didn’t have to study history, geography or any social sciences I was bad at. I was able to focus on the subjects I liked. Early specialization gives you more time to concentrate and makes it easier for you to succeed.”
To confirm his inclination toward science, Mike has repeatedly taken personality tests derived from Myers-Briggs Type Indicator, which was developed by Isabel Briggs Myers and her mother Katherine Briggs and introduced in 1943. The test categorizes personalities in four areas: 1) energy source---either introversion or extroversion; 2) perceiving mental process---either sensing, which requires tangible information, or intuition, which leans toward imaginative possibilities; 3) judging mental process---either thinking or feeling; 4) outside world orientation----a preference for judging or perceiving. According to the test results, Mike is an extreme case of introversion, sensing, thinking and judging.
At the MBTI website, there is a short description of this type of person, “Quiet, serious, earning success by thoroughness and dependability. Practical, matter-of-fact, realistic and responsible. Decide logically what should be done and work toward it steadily, regardless of distractions.”
This description suits Mike fine. “I don’t see why anyone would want to be any other way,” he says. “Why would you prefer groundless feeling over logical thinking, or perceiving over judging?”
Mike makes all his judgments analytically. He chooses not to become an American citizen mostly because jury duty would be too time-consuming for his busy work schedule. Besides, he argues, his vote wouldn’t make a difference in the liberal state of California because he agrees with the majority view. There are no emotional reasons behind this decision, he insists, no sense of patriotism for the Britain he left behind. “I don’t understand why people get so excited about their countries,” he says.
Mike makes fun of the Olympics. He doesn’t see how an athlete’s medal becomes a nation’s glory. One person’s outstanding performance in a certain sport doesn’t mean all his or her compatriots are good at it. How can they be proud of that medal?
While most people tend to emotionally associate themselves with their countries, hometowns and schools, Mike doesn’t take pride in any of these, including his world-class alma mater Cambridge. He chose not to go to Cambridge for his undergraduate studies because it would have required him to take courses in general engineering. He only wanted to study computer engineering, so he selected a more focused program at London University. But after that, he went to Cambridge for a Ph.D.
“Before going to Cambridge, I had an interview for a job that looked quite desirable.” Mike recalls. “But I argued about the company’s technology with every single interviewer on that panel. So, I ended up at Cambridge.”
In 1987, Mike’s Ph.D program chair Roger Neden recommended him to be a summer intern with Digital Equipment Corporation, a pioneering American hardware company. It produced the most popular mini-computers---huge by today’s standards---for the scientific and engineering communities in the 1970s and the 1980s before the advent of desktops. The internship brought Mike to Silicon Valley for the first time. Unlike other interns who tended to be cautious and quiet, Mike spoke up at staff meetings and brought up new ideas that impressed his colleagues. DEC offered him a full-time position before he finished his Ph.D studies at Cambridge. After his graduation in 1988, he came to Silicon Valley again to officially work for DEC.
Alta Vista was initially a project Mike worked on at DEC, which allowed researchers to initiate their own projects. Mike had been interested in search since his Cambridge years. He had made an index system that could search all kinds of information in his own e-mail account. But to search on the Internet would require a far larger scale. It seemed almost too daunting to accomplish. Mike struggled. Then one day he met a fellow computer engineer who seemed to have some of the answers he was looking for.
In a conference room on the second floor of Google’s Building 43, Louis Monier sits down on a red sofa away from the conference table, a bottle of mixed fruit juice in his right hand. He takes a slow sip of his drink.
A tall, broad-shouldered handsome man with salt-and-pepper hair and beard, Louis Monier is surprisingly soft-spoken for his robust frame. His mild French accent comes from his native Provence. The silky teal shirt and off-white khakis he’s wearing also appear to be more European than American, although he has lived in the United States since he first came for a post-doc fellowship at Carnegie Mellon University in Pittsburgh in 1980.
Louis recalls that his collaboration with Mike began in spring 1995, when DEC was losing business to Sun Microsystems. DEC’s researchers were wondering what they could do to turn things around. Since the Internet was on the rise, someone suggested using DEC’s big machines to do something on the web. Louis and Mike happened to be both interested in search. Although they belonged to different labs at DEC, they teamed up to start building a search engine.
