We’ve entered the age of the algorithm.
In a way, it was inevitable: thanks to the rise of smartphones and social media, we’re surrounded by vast, unfiltered streams of information, dripped to us via “feeds” on sites like Facebook and Twitter. As a result, we needed something to tame all that information, because an unfiltered stream is about as useful as no information at all. So we turned to a type of algorithm which could help separate the signal from the noise: basically, a set of steps which would calculate which information should be prioritised, and which should be hidden.
Facebook or Twitter are apt demonstrations of this effect. Both collected users and incentivised them to post as much as possible, only to realise that the wash of information was turning off users. So both, at different times and in different ways, introduced curation algorithms to give us what they hoped we wanted: Facebook, in the form of a totally curated newsfeed which, as a Washington Post reporter discovered, often hides over half the posts our friends churn out; Twitter in a small “while you were away” section at the top of the otherwise chronological feed. Google, too, algorithmically selects the most “suitable” posts for you based on user data and location. It’s this that gives it an edge over more privacy-oriented engines like DuckDuckGo or Bing.
Until now, image-sharing app Instagram (owned by Facebook, incidentally) has resisted the draw of the algorithm. But in a short blogpost earlier this month, the company announced that “your feed will soon be ordered to show the moments we believe you will care about the most”. The move seems inspired by the fact that, according to the company, the average user follows upwards of 800 accounts, and therefore may be overloaded by posts which they aren’t interested in at the cost of those that are. As with Twitter, the site fears losing users to a kind of information fatigue.
The reaction? Panic, mostly. Users began posting memes featuring an arrow pointed at the upper right hand corner of the screen, where you can subscribe to a user’s posts. The aim is to get users to subscribe directly, so they receive an invidividual notification to their phone when their favourite Instagrammers post – despite the fact that, as multiple commentators have already observed, this would get incredibly annoying pretty quickly.
But the desperate attempt to load fans onto the subscription lifeboat is understandable. As I’ve written before, many media personalities make partial or entire livings on the image-sharing app, and therefore what may seem like a tiny tweak in the app’s operations has come as a massive blow to many users.
Others, perhaps those who don’t rely on Instagram for their income, kicked back with superior posts like this one, assuring viewers that the algorithm wouldn’t be penalising them:
Of course, @theinfluencersmovement doesn’t know that for sure – none of us do. All we know is that the algorithm will forefront some posts over others, and Instagram claims this decision will be based on…
…the likelihood you’ll be interested in the content, your relationship with the person posting and the timeliness of the post. As we begin, we’re focusing on optimizing the order — all the posts will still be there, just in a different order.
Of course, the implication here is that eventually, some posts may not appear at all. Personally, this annoys me: I use the site so I can see pretty much everything posted by the 200-odd accounts I follow, and I use my chronological timeline to make sure I’ve seen everything posted that day. This new system will make it more difficult to see the new posts when I quickly visit the app. Instagram faces the same issue as all social networks: its users don’t all use it in the same way. (Instagram hasn’t yet confirmed whether users will be able to opt out of the new timeline.)
Then, of course, there’s the question of how exactly the algorithm will know what you’re “interested” in. I’d hope I’d be interested in everything posted by the majority of those I follow, and it’s not always true that I “like” or comment on my favourite accounts enough for Instagram to know which they are. Most users tend to follow a small number of close friends, plus a larger number who post images relevant to their interests. Will the algorithm be able to figure this out? Or will it just forefront sponsored posts and adverts, plus a few big-name users?
It’s impossible to say that algorithms are “good” or “bad”, just as humanity isn’t overridingly either. Algorithms are designed by humans, and therefore carry forward whatever prejudice or bias they’re programmed to perform. As Victoria Turk has written at Motherboard, algorithms can be sexist, because many humans are, too. Facebook’s creepy experiment that manipulated users’ emotions via their timelines showed that algorithms are far from benign, simply because they aren’t designed to be.
Annette M Markham, a communications researcher, argues in her paper The Algorithmic Self that we should view algorithms as “actors” (basically, people) within our social worlds. In fact, they’re more important than any single person you friend on Facebook, because they dicate how and when you see every piece of content on the site.
Perhaps, in an information-sodden online world, we do want an algorithm hand-picking our timelines for us. But first, we need to ask who they, and their creators, are – and what it is that they want.