Krishna
This is a book on Google and how it functions. Should be an immensely interesting book.
But the tone set, at least in the initial pages makes the book feel like hagiography. (Remember our review of Beyond the Last Blue Mountain? I need not have worried.
Now, there is nothing wrong about praising high achievement, and Google definitely deserves praise for changing the way people think of and interact with computers. The problem with one sided review is that for readers like me, it does not give a balanced view. “Google can do no wrong; Google is the best there is” kind of books do not give even a whiff of any missteps it may have taken and therefore does not give an unbiased look. Maybe I am misjudging the book too soon. Let us see how it pans out.
Luckily, the author pulls out of the adulation and moves on to the world before Google’s PageRank algorithm changed the search results forever. He talks of the previous reigning monarch AltaVista from DEC, which was used by most folks before Google turned up.
It is interesting to note that neither Brin nor Page (in Stanford, but a year apart) had no interest in starting a company but wanted to go do their PhD. However no company was willing to buy their PageRank (or BackRub) algorithm including the then reigning king AltaVista in search and Excite, another search company. So they were forced to start the company.
Despite their professed interest in research, they were reluctant to even write a paper on the new search technology they developed!
They start small and the growth is high. But given the cheap nature of their infrastructure, they have multiple problems; Python is not the best for the volumes handled and the data architecture is fiddly. All of these are rewritten by bright folks who joined early.
The book gets more interesting. The Google first was famous for search not because of its incredible speed but because of Larry and Sergey’s insights to use cross links as a fundamental marker for relevance of the results. The speed came later – through others – using in memory indexing, compression and rewrite of the search multiple times, and also recognizing that proper names had to be dealt with differently.
Until you read the details in this book, you do not realize how difficult answering a search query can be. How do you rank pictures that come with web pages? How do you determine the relative importance, all the while learning from failures of actual queries? In addition to the bright engineers who flocked to the company almost from inception, you also learn the problems, challenges and fixes they had to constantly make.
There are unexpected faux pas created by Ad Sense which “analyzed the content and placed appropriate ads in the pages” – an expansion of Google into other websites for advertisement purposes only. (Ad Sense was created by another company and bought up by Google which then proceeded to recode the entire Ad Sense. I guess the takeover was for IP rights). Examples : When there was a gruesome report in an online news site about bodies chopped up into plastic bags, Google’s ad showed ‘Plastic Bags’. An airline disaster page may advertise airline tickets; An article on Mumbai attacks titled ‘Terrorists kill the man who gave them water’ was accompanied by ‘Terrorism : Pursue a certificate in terrorism 100% online . Enroll Today. Ads by Google”. (Not kidding). And one more : A report about massive food poisoning at Olive Garden in Los Angeles featured a coupon offering a “FREE Dinner for Two at Olive Garden”.
There is a lot to grasp here, and much of it is really interesting. Stephen does not pull his punches when he discusses the false steps of Google or Sergey/ Larry combo. The IPO is an example. The Gmail release is another; the tenacity with which they stuck to SAT scores for recruitment (and grades) even after the data (yes, the data!) showed them that it was irrelevant; Lots to enjoy.
Also some trivia, at least for me. We learn that Sheryl Sandberg spent several years in Google before moving to Facebook (as Meta was then called).
The shock inside and outside Google on the decision to display adds (“Is Google snooping on my private email?”) is fascinating as is the decision to go ahead with it.
Also, the book describes how Google was establishing the frontier edge of computing by thinking decades in advance! Sanjay Ghemawat and Dean found out how to use a huge ‘farm’ of computers for a single task by the then revolutionary method of splitting out the tasks and farming it out to different processors on different computers and how to assemble everything into a wholistic answer afterwards. They called the technology MapReduce. In spite of Googles obsessions with secrecy – it even refused entry into its server farms to the Belgian King when he came to visit the first data centre outside USA built in Belgium – they were allowed to publish the paper explaining how it works. The open source folks seized upon it and wrote an open source database that is today famous as Hadoop and is massively powerful.
The part about Chrome’s development and how Firefox, the open source system used by Google, felt betrayed by the ‘Do No Evil’ company. Also this gave rise to using Chrome as a Cloud Operating System. (Chromebooks). For IT geeks this is all old hat but I am sure that there are some who will find it interesting.
And therein lies the dilemma in this book. You need to be interested in Software and computer history to read this book but it helps if you do not know deeply the subject earlier. Even for a computer nerd (and I claim I am a nerd light at least!) this book is interesting.
The author’s description of how Android on phones made Steve Jobs furious and the two companies parted ways is equally candid.
