words William Wiles
Sergey Brin and Larry Page founded a company together on 7 September 1998. They were both just 25. The company was called Google. Exactly 10 years later, they are worth $18.6 billion. Each. They employ 19,000 people.
It seems superﬂuous to introduce Google’s ﬂagship product: it’s a search engine. It ﬁnds and retrieves information on the web. It happens to do that job very efﬁciently. But that alone doesn’t account for its epochal impact both on the market in which it competed and on the culture of the internet as a whole. To understand that fully, it’s necessary to remember what the internet was like before Google. Which is a difﬁcult feat in itself.
We may use it for communication, but at heart the internet is a database. From the beginning, this bewilderingly huge storehouse of knowledge needed a search function. And the ﬁrst ones just weren’t very good. The results were patchy, disordered and slow. It was a common experience to wait for what now seems an unthinkably long time for a selection of results that were not only irrelevant, they failed to include sites the searcher knew to exist. It was at best frustrating and at worst agony.
Google was a revelation. Results appeared in fractions of a second. They were comprehensive. And they were pretty well ranked in order of relevance – not always perfectly, but normally well enough to save a huge amount of time. It was a memorable moment that drew a line across your experience of the internet – before Google, and after. Such was its obvious superiority that it went from unknown to ubiquitous in two years through word of mouth alone, entering the OED as a verb in 2006; today, it maintains a near- monopoly over search worldwide. In the UK, eight out of ten searches are on Google.
How was this technological superiority achieved? By cunning and brute force. The real innovation in the Google design is the Pagerank system, which rates web pages according to how many other pages link to them; the more sites that link to a webpage using text from a search string, the more relevant Google thinks it is. The brute force part is that Google has made a copy of the internet. Well, not all of it (the percentage is unknown), but its web-crawling programs have taken a representative snapshot of the internet and stored it on Google’s own computers. When Google is searching, it is mostly searching its own memory, not scores of remote machines; that is what slashes irksome seconds to the blink of an eye. This system needs a huge amount of storage space – and Google built that, not using bespoke supercomputers, but off- the-shelf CPUs in awesome numbers. The speed of the search is a testament to the incredible power of parallel computing. It’s the entrepreneurial, DIY ethic of the internet at its best.
But the internet is forever growing, and changing, and Google’s index of it has to be perpetually updated by its army of web-crawlers. Its thirst for data is insatiable, and as its range of products and services has grown through expansion and acquisition – email, documents, books, maps, Blogger, YouTube, the Google phone is coming soon – so has the amount of information, public and private, it is privy to. And once it gathers a piece of data, it doesn’t let it go. Even crediting the company with wholly benign motives, this unprecedented and growing stockpile is inherently authoritarian. Google knows more about you than you should be comfortable with – unless your name is John Smith.