Herald Sun (Australia), June 29, 2003
By GRAEME HAMMOND
For millions of web users, Google is the friendly face of the Internet, the brisk, clean and efficient pathway to exactly what you want.
Whether you search for “deep sea diving” or “Deep Purple” — it depends, after all, on your interest in reefs or riffs — within seconds you will have an array of thousands of websites to choose from.
Google is an amazing search engine with an amazing history. It is already an indispensable web tool, despite this year celebrating only its fifth birthday.
But questions are now being asked about the power it wields over the Internet, the level of clandestine censorship it operates and the extent of personal information it is amassing on every one of its users around the world.
Though it claims to operate in a way that reflects the “unique democracy” of the Web, it is now the subject of the same type of scrutiny and suspicion long given to another technology behemoth, Microsoft.
None of this was the probable intention of its founders, Stanford University computer science graduates Larry Page and Sergey Brin, who first met in 1995. By January 1996, they were collaborating on the biggest challenge of the emerging Internet: how to retrieve relevant information from a massive store of data.
The result was a search engine called BackRub, named for its unique ability to analyse the “back links” pointing to a given website from other sites. The more links from outside sites, the greater its assumed relevance to search terms used.
Housed first in Page’s dorm room at Stanford, BackRub was soon renamed Google, a play on the word googol (the number represented by the numeral 1 followed by 100 zeroes), reflecting the pair’s mission to organise the apparently infinite amount of information available on the Web.
By the time it moved into an office space in September 1998, with a staff of three, Google was already handling 10,000 search queries a day. Five months later it was taking more than 500,000 queries a day.
A year after its launch, Google was attracting corporate backers and was housed in a two-storey complex in Mountain View, California.
Its office was like a hi-tech Willy Wonka chocolate factory: staff sat on large rubber exercise balls instead of chairs, computers were placed on doors held up by pairs of sawhorses, the interior was lit by lava lamps and its office space was roamed by large dogs. In June 2000, Google became the world’s largest search engine, accessing one billion web pages and answering a staggering 18 million queries a day. It introduced its own toolbar, was launched in a range of languages and began selling advertising.
Later that year, the first rumblings about the darling of the Internet began to surface. It was a Left-wing political activist from Texas, Daniel Brandt, who started it, noting that his website – a database of news clippings and facts he had been creating since the 1960s on a list of 100,000 public figures – was coming up only sporadically on Google searches.
Brandt was already aware that one of Google’s key drivers in assessing the relevance of returned search results was its “PageRank” system, in which every website is accorded an arbitrary level of quality and importance.
This, in turn, dictates the importance of links to and from those sites, which then influences where a site will turn up in a search result.
It was only when Brandt downloaded the Google toolbar that he realised PageRank’s impact. The toolbar displays, with a crude bar indicator, Google’s ranking of any website called to the screen. Websites run by large corporations scored a consistently high PageRank; Brandt’s own site, NameBase, seemed to rate a zero.
To Brandt, it was bad enough that links to or from his website were so poorly regarded. The effect was magnified by the fact that the so-called robot crawlers or “spiders” that trawl the Web’s three billion pages on month-long cycles in search of links and key words are programmed to crawl deeper, and more frequently, into sites with a higher PageRank.
The effect, says Brandt, is a vicious cycle: with a low PageRank, your site may not be crawled and therefore will not appear in Google’s index. The result: no links from other sites and a perpetually low PageRank.
“In other words,” he says, “the rich get richer, and the poor hardly count at all. If you don’t get on Google and people don’t know about you, there’s no point in maintaining a website.
“This is not uniquely democratic, but uniquely tyrannical. It’s corporate America’s dream machine, where big business can crush the little guy.”
Google admits new and minor websites may not be crawled, even if their URL is submitted to Google. Their advice: seek a link on Yahoo! and you will stand a better chance of being crawled. In the incestuous world of search engines, Google crawls Yahoo!, which in turn pays Google to use its search engine. There is a complex interplay between all major search engines, each feeding off one another.
The PageRank system has long been a target for abuse by devious webmasters desperate for a high listing on Google. In March last year, it was discovered that the Church of Scientology had created a mass of websites known as “link farms”, many consisting of nothing but links and duplicated data, in a bid to ensure their website, Scientology.org, scored as the top website.
In fact, Scientology websites are outnumbered on the Web by sites critical of the cult, many of them academic and reasoned in tone. Last year, the organisation stepped up its campaign against rivals with a letter threatening Google with legal action if links to one of the most prominent anti-Scientology sites, Xenu.net, were not removed.
Google acquiesced, and for a day Xenu disappeared from its listings. The next day, after a barrage of complaints, Xenu.net was re-listed. Google did, however, begin banning on-screen ads critical of Scientology.
Scientology opponents and conspiracy theorists have also noted that the Open Directory Project, a human-edited web directory that is one of the major feeders to search engines of new links, readily accepts pro-Scientology URLs but ignores or delays the addition of links to sites critical of the “church”.
Exhaustive studies by respected researchers at Harvard University have revealed that Google also quietly removes links to certain sites – generally hate sites by white supremacists and other fringe groups – at the request of foreign governments.
Does Google release this information to third parties? Only if forced to by courts or law enforcement agencies, they say. How much of such data on individuals’ surfing habits have they supplied to police so far? They refuse to answer that question.
“The privacy of our users is of the utmost importance,” a Google spokesman said this week. “But I can’t speak on legal issues.”
Among web designers in Australia, Google seems well regarded. Tim Fouhy, a director of Reactive Media, whose clients include Melbourne Airport and John Farnham, says Google is constantly changing its algorithms to prevent abuse of its rankings.
“It’s a good company. I trust them,” he said. “We submit new websites to them and they’re posted incredibly quickly, sometimes overnight.”
Rodney Ferro, head of Protocol Networks, which offers a website promotion service that uses sophisticated methods to ensure a high ranking of sites in search engines, says: “I think Google is one of the better search engines. We’ve never had a problem.”
A Google spokesman said this week the company’s goal was to provide as comprehensive a directory of the Web as possible.
“Our goal is to get information. It wouldn’t make sense if we worked against that by ignoring websites,” he said.