This is it. This is goodbye.
The rot set in several years ago when I stopped being interested in stalking my ex-boyfriends and your primary function for existing was lost, then you started harassing me with slipper advertisements, and finally this article by John Lanchester in the LRB (https://www.lrb.co.uk/v39/n16/john-lanchester/you-are-the-product) has me terrified of you.
Now as a housewife I realise I’m a perfect fit for slipper advertisers. A shoe-in. But do I really need to be exposed to high-end slipper advertisements every single time I go on facebook for months on end? Do they need to appear multiple times in one session? Do these advertisements need to follow me around so that they pop up everywhere else on my web browser too? And fine, yes, after months of this war of attrition by slippers I bought the damn things. So did my mother-in-law. But STILL THE ADVERTISEMENTS APPEAR. All this information you have on me, the reams of data you have stored in relation to me, and slipper manufacturers are the only companies you can flog my advertising space to? Either you and your algorithms are useless, or all I want in my secret heart of hearts are a pair of comfortable yet durable slippers. Perhaps that’s all anybody really wants.
From all the things I’ve read about you though, I worry that you will get better at collecting and putting together all the data that you have. Not so much for me (my only identifiable desire being slippers, I am fairly incorruptible) but for my daughter.
I worry that your facial recognition technology will improve to the point that you will be able to link up all the non-tagged photos I have posted of her as a child across social media with photos of her as a teenager when she joins the site/instagram/whatever new variation is around that you have bought up in 15 years time, so that you have a complete visualisation of her as she develops, coupled with all the other data you have collected on us her parents, and her network of friends and family.
You’ll know if she has siblings or pets, how often she has moved house as child, whether she grew up in an urban or rural setting, whether her parents are together or end up getting divorced, whether there has been the trauma of a close family member or friend dying during her childhood. You’ll know what parenting advice her parents have ‘liked’ and/or doled out, (probably somewhere is a comment on how long I breastfed her, and whether I let her cry it out or not), you’ll know if and when I go back to work, which playgroups and schools she attended. You’ll know what political views her parents and she have through what news items we share. You’ll know if either of her parents are part of a patient support group on facebook and whether or not their health problem is hereditary. You’ll know what she did and where she went on holiday. You’ll know from her photos and the facebook groups she’s part of what she’s interested in, how much she participates in sporting and outdoor activities and whether or not she is religious. You’ll know her sexuality and relationship history. You’ll know whether she has cyber bullied or been cyber bullied.
You’ll extrapolate from how many likes there are on each photos of her as a child whether or not she has a strong network of friends and family who take in interest in her development, and later when they are her own photos whether or not she has a close knit group of friends. From data on her and her parents usage of the site and number of time they use specific words/phrases you’ll determine their moods across her life to date and be able to detect any patterns of depression or anxiety. Matching up data on where she has lived with maps of pollutants you’ll know for how long it is likely she has been exposed to what degree of air, water and noise pollution with what likely effects. You’ll be able to estimate her BMI and calculate her waist to height ratio from photos over the course of her life. You’ll be able to work out with a fair degree of accuracy how much disposable income as a family we have and define which socio-economic class she is in.
All of this data you have on each of your facebook members will be compiled and stored and searchable to your algorithms. Without control groups and very detailed analysis you won’t be able to figure out patterns of causation from this data, or answer any ‘why?’ questions but correlation alone will be enough for you. You will mine this data for correlations with various mental and physical health outcomes, academic successes or failures, employability in various sectors, alongside of course improving your ability to predict what products people are likely to want, what their individual price points are for various products, and when they are most vulnerable to marketing.
Will you be able to calculate with a fair degree of accuracy how long she will live? Will you give her a mental and physical health rating? Or an overall well-being score like a credit rating that individuals can bump up with their fitbit data if they wish to (do not get me started on fitbits)? Will you have a rating for her ‘resilience’? For how easily she is manipulated or ‘nudged’?
Will you then sell on these ratings to private health and life insurance companies? To mortgage providers? To future employees of her, so that when they review her C.V. (or get an algorithm to review her C.V.) they can check this against her ratings and the data you have on how well her personality will fit with the corporate culture they have and how much of a pension she is likely to need? To security firms who want to use the data to work out the likelihood of her being a terrorist threat and how much airport security she needs to be subjected to?
I do not want her to be reduced to a rating, or series of ratings. For all her data to be fed into self-reinforcing echo-chambers and loops (from her data we think she will be interested in X activity/career/relationship/product/political party, therefore we will connect her, i.e. market to her heavily X. She then does or buys or votes X, thus feeding into the correlatory data that people with similar data points are interested in/do X). I do not trust you, and I definitely don’t trust private health and life insurance firms with her health data. I don’t want potential universities or employers to turn her away based on your impenetrable, unverifiable and very likely flawed algorithms.
You may think it’s worth it for us to freely give all this data to you. That you won’t use it/sell it on for personal profit (HA) but will mine it and correlate it and use it come up with cures and prevention’s for diseases (will you have helpful AI therapy bots that will intervene when you notice a user is in mental distress? A bot that sends you tailored recipe ideas/healthy deliveroo options if your BMI is looking a bit high?) and that if we have nothing to hide we have nothing to worry about. Airport/immigration checks will become a breeze for us if we have low terrorist risk scores. I can stay connected to my friends and through my local community facebook groups get parenting advice and swap children’s toys. What could be more benign? The only price I and my daughter in the future needs to pay is constant aggravating slipper advertisements.
But in this all seeing environment of course I have things to hide. My Black Mirror-esque paranoia which may count as a negative mental health marker for my daughter for one. And of course my daughter will have things she’ll want to hide. I don’t want her making an innocuous ‘mistake’ when she is young like breaking her leg flagging her up as having a predisposition for athletic risk taking and future joint problems, which means she is heavily marketed to by sporting adventure and untested designer pain pill companies, and has to pay high health insurance premiums when she is older. I don’t want her life to have a pre-destined predicted outcome based on her childhood that becomes a self-fulfilling prophecy.
I do not trust you with the story of her life.
So I’m deleting you facebook. And Instagram. (As well as all of the above issues I’m trying to fly less in a stand against its effects on climate change, and looking at my friend’s holiday photos is not conducive to this. Also I don’t care what anyone else is eating).
To all my friends you know where to find me, and I wish we saw each other in person more often. To my ex-boyfriends I apologize for all the stalking. It was done from a place of love.
P.S. Facebook if you change and become socialized ala this Ben Kunkel article I may reconsider my decision, because I do think that internet platforms like yours can be used for good: https://nplusonemag.com/online-only/online-only/socialize-social-media/
P.P.S If you think any of this is far-fetched, have a read of Infoglut by Mark Andrejevi. If you think algorithms are value free, try Weapons of Math Destruction by Cathy O’Neil. And for a mind blowing read on how we make sense of the future, self-fulfilling prophecies and investment decisions and all, go for ‘Imagined Futures’ by Jens Beckert.