Recommendation algorithms — Media, 14–17 years
See how platforms choose what to show next, using signals from your behaviour, and why that can narrow or shape your view.
A feed is selected
A platform does not simply display everything in time order. Software estimates what may keep your attention by using signals such as views, likes, follows, searches and watch time. It then ranks posts, videos or songs, so two people can receive different feeds.
The problem of too much content
There are more posts, videos and songs than anyone can inspect. Recommendation systems were built to help people find something quickly and to help platforms keep attention. Their goal is usually relevance or engagement, not a balanced picture of the world.
How one signal changes a feed
You watch three videos about repairing bicycles and finish each one. The system records strong interest in that subject. Next, it may place more bicycle videos near the top, while showing fewer cooking videos. This is not mind-reading: it is a prediction based on your recent actions.
A personalised feed is not the whole world
If a topic appears repeatedly, it can feel as if everyone is discussing it or agreeing with it. That feeling is reasonable because your screen really is full of it. But the feed reflects a selection shaped by data and design, not a random sample of all people or facts.
Making your information diet wider
When you need to understand an issue, do not rely on one feed. Search deliberately, follow different credible viewpoints and check what the system did not show you. You can also use controls such as “not interested” or chronological views, where they are available.
Keep exploring
Other languages
Loading MyLeoNes™…