Erica: Are you sick of your internet algorithm? Then why not take matters into your own hands?
Hello, I’m Erica McKoy and welcome to Which? Shorts - your free weekly insight into Which? magazine, as well as our Money, Tech, Travel and Gardening titles too.
Today I’m bringing you a piece that our tech expert Nicole Kobie wrote for the August/September issue of Which? Tech magazine, all about algorithms.
These are the internet machines that learn what you like - or at least engage with - online. They then process that information to help create the output that you then see.
But as a chronological feed of your friends’ posts seems to be a thing of the past, and a ‘For You’ page designed to keep you scrolling for longer now seems to the norm, we explain how to take control of your internet algorithm
Here’s Nicole’s piece, adapted for the podcast, this week read by myself, Erica McKoy.
How does the vast richness of the online world get filtered down to the fragments that appear on your screen? It’s not always up to you to choose; instead, it’s down to algorithms. These decide what news stories you see, what posts hit your social feed, and even what music you hear. These recommendation machines are obsessed with engagement, not what’s good for us.
Search for a sensible weight-loss plan and you’ll be inundated with dodgy diet pill ads. Love a specific TV show? And a Google News visit risks spoilers by showing recaps.
On social media you’re no longer shown friends’ posts in chronological order. Instead, algorithms decide the content of your feed using ‘signals’ such as your friends’ activity, accounts you’ve ‘liked’, and interactions with different posts. The longer a topic keeps you engaged, the more likely you are to be shown that again, in order to keep you hanging around to view ads.
And that’s sparked problems.
Engagement causes smartphone addiction and doomscrolling on topics that upset us, and traps us in echo chambers of our own opinions, where we’re recommended one-sided content that triggers emotional responses – anger is great for engagement, after all.
These complex systems are difficult to see, let alone understand, but you aren’t wholly powerless. Algorithms interpret the signals we send – your clicks, likes and even how long you look at part of a webpage – and that means we can influence the results. Here’s how to dodge the algorithms and take back control of what you see online.
Start with privacy settings.
For example, Google Chrome - the world’s most-used browser - has a ‘privacy guide’ that explains key controls to customise ads and disable third-party cookies. Go further by selecting ‘Settings’, then ‘Delete browsing data’ to remove all of your internet history. You’ll sharp see a difference in everything from search results to ads.
As you continue with a clean account, consider what you’re telling the algorithm. Want to see more of your friends on Instagram? Open their pages directly and like their posts when shown in your feed. Find suggested posts from influencers annoying? Click to say you’re ‘not interested’.
Think of it like training a pet: a bit of consistent feedback now will pay off in the longer term. But unlike a new puppy, you can always delete your account if the effort becomes tiresome.
You could always try to find a better way to search.
Try anti-tracking alternatives to Google such as brave.com and duckduckgo.com. If you want to stick with Google, try startpage.com, which acts as an intermediary so results are based on your search, not your past browsing behaviour. That said, if you’re trying to find an article you read last week, Google’s tracking will be handy.
The same follows for browsers.
Choose one that blocks third-party trackers, such as Brave or Firefox, or install anti-tracker extensions that also ban ads, such as uBlock Origin. You can even rearrange the web: the News Feed Eradicator extension removes algorithmic feeds from social media.
RSS feeds collect updates, such as news, in one place and are a way to find something to read without algorithmic influence. Choose a tool such as Feedly or Inoreader, then add feeds from your favourite sources or topics. Not all sites have RSS, but tools like Fetch RSS can generate them for you.
However, dodging personalisation means you’ll need to put effort into curating your own content. You may miss material you’d like to see, and see ads that aren’t relevant – which may well be good for your bank account.
Don’t think you have to be nice to your algorithm - you can confound them instead.
Data poisoning sounds very dramatic, but could be as basic as using a fake birthday when registering an account. By giving these systems false data, you skew their algorithms.
Examples include clicking random ads, following an account that bores you, or going into settings and selecting ad topics or locations that don’t apply. This works best if you’ve cleared out your data history first, as otherwise those signals may appear as outliers and be disregarded.
Automate your efforts with the AdNauseum browser extension, which follows all the ads on a page to build up digital noise. Users see no ads but the system ‘clicks’ them in the background. TrackMeNot does a similar trick, randomly running searches to obfuscate your real ones.
Once you’ve worked through this for every website, app or service that uses algorithms in this way, they will again start gathering data to make decisions about what to show you.
Unfortunately, this is now the start of a lifetime of digital hygiene. It needs to be easier.
Digital companies that feed our data into algorithms for personalisation and behavioural advertising should make it simpler for us to start afresh. This would also benefit web companies by ensuring they’re serving up what we actually want to see.
Instead, we must become experts on training algorithms just to use the web as we’d like – but that gives us more power, at least. And that’s a start to winning back the web we want.
