🔮 Patterns and Predictions

3rd Grade 🤖 Artificial Intelligence

A pattern is something that repeats in an orderly way. People spot patterns all the time — and artificial intelligence is built mostly on this: it finds patterns in data and uses them to predict, which means to make a smart guess about what will probably happen or what the answer probably is.

Free · US Letter · no sign-up
Google Classroom Save to Pinterest Email me

Patterns all around us

Red, blue, red, blue, red… what comes next? Almost any kid will answer “blue.” That’s a simple pattern. But life is full of patterns too: there’s more traffic in the morning than at midnight, some seasons bring more rain than others, and at recess the playground fills up with kids.

When you see dark clouds and say “It’s probably going to rain,” you’re making a prediction based on a pattern you learned from experience. You didn’t see the future — you just remembered that the last times you saw clouds like that, rain came. AI works exactly like this, only with far more data and at lightning speed.

How does a computer predict?

A computer with AI gets a large amount of data from the past and looks for connections in it. For example, a navigation app collects information about the speed of many cars on the roads, at every hour of every day. From that data it learns that on Monday at 8 a.m. the main road into town is jammed, so it can predict how long the drive will take and suggest another route.

The spam filter in your inbox also works by patterns. It doesn’t read the message the way a person does. Instead, it checks whether the message has signs that showed up in lots of spam before — certain words, suspicious links or an unknown sender — and predicts whether the message is probably annoying or dangerous.

The keyboard on a phone does something similar. When you type “Good,” it suggests “morning,” because in lots of sentences it has “seen,” the word “morning” came after “Good.” The keyboard doesn’t know how you feel or what you want to say — it just figures out which word best fits the patterns it learned.

Weather forecasts use patterns too. Measurements of temperature, air pressure and wind from many places go into powerful computers, which calculate what’s likely to happen over the next few days. In recent years, AI has also been used to improve forecasts like these.

A prediction isn’t a sure thing

This is really important to remember: a prediction is an educated guess, not a promise. When a forecast says “70% chance of rain,” it means that in similar situations in the past, it rained most of the time — but not every time. It may not rain at all.

AI makes mistakes too. If something happens that wasn’t in the data — like a sudden accident on the road or a special event in town — the prediction can be wrong. So always use common sense: if the navigation app suggests a road that looks closed, listen to the signs and to the adult who is driving, not just to the phone.

A trap: things that happen together

Sometimes two things happen together, but one doesn’t cause the other. For example, people buy more ice cream in summer, and more people swim at the beach in summer too. If you look only at the numbers, you might think ice cream makes people go to the beach — but really, the heat causes both.

A computer looking for patterns can easily fall into this trap, because all it sees is that the things show up together. That’s why people need to check the findings, think about what the real cause is and not blindly trust every pattern a computer finds.

Try it yourself

For one week, write down every morning whether the sky is cloudy and whether it rained that day. At the end of the week, look at your list: did you find a pattern? Now try to predict tomorrow’s weather, and check whether you were right.

Notice how hard it is to predict with so little data — just seven days. The more examples you have, the easier it is to find a real pattern and not one that happened by chance. That’s exactly why AI systems need huge amounts of data.

Key facts

  • A pattern is something that repeats in an orderly way.
  • AI finds patterns in data from the past in order to make predictions.
  • The keyboard on your phone suggests words based on patterns from many sentences.
  • A prediction is an educated guess, not a certainty — it can be wrong.
  • When two things happen together, that doesn’t mean one causes the other.

Quick quiz

  1. What is a pattern?

  2. Why does a phone keyboard suggest “morning” after “Good”?

  3. How does a navigation app know there will be a traffic jam?

  4. A forecast says “70% chance of rain.” What does that mean?

  5. In summer, people buy more ice cream and more people swim at the beach. What’s true?

  6. When might an AI prediction be wrong?

  7. Why is it hard to predict with only a little data?

  8. What should you do when a navigation app suggests a road that looks closed?

💭 Think and talk about it

  1. Find three patterns that repeat every day at your school. Which of them could you predict ahead of time?
  2. Why do you think a forecast for tomorrow is more accurate than a forecast for two weeks from now?
  3. Think of two things that happen together in your life. Does one cause the other, or is there a third reason?

Questions kids ask

Does AI know what will happen in the future?

No. It can’t see the future — it calculates what is likely to happen based on what happened in the past. That’s why its predictions can be very good, but they’re never certain.

What’s the difference between a prediction and a guess?

A guess can be totally random. A prediction is an educated guess — based on data and on the patterns found in it.

Do people spot patterns too?

Absolutely. Our brains are great at spotting patterns — that’s how we learn to read, recognize tunes and expect what will happen next. AI does something similar in a very different way, using calculations on huge amounts of data.

🔎 Ask Thinkapedia another question