How Recommendations Work
Have you ever wondered how Netflix always seems to know what you might like to watch? Or how Spotify picks songs you will probably enjoy? That is recommendation AI at work!
How it knows what you like
Recommendation AI pays attention to everything you do:
- What you watched or listened to
- How long you watched before stopping
- What you searched for
- What you skipped or did not finish
Over time, it builds up a pretty good picture of your tastes.
The clever bit: other people like you
The AI also compares you to millions of other users. If lots of people who like the same shows as you also loved a particular movie — there is a good chance you will like it too. This is called collaborative filtering — using what groups of similar people like to make suggestions.
Is it always right?
Nope! Have you ever had a weird recommendation that made no sense? That happens when the AI makes a wrong guess about what your choices mean. Maybe you watched one cooking video and now it thinks you love cooking — when really you were just bored!
Think about it
Is it good or bad that apps know your taste so well? What might be the downside of always watching things a computer picks for you?