A search engine consists of three main parts: a crawler that collects web pages, a back end that processes the pages behind the scene, and a front end that interacts with users. The crawler Louis built was an innovation because it could cover 1,000 times as many web pages as what had been done before. The user interface in the front end was also Louis’s creation. He opted for simplicity in a very similar way to what Google did later and gave access to as many people as possible. Meanwhile, Mike was making an index in the back end. It was painstakingly difficult, given the index size and the volume of queries. But Mike overcame all the difficulties with his intelligence, persistence and diligence.
“When we were working together, Mike really impressed me as a very smart, very technical, incredibly hard worker, not interested one bit in getting credit for what he was doing,” says Louis, “Also with incredibly high standards, especially for himself. Mike is never happy with what he does”
For months, Louis and Mike worked day and night. They finally completed their new search engine at the end of 1995 and named it Alta Vista, meaning “high view” in Spanish.
“It was our Christmas present to the Internet.” Louis says, smiling.
When the press came to inquire, Louis did the front-end work again. He was always the one who talked to reporters.
“Mike never wanted any exposure,” recalls Louis. “This is probably the first time ever that he lets someone interview him. He never did when Alta Vista was in the news.”
The two men’s invention attracted thousands of users every day. Louis and Mike worked very hard to direct the heavy traffic on Alta Vista. Mike put in at least 14 hours a day. Then his health suddenly deteriorated. He had contracted Chronic Inflammatory Demyelinating Polyneuropathy, a disease that caused his immune system to destroy his own peripheral nerve cells. Fortunately his condition was controllable. He continued working.
While Alta Vista was a free service that didn’t generate any revenue, it brought more positive press than DEC had received in years. Management was thrilled but didn’t really know how to make the best out of the unexpected success. The company did little more than use Alta Vista on demonstrations of new machines and create a software division to oversee the search engine.
It seemed to both Mike and Louis that the people in charge of the software division were not as enthusiastic about search as about developing other software products, none of which turned out to succeed. Mike blames them for hiring a “marketing idiot” who posted flashy ads that users disliked on the Alta Vista website and suggested adding an orange stripe across the website to make it eye-catching. After the orange-stripe idea was rejected, the marketing side even sent out spam to advertise Alta Vista.
Upset with the managers and marketers of the software division, Mike stopped working on Alta Vista in 1997 and moved on to other projects. He completely let go of his groundbreaking invention, from which he never gained publicity, nor did he ever care to take any credit. He simply wanted to continue pursuing his passion in life, hi-tech research.
Mike left behind the most efficient search engine of the pre-Google era. Before the so-called Google guys, Larry Page and Sergey Brin, decided to launch a new enterprise, they once hoped to sell their newly invented technology in order to continue their Ph.D studies at Stanford University. They targeted Alta Vista, which had 54 percent market share of search as of March 1998, and talked to Paul Flaherty, a Stanford Ph.D researcher working with Alta Vista at the time. Flaherty liked the new invention, which was superior in ranking the importance of web pages, but it wasn’t his decision to make. Alta Vista belonged to DEC, which was undergoing a merger with Compaq and had no desire to buy any outside technology.
During the pricing process of the buyout, Compaq put a zero price tag on Alta Vista. After buying DEC in 1998, Compaq didn’t take a close look at Alta Vista until a year later, when several companies attempted to buy the search engine. AOL offered $750 million. Then Compaq, realizing it had a very valuable asset, decided to keep and make better use of Alta Vista. But management could only think of following the Yahoo! model by creating multiple functions and categories at the website. To make a second Yahoo!, dozens of Compaq’s Houston employees relocated to Silicon Valley. But they were all hardware engineers with no knowledge of search technology. Louis felt terribly disappointed. He resigned in April 1999 to start his own e-commerce business.
In August 1999, CMGI, an Internet investment firm, acquired majority ownership of Alta Vista from Compaq in a $2.3 billion stock swap. CMGI allowed Alta Vista to spin off. Alta Vista planned to go pubic the next year, but the stock market crashed right before the scheduled date. The dot.com bubble burst. Alta Vista remained private. In April 2003, Overture bought Alta Vista. A few months later, when Yahoo! acquired Overture, Alta Vista also became part of Yahoo!.
