Netflix, for me, used to be a place to discover new and exciting content or, since they’ve got quite the back catalogue, a way to catch up on movies and series that I missed the first time around. Now, it’s a place where I scroll around aimlessly for a while before, more often than not, watching something I’ve already watched before.
And the problem isn’t with me… okay, it’s a little with me… the issue comes in with how the streaming service monitors every single move I make. The reason for this, besides informing them what combination of genres, scripts and actors are likely to result in a hit film or show, is to try and keep me watching Netflix for as long as possible.
The Netflix algorithm wants to take over my (and your) entertainment choices. On paper, it does this better than you do. After all. It knows all the films, it knows what other people who like the films you like have also enjoyed, and that little profile it has built of you? Yeah, it really, really works. For a while. And then it stops working.
Not only Netflix
It’s not just the video streaming service that fails to do what it claims. All of the algorithm-powered services — Google, Facebook, Twitter, Spotify, Amazon — all of them are at the mercy of their own algorithms. And so are you.
The easiest way to see this in action is by messing with Amazon. Log in to your Amazon account, choose a weird search term (like, say, pooper scoopers). Something that you’ve never searched for before. Add one to a wish list, just for fun. Do this for a couple of days, and watch how long it takes Amazon to send you an email (depending on your communication settings) recommending all sorts of dog excrement disposal tools.
Dear Amazon, I bought a toilet seat because I needed one. Necessity, not desire. I do not collect them. I am not a toilet seat addict. No matter how temptingly you email me, I'm not going to think, oh go on then, just one more toilet seat, I'll treat myself.
— Jacqueline (Jac) Rayner (@GirlFromBlupo) April 6, 2018
And this only gets worse if you actually buy something unusual — Amazon’s algorithm is older than most so its flaws are really obvious. It only really knows you want something once you already have it, so it’s going to keep on suggesting items it (finally!) knows you like — even if there’s no sensible reason for you to want more of them.
Swimming in circles
Entertainment services and social media have a better time of things, though Facebook’s always going to use a shotgun approach to getting you to buy something (until they get that whole Facebook currency thing off the ground). But even here, the algorithms governing the suggestions you see are doomed from the start.
Take Spotify (because why only pick on the video streaming guys, right?). When you start using it, you’ll likely start listening to music you know. Your profile builds and you’re introduced to similar stuff. You like those tunes, so you listen to more of that. Spotify’s net tightens over the space of a couple of years, until you’re encased in a rigid shell of music genres that you used to listen to. And that shell is hard to pierce.
Prolonged exposure to a suggestion algorithm always only ever leads in one direction — stagnation. You’ll eventually run out of new things like the things you like, and algorithms don’t compensate well enough to start suggesting new content without pissing you off. Death metal fans are only going to be confused if Spanish folk music turns up on their Discover Weekly (this actually happened). But once you’ve gotten to this point, you’ve got your work cut out for you.
Breaking the fish bowl
Eventually, you are swimming in a digital fish bowl. You’ll only see the movies you like (and movies like them), the music you like (and music like it), the news you’re prone to reading (rather than what’s important), and suggestions for products you already own and don’t need to replace. Essentially, you’re going to be bored. Finding new and exciting things to experience becomes exponentially harder when you’re fighting against an algorithm that wants to engage your attention all the time.
The services we use want us to use them, but they only know what we’ve told them — so past a certain point it’s only ever more of the same. Eventually you get to a point where you’d have greater success scrolling through an unsorted list of, say, comedies. Only that’s been optimised too, so you’ll see the ones Netflix reckons you’re most likely to enjoy first. And there’s no way to go back to an unordered list, because that’s not ‘efficient’, apparently.
In order to get anywhere against the algorithm, you have to already know what it is you’re looking for. Exploring for something new eventually calls for extreme action — a new Netflix profile, deleting your Spotify and starting over, deleting Facebook… no, that’s it, that’s the whole thing… or signing up for a whole new Gmail/YouTube account.
Either that, or you’ve got to set out to retrain the algorithm making all of your choices for you. Had an ill-advised flirtation with nightcore on YouTube, or watched a few too many reality TV shows on Netflix? You’ll have to do some pruning, and even then you’re likely to keep getting suggestions for something you don’t care about for years afterwards.
The point is: all of those lovely services that take the work out of your hands will eventually collapse on you, trapping you in a bubble of your own making. They don’t do it on purpose, it’s just a natural limitation of current algorithms. Breaking out of it is much, much harder than constructing it, unless you just burn it all down and start over. Next time you’re frustrated with the suggestions being put in front of you, spare a little thought for the computer brain that is only able to suggest what it already knows you like. You’ve probably both hit your saturation point.
