Skip to main content
Machine Learning & Deep Learning

⏱ About 10 min10 XP

Everyday Predictions

You might think predictions are something only scientists or fortune tellers use. But predictions are hiding inside things you use every single day — and you probably never noticed! Let us go on a little tour of predictions in the real world.

Predictions Are All Around You

The weather forecast is a prediction. Meteorologists — people who study the weather — collect clues like temperature, wind, and cloud shapes. They feed all that information into computers that have studied millions of past weather events. The computer finds patterns and predicts: rain tomorrow, 80 percent chance. Video apps predict what you might like to watch next. Every time you pick a video, the app notices. After many choices, it sees your pattern: you love dinosaur videos and art tutorials. So it predicts you will like the new dinosaur drawing video and suggests it to you. Word prediction helps you type. When you start writing a message and your keyboard suggests the next word, that is a prediction. The phone studied billions of sentences and learned which words usually follow which other words.

The Big Idea

Predictions are built into weather apps, video apps, keyboards, and many more tools. They all learn from past examples and patterns to guess what comes next.

Here is a day full of predictions. Morning: Dad checks the weather app. It predicts sun, so he does not pack an umbrella. Afternoon: You open a reading app. It predicts you might like a new book about space because you loved the last space book you read. Evening: You start typing a message to your friend and your phone predicts the word 'tomorrow' after you type 'see you.' Predictions made your day easier three times — without you even asking for them!

Match each everyday tool to the prediction it makes.

Terms

Weather app
Video streaming app
Keyboard on a phone
Music app

Definitions

Which song will you enjoy after this one?
Which show will you want to watch next?
Will it rain or be sunny tomorrow?
What word comes next in your message?

Drag terms onto their definitions, or click a term then click a definition to match.

All of these tools are making predictions by studying past examples and looking for patterns — just like you learned in the last few lessons. The weather app studied millions of weather events. The video app studied what you and millions of other people watched. The keyboard studied billions of sentences. Big collections of examples make big, useful predictions.

Notice Predictions Today

The next time an app suggests something to you — a video, a song, a word — pause and think: what examples did it probably learn from to make that prediction?

When a video app suggests your next show, what has it been studying?

Your phone keyboard suggests the word 'school' after you type 'going to.' What kind of information did it learn from?

Prediction Spy

  1. For the rest of today, become a Prediction Spy.
  2. Every time an app or website suggests something to you — a video, a song, a word, an article — write it down.
  3. Next to it, write: what past examples do you think it learned from?
  4. Try to find at least three prediction moments in your day.
  5. Share your list with a classmate or family member and compare notes.
  6. Were any predictions surprisingly good? Were any surprisingly wrong?