So is the episode where Kai-Fu Lee was recruited when Google decided to enter China and Steve Ballmer, the then CEO of Microsoft went ballistic. He threw chair against the wall in rage, as reported by a sworn affidavit from Kai-Fu (denied, not under oath, by Steve as ‘exaggerated’). The author also mentions an earlier episode where Bill Gates ranted to him saying that ‘China has fucked Microsoft’ making him feel that all his effort in Microsoft China has been in vain. Interesting facts, indeed – not to mention dirty gossip! Gates denied the episode happened.
The filtering was effective but got bad press for Google (when Google.cn opened). For instance if you searched for ‘Falun Gong’ in China, all you got were anti Falun Gong sites. There were, interestingly, two sides to the story. One was that by declaring at the bottom that ‘some results were filtered out’ Google was being transparent and ‘pressure would build up for freedom within China’. The other is that ‘Google compromised on its basic principle of “Don’t be Evil” just to make money in China’s vast marketplace’.
Human rights committee convened a grilling of Google, Yahoo and others who had helped the Chinese government in filtering news to the populace that they did not like. Human rights groups gave pamphlets that accused IBM of earlier cooperating with German Nazi regime, even providing technology that helped exterminate Jews in the Holocaust. Yahoo had provided the identity of a dissident blogger to the Chinese government, causing him to be arrested, which was seen as traitorous by human rights groups. The implication was that the information companies – which claimed to do good to the society – were no better than ‘evil tobacco executives or, indeed, mobsters’.
Representative Lantos, who had lost family members in the Holocaust, was livid and shouting.
However, no laws were changed to stop them all. The naming was controversial. Google could not be transliterated since it was close to ‘Gou-Gou’ which meant ‘dog-dog’. So they settled on ‘Goo-go-a’ which unfortunately earned it a bad name because one of the translations would mean ‘wandering and enough’. So some wondered how a company that cannot even properly pick its name would help Chinese get the right information. After two years, it was changed to GuGe which had no negative connotations!
The Chinese population did not like the new name! But Google decided to go with it anyway.
Meanwhile Baidu was a trailblazer in search in China. It was started by Robin Li, who decided to go back from California. After trying to get other companies to use his search technology (very similar to Larry and Sergey’s in parallel time), he decided to launch his own website. The name Baidu comes from the initial letters of a poem that says ‘I tried to search for a girl in chaos, but suddenly by chance, there she stood’. Early in its time, it was accused of copying Google’s simplistic design and due to the then lax copyright in China, searching for ‘music’ took you to free download sites for music!
When Robin was told to filter results, his dilemma was cost not morality. He complied.
While the author is candid enough to praise Baidu for getting better results in ‘fresh information’ queries, he claims that Google ultimately caught up but the government skewed the tables in favour of Baidu – not to mention a mocking advertisement by Baidu, in the vein of PC vs Mac ads that Apple telecast in the older days.
But a combination of factors (including, the author claims, a deliberate Chinese government attempt to cripple bandwidth for Google China) made the public conclude that Google was inept and inefficient and that Baidu was far better in comparison.
By the author’s telling, Google was doing very well in China (though a distant second to Baidu) but decided to pull out rather than subject itself to increasing government demands and restrictions.
Kai-Fu Lee, who was an “IT rockstar” was poached from Microsoft (and successfully navigated a lawsuit from Microsoft alleging breach of non compete clause in his earlier contract). Due to increasing pressure from the government and (from American guidelines) from Google, he decided to quit. At the very same time, Google experienced a hacking theft and lost some of its intellectual property that was so serious that Google never revealed what was stolen. An internal investigation after the fact in Google concluded that there were ‘strong indications’ that the hack probably had the government’s support. An ‘advance predictor’ of Google China created a huge controversy when all its suggestions were either risque, vulgar, or pointed to porn sites! Google was hauled over the coals by the Chinese government over this.
When one of the sources of the hack was tracked to the laptop (as an interim hop) of a Chinese student at Stanford, the company got hold of the laptop but the program was so sophisticated that it had already self destructed. Sergey Brin had been incensed – partly because of his family’s persecution in Soviet Union earlier – that some opposing views may have been stolen in the hack. He was even more angered when he learnt that some other American companies had fallen prey to hacking but in the commercial interests, had chosen to bury the incident.
His response was to stop censoring altogether! Sergey threatened to expose other companies (more than forty, to his knowledge) but was dissuaded by the legal experts. Google decided to pull out of China. Eric Schmitt, the CEO was a dissenter till the end, and people say that his hitherto excellent relationship with the founders suffered a strain that lasted several years after this. The decision taken was to ‘stop fully censoring the Chinese website’. The results were predictable, Google being banned from China.
To his credit, the author does not spare Google when it does things in stealth. Even if its lawyers are of the opinion that it has the right. One example is its effort to scan all the books. This really put a company which had the motto of ‘Don’t Do Evil’ into a gray area and seems to have landed it in a cesspool of backlash from intellectual property rights advocates.