Now Alta Vista is only a web page that keeps its old name to retain faithful old users. Those who click on Alta Vista are actually going into Yahoo!.
Louis looks back on the entire process with some regret.
“Alta Vista could’ve done better, but it didn’t get good management from the start,” says Louis, “Then Google came out with business experts in the management. Technologically Google was able to remove spam from search results. Alta Vista couldn’t do that. But the two search engines were pretty close. Alta Vista fell far behind later only because it didn’t get enough management support to keep improving itself when Google did. I wish things could’ve been different with Alta Vista, but what’s done is done. You can only move on.”
Louis has redirected his interest in search into his current job at Google, where he started to work in 2005. He is happy with the way Google keeps pursuing innovation to remain on the cutting edge. He predicts that Google will continue to be number one in search.
It is dark when Mike walks out of Google’s Building 43. The temperature is in the low 40s, quite chilly for California. Mike has his black jacket on.
There are many cars in the well-lit parking lot. Mike is going to none of them. He is walking home, where he lives alone.
It has been Mike’s habit to leave work late ever since his DEC years. He became a Compaq employee upon Compaq’s acquisition of DEC. After CMGI bought Alta Vista from Compaq and let it spin off, Mike turned down the invitation to work for his own invention. He knew Alta Vista was financially unable to hire top-notch researchers, and he feared he would just be maintaining the existing product, nothing innovative. He preferred staying with Compaq to take on new projects. To him, work has to be interesting. He wants to work with top computer scientists, to keep collaboratively generating brightest ideas. He doesn’t have to be the head. He would decline an executive position for all the management duties it entails, which he deems as distractions from his research.
After HP bought Compaq in 2001, Mike decided to leave, because HP assigned projects from the top rather than let researchers brainstorm new ideas. To Mike, this was not the proper way to do research. “Alta Vista would’ve never come out of an environment like that,” he says.
Before the final merger, Mike jumped ship to Microsoft, where some of his former Compaq colleagues had gone. He worked on preventing spam for Microsoft. He came up with an idea to add 10 more seconds to the time for each email to get through. Since Internet servers such as AOL and Earthlink charged users for their time on line, if the time of delivering each email increased, it would be costly to spammers sending a large number of emails simultaneously and would discourage them from doing so. But 10 more seconds wouldn’t bother a normal user sending only a few emails at a time.
While Mike dedicated himself to the spam prevention project, one of his colleagues refused to cooperate. But she started another project also about computation time. Her intention to compete irritated Mike. He believed there shouldn’t be competition among colleagues in an industry research lab. Unlike university labs, where tenured positions are few and the need to compete is intense, an industry lab offers every researcher a long-term stable job. It should be all about cooperation.
Mike didn’t get to participate in much teamwork at Microsoft. Most of his colleagues in the Silicon Valley R&D lab were older, had families and went home at 5 pm. Mike found the almost empty lab in the evening “very boring”. He also disliked taking frequent trips to Microsoft’s headquarters in Redmond, Washington.
In 2003, Mike left Microsoft for Google. What attracted him most was the creative Googlers. But he had a concern about Google’s medical insurance coverage. Mike had never fully recovered from Chronic Inflammatory Demyelinating Polyneuropathy. For him, Google made a special case in health insurance.
Mike goes for a regular treatment called Intravenous Infusion of Immunoglobulins at Stanford Hospital. The cost: $10,000 per visit.
A needle inserts the inner side of Mike’s left forearm. It gets into a vein at the first try. Both Mike and Stacey, his nurse, breathe out with relief. Stacey removes the syringe from the needle, which she instead connects with a thin plastic tube that leads up to an IV bottle hanging from a wheeled tripod.
When Stacey walks away to take care of other business, Mike remains seated in a brown leather armchair, a small silver laptop on his lap. He is prepared for the next six hours of treatment. In the Ambulatory Treatment Infusion Center, Mike can walk around with the IV on the tripod following him. But he stays in his seat most of the time, doing a little work, reading the news or watching a DVD on his laptop.
