How to undermine Facebook

Like many people, I have a Facebook profile, but it isn't something I use much. And like many people I've become increasingly uneasy about the way our interactions are monetised and manipulated. I don't mind this to an extent - I'm willing to trade off free use of something for some adverts I ignore, for example. But it's begun to feel all rather insidious with Facebook.

Alan Levine posted that he wasn't going to dramatically quit Facebook, but his plan was to maintain a non-presence:

But quitting seems to pointless. Or impactless. So I have an evil plot. I am keeping my facebook account, but I have completely neutralized its presence. I belong to no groups, like nothing, or use anything that sucks in my information 

But I think you could take this one step further. In those fantasy books and comics, when a human needs to slay a monster, they nearly always have an epiphany where they use it's own strength against it. The same approach could be applied to the behemoth of Facebook.

What Facebook really sells to companies is meaningful data and connections. The key to this is that they are meaningful. If I 'Like; a company it matters because my friends will believe that I really endorse their products. If I install an app it's because it appeals to me. And if people who Like X tend to also Like Y then that becomes valuable data.

But what if those actions weren't meaningful? What if they were genuinely, and deliberately random? A company may get some benefit along the lines of 'all publicity is good publicity', but the real value of knowing it's something I genuinely like is removed. And this corrupts the data set, because those connections between points are no better than a random number generator.

So, if I was a real hacktivist type, I'd create a tool that searches Facebook and randomly suggests companies to Like and apps to install. Then we'd all do it, and Mark Zuckerberg would cry 'I'm melting, I'm melting.'

This is only a half-serious suggestion - the point I want to make is that in social media, it is integrity that counts - data integrity, personal integrity and behavioural integrity. Which prevents marketing firms and retailers with a problem as often they are trying to modify or coerce this integrity. But as soon as you do that, the whole approach becomes meaningless. If I retweet something because I might get a prize, then people don't value that tweet. And so the moral is, don't sell your integrity cheaply. And don't manipulate our integrity or we can use it against you.

Comments

[twitter.com/cogdog]

I did consider that as an idea, Martin- and curious if anyone could estimate what the impact would be or what sort of numbers it would take to generate some disturbance ripples.

But I imagine it is more than just recording a string of bad likes; the traffic Facebook exchanges is all about our habits or our profiles (do we even know what data they ship to companies).

Yes, your plan takes the deviousness to a delightful higher level.

[Martin]

You're right Alan - in reality it would require the majority of users to do it for it to really impact. They undoubtedly build a certain amount of noise into their data and analysis anyway.
And to really undermine it, as you suggest you'd have to put random info in your profile (another tool for someone to develop).
But the principle is sound I think - social data only becomes valuable if the individual elements have integrity.

[Nick Kearney]

The logical next step would be a tool that generates "bot-like" self-replicating profiles that like everything they find and interact randomly, multiplying eventually to a larger number of profiles than the population of the planet.
This would probably lead to the spam-death of the universe.

[PeteJ]

I'm fascinated by these ideas of exploiting the gaps between the intent of the creator of these various social media markers and their interpretation(s). Your post reminded me of this First Monday article by Finn Brunton and Helen Nissenbaum

http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3493/2955

which explores how individuals and groups adopt various tactics of "obfuscation" to try to counter exploitative monitoring, mining and profiling.

That piece in turn references the work of James Scott, which I recall enjoying as an undergraduate, on how groups with little economic or political power use an often shifting range of ad hoc tactics to more or less quietly disrupt or confuse those seeking to control them.

Even leaving aside the element of active obfuscation, I do wonder whether ascribing some simple, uniform interpretation to a "like" or a "favourite" isn't always somewhat problematic, whether there is always an element of ambiguity. It seems to me there are all sorts of possible intents behind a "like" of, say, the Fb page of university XYZ, from an active signal of my enjoyment/valuing of my time as a student there, to an expression of identity through association ("University XYZ is cool"), to opting in to some social reward thing ("Like our page today and you'll be entered in our raffle for an IPad!"), to an accidental click on the wrong button in my phone app. This is one of the reasons I find Brian K's exhortations to use these markers as "lightweight" measures of "impact" and "performance" so problematic. But that's another comment, and one for his blog rather than yours! :)

[Scott Leslie]

Check out TrackMeNot http://cs.nyu.edu/trackmenot/ - not Facebook specific but more general. Hide in plain site. Oh, and for good measure, damn the man.