Everyday AI
A recommendation system is AI that predicts what you will want next, such as a video, song, or product, based on your behavior and that of similar users.
Every 'you might also like' is a recommendation system at work. It watches what you click, watch, skip, and buy, compares you with millions of other people, and ranks what to show next. It is the most common AI you interact with, often without noticing it is there.
Two ideas do most of the work. One finds people whose history looks like yours and suggests what they enjoyed. The other looks at the items themselves and suggests things similar to what you already liked. Modern systems blend both, using embeddings so that users and items live in the same space.
Recommendation systems shape culture and attention. They are tuned to keep you engaged, which can mean surfacing outrage or endless scrolling. Critics call this 'the algorithm.' Understanding that it is a model optimizing a goal, usually watch time or purchases, helps you use it on your own terms.
Spotify's Discover Weekly playlist suggests songs you have never heard, chosen because listeners with tastes like yours played them on repeat.