When cornered, Page seems to ‘blame the law’ rather than the process. He said that the ‘law was not written right because it was written when no electronic copies were available anywhere!’. I see the anger of (even out of print) authors and publishers as a legitimate reaction to Google’s unauthorized scanning of all the books in the world. However, legal and cyberlaw experts argued on Google’s side in a public debate organized by the New York Library. Page saw the whole thing as an attempt to ‘wrangle money from the deep pockets of Google’ and not a realistic objection. He could not even understand why so many folks were angry enough to sue Google. Wasn’t he both helping them sell more books and have the world’s knowledge searchable?
Whichever side of the argument you subscribe to, the lawyers of Google made a powerful argument : the copyright act exists to protect the rights of the creators. In creating interest in more people and not selling the books and directing the queries to the appropriate sites, Google is indeed respecting the spirit of the copyright law!
The epilog about the (then) brash new company which was trying to outdo Google in ambition, aka Facebook, is fascinating too. By growing in orders of magnitude, poaching Google employees and also buying companies created by Google employees and poaching them that way, Facebook created a lot of competitive fear with Google. (Facebook refused in its early days the offer of Google to buy it out!)
The irony is that one Google employee developed a ‘connectivity cloud’ earler than Facebook. Initially named Friendster, Google struggled to find a commercial non trademarked name for it. Finally it was named Orkut after the creator’s name (as a separate company belonging to him. He was from Turkey originally and was named Orkut Buyukkoten.) The software was ONLY available through invitation, which initially limited its reach. It was killed by identity thieves and spam, as we understand it. Partly because Google did not integrate it but left it to its own devices, to see how it swims or sinks, as recounted by the author in this book.
Facebook meanwhile ate its lunch. By the time Orkut’s potential was realized, it was too late!
The other reason was that Orkut was not fast like Google’s other services – an obsession with Larry and Page. Orkut was extremely popular in two countries – Brazil and India. These were explained (unsatisfactorily in my opinion) by the fact that “Brazil was used to slow services” and India coincided with nighttime in US when the server load was low and consequently the response time was fast.
Another boat that Google missed was location based services. Even though it bought off an app called Dodgball that introduced location based services. The privacy concerns created the idea that you can share messages with friends without the location. Just to your group. Google did not exploit it fully. This was the idea behind Twitter, which grew into another giant of the Internet.
The author argues, convincingly, that the idea of people talking to each other (social networking) without ‘an algorithm in the middle that determined the best way’ was anathema to Google’s founding principles and this is the main reason why it missed the social network revolution completely.
Google tried to get into social networking through a backdoor. Its initial attempt of a portable ID was killed by Facebook refusing to cooperate – and thus ceding control to Google. Then it wanted to buy Twitter but its founder, bitter at the lack of interest shown by Google when it bought his previous company called Blogger, refused to play ball.
Google never created a clone of a successful company elsewhere (As someone said, they could have ‘written Twitter over a weekend’) because, again to quote ‘It has never been successful chasing taillights’. Finally, the company threw its wait behind a product called Buzz that produced groups (like Twitter) based on user interest elsewhere. It was released in 2010.
But Buzz created a privacy concern that was intense. (Using search and interest history to create groups? How many privacy rules does this violate?). When you logged on to Buzz, it gave you all your contacts and you could choose people to come into Buzz. It also gave access to their data and their full social contact network! Imagine disclosing that a friend added a lot of headhunters in your LinkedIn or equivalent when you entered a person into Buzz! Or worse.
And all the people you follow (and those who follow you) are made public. The public was horrified.
Google found that Wave cause lawsuits and unintended consequences in authoritarian countries by exposing the connection of dissidents. Even Canada accused it of privacy violations on a large scale. The backlash stunned Google and spelt doom on Buzz. Google also withdrew Wave at the same time, probably a double failure to the company. Google execs were still probably in denial. Bradley Horowitz said “We thought that after Facebook, this was something that people were used to”. Really?
Meanwhile in another blow to the competitive landscape to Google, Microsoft and Yahoo announced that henceforth, Microsoft’s search engine will serve Yahoo queries. With Bing improving its quality of results, though Google was outwardly calm, it sensed a growing threat internally.
When Bing did not affect Google’s search share much, the urgency was abandoned. Then Mark Zuckerberg became the bigger threat and Google pulled out all the stops to create a social network codenamed The Emerald Sea. With superambitious targets, it seemed to go nowhere while Facebook went from big to bigger (and scarier as a competitor).
Google spoiled its name further by changing its views on Net Neutrality and getting in bed with the chief opponent of the principle, Verizon, even Jon Stewart got a segment created on the descent to evil of Google.
The book ends, since it was written before, without mentioning the change in leadership to Sundar Pichai. The book ends when Larry Page has taken over the rains as CEO from Eric Schmidt, and Google was still Google!
All in all, an excellent, and unbiased coverage of the Google saga. A delight to read.
8/10
— Krishna