Course 3, lesson 21 of 100, Ages 7+

Why apps suggest things

Recommendations, explained

Like I’m 5

If you and your friend both love dinosaurs, and your friend loves a new dinosaur book, you might love it too. Apps guess the same way.

The big idea

When a video app suggests what to watch next, it's using a recommender. It looks at what you watched, liked and skipped, and compares it with millions of other people's choices.

If people who liked the same things as you also loved a new show, the app guesses you'll love it too. Recommenders are useful, but they can also keep showing you more of the same, so it's good to explore on purpose.

Examples

  • Video apps: Watched three dinosaur videos? Expect a fourth.
  • Online shops: 'People who bought this also bought' is a recommender at work.
  • Music apps: Weekly playlists mix songs you love with similar new ones.

How it works

  1. Video and music apps suggest what to watch or play next. That’s AI too.
  2. It notices what you liked, and what people with similar tastes liked, then guesses what you’ll enjoy.
  3. Suggestions can keep you watching for ages. You decide when to stop, not the app.

Check your understanding

How does an app guess what you’ll like?
Options: From what you and similar people enjoyed; It reads your mind; It picks at random.
Answer: From what you and similar people enjoyed. Recommendation systems learn from patterns in what people watch, play and like.
Why do recommenders sometimes show the same kind of thing again and again?
Options: They learn from what you already watched; They are broken; They pick at random.
Answer: They learn from what you already watched. They follow your past choices, so it helps to explore new things on purpose.

Remember

Suggestions come from what you and similar people liked.

Talk about it

Has an app ever suggested something you loved, or hated?

Go deeper

Recommender systems combine collaborative filtering (people like you liked this), content features and engagement signals. Optimising purely for watch time can create filter bubbles and compulsive use, which is why it helps to use the controls apps offer and to shape your own feed.