Even by our jaded postmodern standards the story of Google is jaw-dropping. The company, barely seven years old, has become so pervasive that it has spawned a verb; it also boasts a market capitalisation of roughly $85bn and is relied on by Chinese dissidents to bypass government censors. Worldwide, there are around four billion internet search requests each month; Google handles about half of them. According to John Battelle, Google "fundamentally changed the relationship between humanity and knowledge". This could be dismissed as classic Silicon Valley hyperbole, but in a few subtle ways it is no less than the truth.
Before Google the technologies for searching online were so primitive as to be virtually useless. Online search engines basically scanned websites for the words on each page, and indexed them. The more times a given word appeared in a page, the more likely it was that the page would be about that subject. So a site featuring the word "inflation" more often than any other would rank higher in the results returned to users searching for information on that topic, regardless of whether it was truly the best source (and irrespective of whether it referred to economics or blowing air into tyres). It was a fundamentally idiotic way of finding content, and even lent itself to the creation of false results. (Pornographers, for instance, would invisibly add hundreds of instances of the word "car" to their homepages, thereby guaranteeing themselves a high ranking when people searched for that term.)
In 1997 two Stanford University graduate students came up with a different way of organising content. Instead of just looking for the words on a page, they coupled that with the number of sites that linked to each of those pages. Larry Page and Sergey Brin got the idea from the world of academic citations. An article that is frequently referred to in other papers is probably better, or at least more relevant, than one that is ignored. Google's technology applies the same approach to the web. A page that appears as a link on lots of other websites will rank higher in the search results than one that doesn't. Google, essentially, is the world's biggest popularity contest.
This method has another advantage: while other search engines become less reliable as the size of the web grows, Google, because it relies on the collective intelligence of users to identify content through the links they embed in web pages, actually becomes more accurate.
Every medium is defined by its limitations. The printed page is hampered by physical distribution. Radio and television have always relied on the airwaves, which are cleaved into different channels by regulators. The internet, however, is a medium defined by abundance. Google, whose name is a play on the word "googol" (a mathematical term for the number 1 followed by 100 zeros), is important because it found an elegant way to manage this abundance.
But what of the future? All online companies are media firms to some extent, and all media are, in a sense, search engines. When we are faced with an abundance of information, an editorial filter - whether provided by a record label, a newspaper or a retailer such as Amazon.com - is indispensable. This helps explain why so many online firms are moving into search, and why Google is using its dominance in that area to move away from it, by extending its technology to e-mail, web-based voice-calling and shopping (via Froogle). Google's founders have realised that the same technology that serves up search results can be used to provide anything from a text document to a spreadsheet - which is why Microsoft is getting worried.
Will Google's success continue? Just as Battelle's book is marred by occasional bombastic statements, so the biggest threat to Google comes from its inflated self-importance. One fears for any company that has as its motto "Don't be evil", which, aside from being obvious, says next to nothing about such activities as competing with rivals or operating in China. It would be a pity if Google, after changing the world by making information accessible, stumbled simply because it had too much faith in its own powers.
Kenneth Cukier covers technology for the Economist