AI in everyday lifeBeginnerLesson 844 min read

Why your feed shows what it shows

Videos, posts and shopping suggestions are chosen for you by AI. Learn how it picks, why it can trap you, and how to take back control.

✦ Ask the tutor🧪 Lab · 20 min📝 Lesson test
Lesson in motion

In 60 seconds

Why your feed shows what it shows

Videos, posts and shopping suggestions are chosen for you by AI. Learn how it picks, why it can trap you, and how to take back control.

1/6
In simple words
Imagine a shopkeeper who watches everything you look at. If you stop at the sweets, next time the shop is full of sweets. That is nice at first. But soon you never see fruit at all. Your feed works a bit like that shopkeeper.
Every time you open a video app, a social app or a shopping app, there are far more things to show you than fit on your screen. A recommendation system is the AI that picks what goes at the top.

What it watches

  • What you watch, and for how long
  • What you like, share, comment on or save
  • What you skip quickly
  • What people similar to you enjoy
  • Sometimes your location, your language and the time of day
From these signals, it guesses what will keep you watching, clicking or buying. Its goal is usually set by the company, for example "more time in the app". That goal is not always the same as your goal.
  1. 1

    You watch

    You watch three cricket clips to the end.
  2. 2

    It learns

    The system notes: this person likes cricket clips.
  3. 3

    It shows more

    Your feed fills with cricket, and with clips that got strong reactions.
  4. 4

    The loop grows

    You watch more because it is there. The system becomes even more sure.

The filter bubble

When a feed shows you only what you already like or agree with, you can end up in a filter bubble. You may start to believe "everyone thinks like me", because you rarely see other views. Strong emotions like anger and fear often get more reactions, so upsetting posts can travel further.
Watch out
A post being on your feed many times does not make it true. It only means the system thinks you will react to it.

Taking back control

Letting the feed decide
  • Scrolling whatever comes next
  • Watching angry posts to the end
  • Only following people who agree with you
Steering your feed
  • Using search to find what you actually want
  • Using "Not interested" or "Don't recommend" on things you don't want
  • Following a few trusted sources with different views
Do this
Many apps have options like "Not interested", "Hide", "Mute", "Following" feeds, watch history controls and screen-time reminders. Names differ between apps, but most have some of these.
Try this. Open your most-used video app. For the next five things you don't want, tap "Not interested" (or the app's similar option). Check your feed again tomorrow and see what changed.

Watch and read more

Lab

A one-week feed experiment with notes on what changed.

~20 min

The problem

Pick one app you use every day.

  1. Write down the topics of the first 10 items on your feed today.
  2. For three days, use "Not interested" on things you don't want, and search for two topics you want to see more of (for example a skill you want to learn).
  3. On day four, write down the first 10 items again.
  4. Compare the two lists. What changed? What stayed the same?

You are done when

Hard questions

Try to answer before you reveal. If you can answer these, you understood the lesson.

Q1If a feed is designed to keep you watching, why might it not show you what is best for you?Reveal

Because "keeps you watching" and "good for you" are different goals. A shocking or upsetting clip can hold attention longer than a helpful one. The system is not evil; it is chasing the number it was told to chase. When the measure becomes the goal, the results can drift away from what people really want.

Q2Two neighbours search the same news topic and see very different results. Give two reasons.Reveal

First, their past behaviour is different, so each feed is personalised. Second, location, language and device settings can change results. This is why "I saw it everywhere" can mean "my feed showed it everywhere", not "everyone saw it".

Please sign in to continue.

Questions people ask

Is my phone secretly recording me to choose ads?

This is a common worry. Apps can learn a lot about you without a microphone: your searches, your location, what you watch and buy, and what similar people do. That is often enough to make a suggestion feel "too perfect". You can check which apps have microphone permission in your phone's Settings.

Can I reset my recommendations?

Many apps let you pause or clear your watch history or search history. After that, the feed often becomes more general for a while and then learns again from what you do next.

Are recommendation systems bad?

Not by themselves. They help you find useful videos, songs and products among millions. The problem comes when their goal (your time and clicks) quietly takes over your own goals.

Why do I see the same fake news again and again?

Because many people reacted to it, and the system rewards reactions. Seeing something often is not proof. Check it with a trusted news source or a fact-checking site before you share.

Lesson test

5 questions. Get 3 right (60%) to pass and complete this lesson.

Sign in to take the test and save your progress