Thursday, May 19, 2005

More on Google's dominance

Two days ago I commented on the worrying developments regarding Google's Web Accelerator. This could have serious consequences, see for example this spoof site about Google's Content Blocker (thanks for Anne-Kathrin for the link). I recommend reading their privacy policy, where it states that "Google is committed to collecting every bit of data that it possibly can, and storing it forever." I am coming back to this because I'm reading an excellent book called Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life by Albert-Lászlo Barabási. This is a brilliant book about the science of networks, and it explains the reasons behind phenomena like six degrees of separation and Pareto's 80-20 rule.

In short, Barabási's argument is that networks respond to certain rules that apply almost universally to things like quantum physics, the internet and celebrity links. Networks are not random, they grow at a specific rate, in which the older hubs have the advantage because they can develop the largest collection of links, but also there are hubs that acquire links at a faster rate, and can overcome older ones (fitness of hubs). Because society is a network where each one of us is a link, there are people who collect links faster, and this serves to prove the observation that the rich get richer. If we think about this in internet terms, Google was not the first search engine, but it obliterated its competition because it had better fitness. Why is Linked important for Google? The thing is that Barabási's mathematical proofs can have a strange effect, and it is that in some instances, it can lead to a collapse of competition, and the winner takes all.

According to Barabási, this occurs in physics with the application of Einstein-Bose equations about gases, and his claim is that this applies as well to other networks. If this is true, then what we are seeing with Google is not only predictable, it may be inevitable.

No comments: