Ever notice your feed seems to know exactly which content makes you angry, insecure, or unable to look away — and serves you more of it? That’s not paranoia. Recommendation algorithms optimise for engagement, and the content that engages hardest is often the stuff that makes you feel worst: rage-bait, comparison triggers, doom, and endless dopamine-hit clips. Left alone, your feed drifts toward whatever hijacks your attention, not whatever’s good for you.

Here’s the part people miss: the algorithm is trained by your behaviour, which means you can retrain it. You’re not stuck with the feed you have. With a bit of deliberate effort, you can reshape it into something that informs, entertains, or inspires you instead of quietly making you miserable.

Understand what the feed is doing

The algorithm watches what you linger on, rewatch, react to, and rage-comment on — and gives you more of it. Crucially, it doesn’t distinguish between ‘engaged because I love this’ and ‘engaged because this made me furious or insecure.’ To the machine, a hate-watch is a vote for more. That’s why passively consuming without ever pushing back trains it toward your worst impulses.

How to retrain your algorithm

  1. Actively use the ‘not interested’, ‘hide’, and ‘don’t recommend this’ tools — they’re your main lever, so use them constantly.
  2. Unfollow, mute, or block accounts that reliably make you feel worse, even ones you ‘should’ like.
  3. Stop hate-watching and rage-commenting — engagement of any kind tells the algorithm to serve more.
  4. Deliberately engage with the good stuff — follow, like, and linger on content that genuinely informs or lifts you.
  5. Follow real interests and hobbies so the feed has better material to pull from than just outrage.
  6. Do a periodic clear-out — your triggers change, so re-prune every so often.
💡

A hate-watch is a vote

The single biggest mistake is engaging with content that makes you angry ‘just to see’ or to argue in the comments. The algorithm can’t tell love from hate — it only sees that you stopped scrolling. Every second you linger on rage-bait is you asking for more of it. Scroll straight past, or hit ‘not interested’, and starve it instead.

Curate the inputs, not just the algorithm

Who you follow is the raw material the feed works with, so be ruthless about it — the same logic as curating away comparison triggers. Follow accounts tied to your actual interests, mute the ones that spike envy or anger, and remember you can care about someone and still not want their content in your head daily. A feed full of your genuine interests simply has less room for the junk.

⛔

Don’t confuse a better feed with a fixed relationship

Retraining your algorithm makes scrolling less toxic, but it doesn’t address how much you scroll or why. A ‘healthier’ feed you still doom-consume for four hours a day is still a problem. Pair curation with limits — see how to stop doomscrolling and taking a social media break.

Stay in the driver’s seat

The goal isn’t a perfect feed — it’s to remember that you’re supposed to be the one in control, not the machine. Actively shaping what you see, rather than passively accepting whatever’s pushed at you, flips the dynamic. It’s the same instinct behind reducing screen time and protecting your attention as something valuable rather than something to be harvested. Your feed should serve you; make it.

Checklist

Written by Ashutosh Sharma

Frequently asked questions

How do I fix my social media algorithm?
Retrain it through your behaviour, since that’s what it learns from. Actively use ‘not interested’, ‘hide’, and ‘don’t recommend’ tools; unfollow, mute, or block accounts that make you feel worse; stop hate-watching and rage-commenting, because any engagement signals ‘more of this’; and deliberately engage with content that genuinely informs or lifts you. Following your real interests gives the feed better material to pull from. Re-prune periodically as your triggers change.
Why does my feed show me things that make me feel bad?
Because recommendation algorithms optimise for engagement, and the content that engages hardest is often what makes you feel worst — rage-bait, comparison triggers, and doom. Crucially, the algorithm can’t tell whether you engaged because you loved something or because it made you furious or insecure; to it, a hate-watch is a vote for more. Left unchecked, your feed drifts toward whatever hijacks your attention rather than what’s good for you.
Does hate-watching content make my feed worse?
Yes. Engaging with content that angers you ‘just to see’ or to argue in the comments tells the algorithm to serve you more of it, because the machine only registers that you stopped scrolling and interacted — it can’t distinguish love from hate. Every second lingering on rage-bait is effectively asking for more. The fix is to scroll straight past or hit ‘not interested’, starving that content instead of feeding it.
Can I really control what my algorithm shows me?
To a meaningful degree, yes. The algorithm is trained by your behaviour, so you’re not stuck with the feed you have. Consistently using the feedback tools, curating who you follow, refusing to engage with content that upsets you, and deliberately interacting with better material will reshape it over time into something that informs, entertains, or inspires you rather than quietly making you miserable. It takes ongoing effort, but you can shift it.
Is fixing my feed enough, or do I still need to cut down?
Curating your feed makes scrolling less toxic, but it doesn’t address how much you scroll or why — a ‘healthier’ feed you still doom-consume for hours a day remains a problem. Pair curation with limits on your usage, such as reducing doomscrolling and taking periodic social media breaks. The aim is to be in control of both what you see and how much time you give it, rather than the machine harvesting your attention.

People also ask