by Felix Cohen
Last week saw two very significant pronouncements from two of the most significant online businesses; Facebook and Google. On Tuesday, Google was reported in the Financial Times as having the intention of gathering 'so much information about individual users that it could even offer suggestions on how to spend free time or what career move to make'.
And on Friday, Facebook, the 6th most trafficked site in the US, held a press conference to announce that they were not being acquired at this point, and that instead, they were offering the Facebook service as a universal platform for other companies to use as they wish. For example, openDemocracy could (and might) add facebook membership to our members (the people who receive our e-mails and can login on the site), so that you could discuss articles, have discussions outside the scope of the forums, form activist groups based upon various issues or geography. We would be able to create an 'application' on facebook, deeply connected to the community here. I predict that a lot of online communities such as ours will be considering adding the features that facebook can offer soon*.
Facebooks announcement, as well as Googles reaffirmation of its position as maven of the worlds information, leave me optimistic for Web 2.0, but gives cause for concern. This much data about people's personal life, habits and behaviour betrays deep patterns in life, as betrayed by a mistake Sainsbury's (a UK grocer) made, when through deep 'data mining' of their loyalty card data, they began sending out baby related vouchers to households where spouses were unaware their partner might have been trying to become a parent. While it is fine, nay preferable, for Google to be able to target me with adverts that are appropriate for what I want, I am increasingly reviled by the idea of the company being able to predict my career, activites and other proclivities. Facebooks expansion, and increased utilisation of the 'social graph' similarly concerns me, as it increasingly becomes the portal through which I organise my social life.
But what is going to happen to the ad hoc decisions that are made; last minute party invites, the digital camera that my local shop sells cheap but which I never see as I searched through Google or meeting someone genuinely different and unconnected to me.
This is the problem with 'the algorithm', as so widely touted by Google, Yahoo and even Amazon and Facebook as the reason for their successes. And the algorithm is great for pulling some stuff out of the long tail, but ultimately it is not a catch all net; the algorithm is limited to extrapolating a pattern in data; one person likes this, so as you have behaved similarly to them in the past you are likely to behave similarly to them in the future. People complain when this sort of behavioural profiling is used for terrorism suspects or to profile potential criminals in schools, but we are happy for private companies to analyse every scrap of data they have in order to make more money from us. And Google, despite claiming to 'not be evil', are very opaque about the data retention policies (although some concessions have been made recently); it seems safe to assume that every scrap of your online behaviour is being recorded and used to predict future behaviour. Even openDemocracy records the paths that people take through the site, as well as monitoring a tiny percentage (0.04%) of users even more closely (how far you scroll down the page, and how long you stay at each part of a page). We don't do this to sell more product to you, but it has real value to us in terms of making sure the site is working for people, and generating revenue through advertising.
*This is made possible when websites offer themselves as a 'platform', offering their various capabilities to other websites. This is done through the design and publishing of an interface that operates in parallel with the browser interface that you browse to. The platform interface allows specially written programs that run on one webserver to make the same queries as you or I make; getting search results from Google, looking up books on Amazon, accessing images on flickr and so on. Deeper technical details are beyond the scope of this article, but look at some of the further links below to find out more.
