AIQ
AIQ
Part of the Finding Patterns lesson guide. Teaching a different grade? 🔧 Builder (8–10) · 💻 Hacker (11–14) · ⚡ Architect (15–18)
Before you open the app, dump a mixed pile of small objects on a table where everyone can see — a handful of toy cars, stuffed animals, blocks, and crayons all mixed together works great, or use whatever you have (buttons, school supplies, mixed candy). Don't sort it yourself.
"Uh oh! Look at this big messy pile! 🧸 Nobody told me how to organize it — there's no sign that says 'put the cars here' or 'put the crayons there.' Do you think we could STILL figure out how to group these, even with nobody telling us the rule?"
Let two or three kids come up and start grouping the pile however makes sense to them — most will naturally put cars with cars, animals with animals, without needing to be told. Point at what they've made: "Look at that! Nobody told you the rule, but you found your OWN groups — cars here, animals there! Guess what — A.I. can do that too!"
Ask the class: "How did you know which things went together, even though nobody told you the rule?" Kids at this age will usually answer something like "because they look the same" or "because they're both cars." Celebrate that answer big — "YES! You looked at how alike two things were, and that's exactly the secret trick A.I. uses too!" This one idea — alike things go together — is the whole lesson, and everything that follows is just this same idea, over and over, in bigger and bigger piles.
Open the lesson on the app and read the hook screen together. Point at the four little icons — 🔮 Discover, 🗂️ Sort, 🔍 Patterns, 🧩 Groups — and say each word out loud with the class before moving on.
If your class has extra time before starting the app, do one more quick round: hold up two of the same kind of object that look almost identical (two red crayons) and one that's very different (a green block). Ask: "Which two of these three belong together?" Every hand should point at the two crayons. Say: "You just decided two things were 'alike' and one was 'different' — that's the exact same job A.I. does, just with way, way more things at once!"
Keep the whole warm-up to about five minutes — the point is just to get every child physically handling and noticing similarity before the screen introduces the same idea with words like "pattern" and "group." If a child gets stuck and can't decide where to put an object, that's a great moment to say "there's no wrong answer here — just tell me why you'd put it where you're thinking," since this lesson is exactly the one place in the app where more than one grouping can be reasonable.
The app walks students through three little scenes. Go through each one together, reading the facts out loud in a big, excited voice and pausing to connect every example back to your toy-pile warm-up.
Read the clustering fact together: A.I. can look at a thousand people and, just like your toy pile, sort them into groups all by itself — "these people are kind of the same!" Show the color sorting example: "If you gave A.I. a HUGE pile of paint colors, it could group all the reds together, all the blues together — just by looking, nobody has to tell it!" Bring it back to something they know: "It's exactly like what you just did with the toys, but A.I. can do it with millions and millions of things, way faster than we ever could."
Mention the shopping fact in the simplest possible words: "Big stores use A.I. to notice that some shoppers like the same kinds of things — like some people always buy toys, and some people always buy snacks — and it figures that out just by watching what people put in their carts, nobody has to tell it the groups." And touch the stars fact lightly, since it's a fun "wow" moment even if the science is far off: "Scientists even use this same idea to look at millions of twinkly stars in the sky and find new star groups nobody had noticed before!"
Say: "Now here's something extra cool. A.I. can even find patterns that are HIDING — patterns we might not notice ourselves!" Read the music genre fact: A.I. can listen to thousands of songs and figure out which ones sound alike, without anyone telling it the genre names. Ask: "Have you ever noticed that some songs just feel kind of similar? A.I. can spot that, like a detective looking for clues!" Mention the fraud detection fact simply: "A.I. can even notice when something looks different from usual — like a helper who says 'hey, that's weird, that doesn't match!'" You can act this out: clap a simple rhythm four times the same way, then clap a very different fifth rhythm. Ask: "Which clap didn't match the others?" Every hand should go up on the fifth one — that's the same kind of "this doesn't match" noticing an A.I. helper does.
Keep this scene very concrete. "How does A.I. know which things go together? It's like a detective — it looks really closely and asks, 'are these two things a lot alike, or not?' Things that are a lot alike get put in the same group, kind of like how your red toy cars all went together." Mention gently that a person still has to tell A.I. how many groups to make: "Sometimes a grown-up tells the A.I., 'find me 3 groups,' and it goes and finds the 3 best groups it can!" Add the last, gentle idea from this scene: there isn't always one single "correct" way to group things — "if I gave you and your friend the exact same toy pile, you might each make slightly different groups, and that's okay! There's often more than one good way to sort something."
Close by reading the app's own summary together: "A.I. can sort things into groups all by itself! It looks at what's similar and puts them together — without anyone telling it how. Like organizing your toy box!"
Give the whole activity 8–10 minutes across the three scenes — this is meant to feel light and playful, not like a science lecture. The badge students earn for this lesson is called "Pattern Finder" 🔮, so it's worth telling them directly: "You're not just watching A.I. find patterns today — you ARE a pattern finder too, right now, with your own eyes and your own toy pile!"
If a child asks something like "does the computer know it's a car?" — a very common question at this age — the honest, simple answer is: "It just notices that some things LOOK alike, kind of like how you noticed the red crayons looked alike. It's not really thinking about what a car IS, the way you can." This plants an early, gentle seed for a bigger idea students will meet again as they get older: A.I. finds patterns without truly understanding what the things in the pattern mean.
A note on running this discussion: at this age, "why" answers will often be very literal ("because they're both red") rather than abstract — that's completely fine and exactly the level this lesson is built for. Resist the urge to push toward vocabulary like "similarity" or "pattern" as a required answer; noticing sameness and difference out loud is the actual skill being built, and the words for it can come later in Builder and Hacker mode.
Close with: "So today we learned that A.I. can be a great sorter, even when nobody tells it the rules — it just looks for things that are alike and puts them together, kind of like you did with our messy pile at the start!" Hold up the class's earlier toy-pile groups one more time and say, "Every one of you found real groups, all by yourselves, with no rulebook. That's exactly the kind of thing that makes A.I.'s pattern-finding so special — it's not magic, it's just really good at noticing what's alike."
Extension activity (10–15 minutes): Give each student (or small group) a small mixed pile of objects — buttons, mixed candy, or a handful of school supplies work well. Tell them: "I'm not going to tell you the rule. Sort this pile into groups YOUR way." After a few minutes, have each group show their groups and explain their own rule out loud: "I put these together because ___." Then ask if another group would have sorted the same pile differently — this gently previews that there's often more than one good way to group things, the same idea A.I. runs into. For a Filipino classroom, a mixed handful of coins (1, 5, 10, 25 sentimo) or a mix of candies like Chocnut and Flat Tops works well and is easy to gather.
If you have a few extra minutes, add a second round with a twist: have each group re-sort the exact same pile a completely different way — by size instead of by type, say, or by color instead of by size. Point out that the objects didn't change at all, only the "rule" for what counts as similar changed, and that gave a totally different set of groups. That's a very gentle, very concrete preview of a big idea: there usually isn't one single "correct" way to group something, just different useful ways depending on what you're looking for.
If a group finishes early, give them a bonus challenge: ask them to try sorting the same pile into just two groups instead of many. Fewer groups forces broader, sometimes trickier categories (like "things I could throw" versus "things I couldn't"), and comparing a two-group sort against their earlier, more detailed sort is a fun, very concrete first taste of the idea that "how many groups" changes what the groups end up looking like.