Learning Goals
5 minBy the end of this lesson you can:
- Give two examples of AI you already use in everyday life.
- Explain in your own words what makes a machine "smart" — it learns from examples.
- Tell the difference between a fixed-rules machine and an AI that learns.
Warm-Up · Smart or Not Smart?
8 minThis is our first lesson, so there is nothing to recap. Let's start with a quick sorting game.
Look at these four things. Which ones do you think are "smart", and which just follow fixed rules?
- A pocket calculator
- A torch (you press the button, the light comes on)
- A phone that unlocks when it sees your face
- A video app that always knows what you might want to watch next
Reveal the thinking
The calculator and torch follow fixed rules — same input, same output, every time. They never improve.
The face-unlock and the video app learned from examples — your face, your past choices. That learning is what we call artificial intelligence.
New Concept · What Is AI?
18 minThe everyday-life idea
Think about how you learned to recognise your friend Aisyah. Nobody gave you a rule like "if the nose is 4 cm long, it is Aisyah".
You just saw her many times. Your brain spotted the pattern. Now you know her in a crowd, even in a blurry photo.
Artificial intelligence is when a computer learns in a similar way — from lots of examples, instead of from one fixed rule we type in.
Fixed rules vs learning from examples
A normal program follows rules a human wrote, word for word:
- Fixed-rules machine: "If the button is pressed, turn on the light." It never changes.
- AI that learns: we show it many examples (hundreds of photos of cats), and it works out the pattern of "cat" by itself.
How a machine learns — the big picture
Almost every AI follows the same three-step shape: it is fed data (the examples), it builds a model (the pattern it learned), and then it makes a prediction (its best guess about something new).
This is how your phone groups photos of the same person, how a spam filter spots junk mail, and how an app suggests the next song. Same idea, different examples.
AI does not "understand" the world like you do. It spots patterns in data — and it can be confidently wrong. We always check what it says.
Worked Example · Teaching a Machine to Spot Nasi Lemak
18 minLet's watch the three steps in action with a food example. Imagine Wei Jie wants an app that can tell nasi lemak from roti canai in a photo.
Step 1 — Show examples (data)
Wei Jie collects photos and labels each one:
- 30 photos labelled nasi lemak
- 30 photos labelled roti canai
A label is just a name we attach to an example, so the machine knows the right answer while it learns.
Step 2 — Let it learn the pattern (model)
The computer looks at all 60 photos and works out what makes each dish look different — the round shape and brown colour of roti canai, the rice and little side dishes of nasi lemak. The pattern it builds is the model.
Step 3 — Make a prediction
Now Wei Jie shows the app a brand-new photo it has never seen. The model gives its best guess and a confidence — how sure it is:
What the app shows
New photo → Nasi lemak (92% sure)
Roti canai (8% sure)92% means "quite sure, but not certain". If the photo were blurry, the numbers would be closer together — and the app could get it wrong.
We never wrote a rule for "nasi lemak". We gave examples, and the machine found the pattern. That is the heart of AI.
Try It Yourself
20 minGrab a pencil and your worksheet. These are thinking tasks — no computer needed yet.
List five things you used this week that you think use AI. For each one, write the job it does in a few words (for example: "photo app — sorts pictures of the same person").
Hint
Think about your phone, video and music apps, games, maps, voice assistants, and online shops.
Sort these into two groups — fixed-rules machine or AI that learns from examples: a microwave timer, a face-unlock camera, a traffic light on a timer, a video app's "recommended for you" row, a vending machine.
Hint
Ask one question about each: "Does it improve by seeing examples, or does it always do the exact same thing?"
Pick one AI from task 1. Write one sentence about what examples (data) it probably learned from. For a music app, that might be "songs lots of people listened to after each other".
Mini-Challenge · AI Detective
12 minBe an AI detective. For each scene below, decide AI or not AI, and back it up with the detective's golden question: "Does it learn from examples?"
- Priya's phone suggests the word "Selamat" the moment she types "Sel".
- A school bell rings at exactly 10:30 every day.
- A game character chases Aiman more cleverly each time he plays.
- A lift goes to floor 3 when you press the button labelled 3.
It works if you can say "AI" or "not AI" for all four and give a one-line reason for each.
Show the detective's verdict
- AI — it learned which words usually follow which letters, from lots of typing.
- Not AI — fixed rule, same time every day, no learning.
- AI — it improves from examples of how you play.
- Not AI — a fixed rule: button 3 → floor 3.
Recap
5 minAI is all around us. A machine is "smart" not because someone typed every rule, but because it learned a pattern from examples. Data goes in, a model learns, a prediction comes out — and it can be wrong, so we always check.
Vocabulary Card
- artificial intelligence (AI)
- When a computer learns patterns from examples instead of from fixed rules.
- data
- The examples we feed an AI — photos, sounds, words, numbers.
- model
- The pattern an AI has learned. It is what makes the predictions.
- prediction
- The AI's best guess about something new, often with a confidence score.
Homework · AI Hunt at Home
≤ 20 minFind three things at home or on a family phone that use AI. For each one, write:
- what it is,
- the job it does, and
- your best guess at what examples (data) it learned from.
Bring your list to the next lesson — we will build a class "AI all around us" wall.
Just look and write — no need to sign in to anything. Ask a grown-up before using any app, and never type personal details into it.
Sample · AI Hunt
- Video app "Up Next" — suggests the next clip to watch. Learned from: which videos millions of people watched one after another.
- Phone face unlock — opens the phone when it sees the owner. Learned from: many photos of the owner's face.
- Voice assistant — turns speech into text and answers. Learned from: huge amounts of recorded speech and writing.
Yours will be different — any three real examples with a sensible "learned from" guess are perfect.