Y07W25RC The Algorithm Behind You

What shows up in your feed can feel random, but it usually is not. In this reading, you will explore how recommendation systems notice patterns and use them to shape what people see next. You will also look at how that can influence attention, choices and even beliefs over time. As you read, notice which small actions turn into bigger effects.

Informative — Explanation text

An explanation text is a piece of writing that shows how something works or why something happens. Writers use it to inform you by breaking a process or idea into clear parts so you can follow the mechanism step by step. You will often find definitions, examples, linked stages, headings and sometimes lists or scenarios that make an abstract idea easier to understand. As a reader, you need to track how one part leads to the next, connect the examples to the main idea and notice the consequences that come from the system being explained.

Before You Read

  • Read the title and headings carefully so you can predict that the text will explain a hidden process behind everyday digital experiences.
  • Think about how your feed often seems to 'pick' things that match what you have clicked, watched or searched for before.
  • Expect the article to move from a simple hook into a clearer step-by-step explanation, with a list and a scenario to help make the idea concrete.

While You Read

  • Pause after each section and check what new part of the recommendation process has just been explained.
  • Use the headings, the 'signals' list and the scenario as reading aids to organise the explanation into smaller, clearer parts.
  • Track how the text moves from what recommendations are, to the signals they use, to the feedback loop and its effects.
  • Pay attention to key words such as 'signal', 'loop', 'bias' and 'recommend', and use the surrounding sentences to sharpen their meaning.
  • Re-read any sentence that explains how repeated actions can change what appears next, because those links carry the main mechanism.

Read With Purpose

  • Notice how recommendation systems gather clues and turn them into predictions.
  • Pay attention to the way feedback loops can shape what people see more often and less often.
  • Look for the consequences this process can have for attention, beliefs and digital literacy.

Now read

The explanation text

~5 min read · ~939 words

Why Your Feed Knows You

Open a video app, a music app or a news page, and it can feel as if the screen somehow knows what you want next. One clip finishes, and another appears that seems strangely well chosen. One article about a topic you clicked yesterday is followed by three more today. This can feel mysterious at first, but the system is not reading your mind. It is responding to patterns in your behaviour.

What Recommendations Are

A recommendation system is a digital system that tries to predict what a person might want to watch, read, listen to or click next. Its job is not to understand you as a friend would. Its job is to sort through huge amounts of content and make a guess about what will keep your attention.

That guess is based on data. If many people who liked one kind of video also watched another kind, the system may connect those two things. If you pause on a clip, replay it, finish it or share it, the system may treat those actions as clues. In other words, a recommendation is not random. It is a prediction built from signals.

What Counts as a Signal

A ‘signal’ is a piece of information that helps the system make a choice. Some signals are obvious, and some are small.

Signals can include:

  • what you click
  • how long you watch
  • what you skip quickly
  • what you search for
  • what you like, save or share
  • what time of day you are active
  • what kinds of posts you return to

Not every signal matters equally. Watching a full video may count more strongly than glancing at a thumbnail. Searching for the same topic several times may matter more than one accidental click. The system is always weighing these clues and asking, ‘What does this person seem interested in right now?’

A Simple Scenario

Imagine a student named Noor opens a video app after school and watches one short clip about basketball skills. Then Noor watches another about game highlights, pauses on a training video and saves a post about shooting practice. The next day, the app offers more basketball content. This does not happen because the app ‘knows’ Noor personally. It happens because Noor’s actions sent strong signals.

Now imagine Noor also clicks one video about sports injuries for a health assignment. If that is a one-off action, the system may test a few similar posts and then move away from them if Noor scrolls past. But if Noor keeps watching them, the system may begin to recommend more of that topic too. The feed keeps adjusting because it is built to respond to what seems to work.

The Feedback Loop

This creates a ‘feedback loop’. A loop is a cycle that repeats. The system recommends something. You react to it. Your reaction becomes new information. Then the system changes what it shows next.

That loop can become stronger over time. If you are shown more of one topic, you may click more of it simply because it is there more often. Then the system treats those extra clicks as proof that the topic matters even more. This does not mean the system is forcing your opinion. It means the system is shaping your chances to see some things often and other things rarely.

Because of this, your feed may become narrower without you planning it. You might start with a broad interest in sport, music or current events, but the system may keep pushing one corner of that interest because your behaviour seemed strongest there. The result can feel natural, even though it has been shaped step by step.

Effects on What People See and Believe

Recommendation systems can be useful. They can save time, surface content that matches your interests and help you discover things you might genuinely enjoy. A good recommendation can feel convenient, even helpful.

But there are also limits. If a system keeps showing similar content, it can create a slanted picture of what is common, important or true. This is where ‘bias’ matters. Bias here does not always mean someone is lying. It can mean the system is leaning in one direction because of the signals it receives and the way it ranks content.

For example, if dramatic posts get more attention, the system may recommend more dramatic posts. If strong opinions keep people watching, those may appear more often too. Over time, a person might start to think, ‘Everyone thinks this,’ or ‘This issue is everywhere,’ when really they are seeing a filtered stream rather than the full picture.

Why Digital Literacy Matters

Understanding recommendation systems does not mean being scared of them. It means being alert. A feed is not a neutral window showing everything evenly. It is a personalised selection shaped by signals, predictions and repeated feedback loops.

This matters because what people see often can influence what they notice, what they return to and sometimes what they believe. If you understand that process, you are in a better position to pause and think. You can ask, ‘Am I seeing this because it is the whole story, or because the system has learned I am likely to click it?’

Summary

Your feed seems to know you because it watches for signals and uses them to recommend what comes next. Those recommendations then create a loop: what you see affects what you do, and what you do affects what you see. This can be useful, but it can also narrow your view and strengthen certain patterns. The more clearly you understand that mechanism, the more thoughtfully you can use the digital spaces around you.

Check your vocabulary knowledge

recommendation n.
a suggested item chosen for you by a system
signal n.
a clue the system uses to make a prediction
predict v.
make a careful guess about what will happen next
feedback loop phr.
repeating cycle where responses shape future results
bias n.
a tendency to lean more in one direction than another