Kath Rella, Editor-At-Large
Earlier today, it came to my attention that I have a profile on a site called Muck Rack. Alongside my name sat a selection of articles I’d written. Two of these articles had been pulled directly from this website, which as anybody can see, is nothing but a collection of the ill-managed scribblings of a sleep-deprived madwoman. The other was a link to Sarah Jamie Lewis’ Queer Privacy book, to which I contributed an essay.
Somewhat alarmingly, Muck Rack promised my email and phone details to anybody signed up to their site. Having never heard of Muck Rack before, and thus satisfied that I had provided them with no such information, I had a few questions. Was Muck Rack, in fact, a misspelling of Muck Rake and therefore out to dox all and sundry for whatever nefarious reason? Had somebody created a profile, pretending to be me, to some bizarre end? Am I suffering from a brain tumour causing me to blackout and forget creating public profiles all over the web?
Well, no. Any fears could be quickly allayed by a search of Twitter. From a cursory glance, Muck Rack is nothing more than LinkedIn for PR people and journos. Why am I on there? Because to build its userbase, Muck Rack has deployed an algorithm to scan the Internet and automatically curate the writing of people it suspects are journalists. High-profile journalists are then contacted directly by a bloke named Andrew who no doubt tries to get them to claim their profile to add to Muck Rack’s legitimacy within its industry.
Only, I’m not a journalist, am I? I am, as I described myself above, a madwoman who writes random articles on a blog at an infrequent rate for no discernible reason. That Muck Rack’s algorithm has conflated haphazardly-produced writing on contentious issues such as privacy and censorship with ‘journalism’ doesn’t make it so. At best, it’s evidence of a quirk in its design and deployment.
Lies, Damn Lies, and Guesstimates
But let’s examine this quirk in a broader context. Algorithms are absolutely everywhere online; they are used to decide what adverts you see, what content needs to be censored for whatever reason, and even how you’re treated by service providers. At their core, algorithms are assumptions about you – semi-educated guesswork dependent on box-ticking and herd behaviour.
An algorithm can be the key decision-maker over a vast number of decisions that directly affect your life. You could be paying a higher insurance premium than somebody else solely because of a computer-generated risk assessment. Political parties might discount you as a potential voter because they’ve guesstimated you probably lean a different way based on your age and location. As the New Scientist reported in January, an algorithm might even be the deciding factor in how the police treat your case.
The worrying thing is that we have a minimal idea about what information is being fed into algorithms and what their conclusions are. When we are privy to these conclusions, they often don’t make much sense. For example, Twitter has somehow convinced itself that I have a fondness for superbikes, Carling lager and the music of Taylor Swift. Meanwhile, Spotify, who are well aware of my musical tastes, think that I’m some sort of fitness enthusiast despite my own stark realisation that I’ve scarcely moved since my dog died earlier this year.
The Path to Digital Dystopia
Therefore, when I hear of political parties using algorithms to make assumptions about the electorate, I’m not filled with confidence. What I see is the potential for the views of large numbers of people to be discounted based not on what they believe, but what a computer thinks they believe. When I hear of police forces using algorithms to predict crimes, I think about the generalisations being made by a machine that creates a prejudicial justice system. I worry about the chilling effect on freedom of expression if algorithms are used to police the Internet and ‘online harms‘. My thoughts are with those who find themselves at the mercy of the Home Office’s approach to visa applications.
If Muck Rack thinks I’m a journalist, that’s not a big deal. If Spotify thinks I’m far more active than I am, whatever. If Twitter really thinks I drink Carling, it’ll probably just result in a company wasting some marketing moolah targeting their pisswater at me. For others, however, this reliance on the magic of Artificial Intelligence that has taken over our society can have serious implications, and we should all be wary of what the future might bring.