The solution in the IV bottle is called Immunoglobulins, which is a plasma product formed by mixing antibodies from about 20,000 donors. Since 1985, all of the plasma used for IVIG has been tested for HIV and hepatitis. The complex production of IVIG makes the treatment highly costly.
Chronic Inflammatory Demyelinating Polyneuropathy is an autoimmune disorder that damages the fatty covering that wraps around and protects nerve fibers. Symptoms include tingling or numbness that begins in the toes and fingers, weakness of the arms and legs, loss of deep tendon reflexes, fatigue and abnormal sensations. The course of CIDP varies from individual to individual. Some may have a bout of CIDP and then completely recover. Others may have many bouts with partial recovery, as Mike does.
It is unclear what exactly causes CIDP or other autoimmune disorders but very clear what it does to the body, and regular treatment is necessary. Administration of IVIG depends on the patient’s conditions. Mike needs it about every two weeks. When he feels better, he may wait a little longer to get the next treatment, but no more than three weeks.
Mike usually comes in for his treatment on a Saturday or Sunday. He doesn’t want to miss a day of work, despite the fact that he actually doesn’t need to work. Given his frugal lifestyle and all the high-paid jobs he has had, he has saved up more than enough money to afford an early retirement. He doesn’t have to work for a living.
Mike works for fun. This may sound incredible, but he does.
At a snack stand in Google’s Building 43, Mike grabs an energy bar and a bottle of orange juice. This is his breakfast before a long day of work.
A little sweaty from his 45-minute morning walk, Mike takes his breakfast to his office. On his long desk are two computers, each of which has two connected mirror-like frames of monitors in the position of an open book.
Throughout his career, Mike has worked on all types of computers, from the giants of the 1980s to the latest models like these desktops with double monitors as thin as wafers. Mike has witnessed the light-speed advancement of computer technology in the past two decades as closely as anyone could. He’s happy about the tremendous progress, but he also feels a small sense of loss.
“When computers are more powerful and more popular, more computer engineers are needed, but they need to be less skillful because the computers don’t have a lot of problems,” says Mike. “The work is not as challenging as it used to be, so it could be less fun.”
Mike is determined to find new fun, however. He has always come up with new ideas and new inventions. According to other Googlers, who find Mike a little eccentric but amazingly brilliant, he owns more patents than anyone they know. Although Mike never talks about his patents, word has somehow spread.
Normally it’s considered quite an achievement for a scientist to accumulate a dozen patents through years of hard work. Mike has 50.
Mike says he takes no pride in this astonishing number. He applied for the patents only because his employers required him to. He is pleased with the financial rewards his patents have brought him. But he doesn’t think he deserves all the credit for them.
“Inventors always need to use their predecessors’ work,” says Mike. “Nobody can invent something out of nothing. For example, how can you name who invented the Internet? It was the accumulation of so many people’s contribution.”
Even the most creative inventors owe debts to others. Mike says when he built the back end of Alta Vista, he copied the syntax of query from Infoseek, an earlier search engine.
“All search engines borrow from each other,” he says.
In Mike’s opinion, it was natural that Google in its initial stage adopt Alta Vista’s method of displaying links. He doesn’t mind the fact that Google surpassed Alta Vista. In the river of technology, new big currents are bound to push old ones down. This is the only way for the river to flow ahead.
Mike doesn’t see the decline of Alta Vista as his loss because he doesn’t think of the search engine as his possession. In his view, technology is all about collective efforts and collective sharing. Alta Vista had its moment. When it came out as an unprecedented success, Mike felt a great sense of achievement. For him that was enough.
Perhaps it was the same thrill that kept Isaac Newton sitting under an apple tree. Perhaps all great scientists are closer to being children than the rest of us. They haven’t lost the wild imagination and persistent curiosity of childhood.
Mike is serious about his work, but in some ways he is another Peter Pan who refuses to enter adults’ world, where he doesn’t belong. Wealth, fame and power mean nothing to him. He simply wants to create new technology, in his own way.
“What else matters? In a million years, none of us will be here. You can’t take anything away with you. It only matters what you enjoy while you are here, what you get a kick out of.”
His large blue eyes sparkle as mischievously as those of a schoolboy.
Crystal Lu is studying for a Masters in Education Policy