The Null Device

2018/3/22

So, the Cambridge Analytica revelations: The broader outline (shadowy company, owned by right-wing billionaires and possibly connected to Russian intelligence interests, uses big data, psychological algorithms and targeted ads to fix elections, including the 2016 US Presidential one and the Brexit referendum) had started trickling out since late 2016. They would have remained an ominous undercurrent, known of by the vanquished liberal minority and embellished with layers of speculation, conspiracy theory and fatalistic gallows humour, were it not for a former operative spilling the beans, revealing further details. Details such as that that the data went beyond the usual broad-brush correlations between brand likes one thinks of when considering ad targeting and went into alarmingly intimate data aggregated from sources including Facebook private messages and credit profiles; and that, while the data was taken from Facebook by deception (under the claim that it was to be used for academic purposes; naming one’s company “Cambridge Analytica” can yield dividends, it seems), Facebook did not try particularly hard to enforce the terms of use ostensibly protecting its users (and, it seems, democracy and civil society themselves).

The elephant in the room, of course, is Facebook and its business model, which has been described, aptly, as “surveillance capitalism”. Facebook does not charge for its services, and funds itself through a modest amount of ads. What it does, however, is build up elaborate profiles of its users, sucking up their every online interaction it can to extend them. Where likes, posts, messages and location data don't suffice, it supplements with data from data brokers, gathered from credit ratings, things such as loyalty card programmes, and public-domain data. This yields a vast amount of data, which, when processed with sufficient computing power, can reveal a lot about users: in their less circumspect moments, Facebook's data scientists have revealed that they can tell a lot of things on an individual from their data trail, from political leanings to sexual orientation, to how likely their current relationship is to end; what other things they can determine about an individual given that data is an open question. What is not disputed, though, is that Facebook's very business model depends on using this data to target advertising, and their acuity in doing so to make revenue. In short, everything Cambridge Analytica allegedly did with their haul of illicitly obtained data, Facebook can do, continuously and at far greater scale.

This has not escaped notice, and the backlash has hit Facebook, not helped by Mark Zuckerberg's protracted silence on the matter. The hashtag #DeleteFacebook is trending on Twitter, urging users to nuke their Facebook accounts, and giving instructions on how to actually force Facebook to delete their data permanently, rather than just flagging it as temporarily inactive; the latest to add their name to the chorus is one of the founders of WhatsApp, a messaging system now owned by Facebook. Of course, that is easier said than done.

In some ways, it is easy: the instructions are easy to follow, and involve clicking through a few screens, entering the text in a CAPTCHA and watching one's digital life and/or surveillance dossier go up in virtual flames. What comes after is the hard part. Facebook have worked hard to own the social graph, and to make it difficult to take your friends with you when you leave. (The timeline you see on Facebook is, famously, not a chronological firehose of all your Facebook friends' posts, but a selection of a small proportion of those, chosen by the same all-knowing algorithms that know which ads to show you. It is said—though is, of course, difficult to verify—that those algorithms specifically score posts giving or requesting means of contact outside of Facebook down.) Or course, downloading a list of one's friends' names and email addresses/phone numbers is a nonstarter from the outset, ostensibly for reasons of privacy, so one's only bet is to individually contact the friends one wants to keep, one by one, to exchange details, resigning oneself to losing contact with the rest.

Other than owning the social graph, Facebook have managed to become the hub of socialisation in the 2010s by reducing friction. It is a lot easier to create a Facebook event for, say, a birthday party or an excursion to see a film, than to individually invite people to add the event to their calendars, and conversely socialising without Facebook has been consigned to the realm of Victorian-era calling cards. It is possible to socialise deliberately, with a few close friends one has in one's address book, but this does not scale. Social software such as Facebook has genuinely reduced friction and saved time, allowing people to keep in minimal touch with people they are peripherally connected to, and also reducing the time required to keep in touch with other, less peripheral, friends; the hectic pace of modern life has filled the time thus freed up. Thus, most people have no more time for phoning/texting/emailing non-core friends, who all are on Fscebook anyway, than they do for composing a multi-paragraph blog post which nobody will probably read. The sad reality of the 21st-century human condition is that Mark Zuckerberg owns the space between us, and leaving Facebook (and Instagram and WhatsApp) is, in many ways, to retreat from modern society; a radical, deliberate act, equal parts Ted Kaczynski and Into The Wild.

Perhaps this will change; perhaps Facebook will end up going the way of former unassailable titans such as MySpace, and be replaced by something else. If that something else is run by another venture-capital-funded surveillance-marketing organisation, the situation will not improve. (Facebook itself, at the start, seemed like a much cleaner and less spammy/scammy operation than MySpace.) There are glimmers of hope in the decentralised sphere, where developers are creating open, decentralised alternatives to corporate-owned monolithic silos. One example is Mastodon, a Twitter-like microblogging system. One thing those services don't currently have, though, outside of mass adoption, is granular privacy settings: it is not possible yet to make non-public posts to one, or to filter posts by group. Once this is implemented, in a way that works robustly across instances, and presumably uses cryptography, there might be a system ready to step into the gap when Facebook stumbles that is not inherently predatory on a structural level.

For now, however, those who are not willing to brave the wilderness are more or less stuck with Facebook; however, it is possible to reduce your profile, by reducing the amount of personal data one posts to it, and removing it from one's mobile devices (or, if not possible, revoking the app's permissions to access things like location information). I would add to this list: not logging into Facebook from one's main web browser, keeping a separate browser with separate cookies for using it, ideally running in a separate virtual machine.

cambridge analytica facebook postdemocracy privacy surveillance 0