Teaching "AI Scavenger Hunt" to Builder mode (ages 8–10)

Hook & Warm-Up

Teacher: "Time for a real mission! 🕵️ Look around your house — how many things use AI? I bet there are more than you think!" (Let that sit for a second.) "Before we open the app, quick prediction: on a piece of scrap paper or just in your head, guess a number — how many things in your house do you think use AI? Write it down or remember it. We'll come back to it at the end and see how close you were."

This small prediction step matters for Builder age — it turns the activity into something they can be "right" or "surprised" about, which fits the XP/badge structure they already know from the app (this lesson earns the "AI Hunter" badge). Ask a couple of students to share their guess before moving on, without judging any answer yet. You'll likely hear a wide spread — some students guess "one or two," others guess "everything" — and both extremes are useful to have on the board, since the real answer (a meaningful chunk of a home's devices, but far from all of them) sits between the two and gives you a natural way to introduce the idea that intuition about AI is often off in one direction or the other.

Open the lesson and let the hook play, then move into the three-room hunt.

Write the range of guesses on the board as you collect them (a quick tally is enough — you don't need every single number). It gives you something concrete to point back to later, and it turns the eventual reveal into a small class-wide moment rather than something only the reporting pairs experience.

This lesson runs roughly 10 minutes inside the app for Builder age — quick by design, since it's meant as applied practice for ideas already introduced in "AI Is Everywhere" and "Smart vs. Intelligent." Plan to spend at least as much time on the discussion and quiz walkthrough as on the hunt itself; the value here is in students explaining their reasoning out loud, not just clicking through the app quickly.

Main Activity

Have students work through the app's three scenes — Kitchen, Living Room, Bedroom — in pairs or small "squads" if your class already uses that setup (Builder mode's squads are opt-in in the app itself, so not every student will be in one — pairing works just as well for this activity). For each object, students should try to guess AI or not-AI before tapping to reveal the answer, and be ready to explain why in one sentence — not just "yes" or "no," but "yes, because..." This one habit is what turns the hunt from a guessing game into actual practice with the underlying test.

Give each pair a small scoreboard: one point for correctly guessing AI or not-AI before the reveal, and a bonus point if their explanation names the actual reason (learns from data vs. fixed behavior) rather than a surface guess. This lightweight scoring fits naturally with the XP structure students already associate with the app, and it gives you an easy way to spot-check understanding as you circulate — a pair racking up points on the guess but not the reasoning bonus is a sign to stop by and ask a follow-up question.

Kitchen 🍳: Smart Speaker (AI — "Alexa and Google Home use AI to understand your voice"). Toaster (not AI — it just heats bread on a timer, no matter who's using it; ask a pair "would it toast differently for two different people?" — no, which is exactly the point). Smart Fridge (AI — some models use AI to track what food is inside, spotting what's running low). Spoon (not AI — pure, simple engineering, nothing to figure out).

Living Room 🛋️: Smart TV (AI — learns what shows you like and recommends more over time). Game Console (AI — used for enemy behavior and adjusting difficulty; this is the one most pairs get wrong on their first guess, since a console doesn't "talk" the way a speaker does). Couch (not AI — cushions and a frame, holds weight the same way every time). Power Strip (not AI — just distributes electricity, no decisions involved).

Bedroom 🛏️: Smartwatch (AI — detects whether you're walking, running, or sleeping from your movement patterns). Bed (not AI — you do the dreaming, not the mattress). Phone Camera (AI — adjusts focus and lighting and recognizes faces, all before the photo is even taken). Desk Lamp (not AI — on and off, that's the whole job, forever).

After each room, ask: "What's the ONE thing that separates the AI examples from the non-AI examples in this room?" Push students past surface answers ("the AI ones are more expensive" or "the AI ones have screens" — both wrong, since a game console has AI without a huge screen advantage over a TV, and a smart fridge might not look different from a regular one). Land on the real test: does it learn from data and adjust its behavior, or does it do the exact same thing every time?

Close with the app's summary: "AI is everywhere in your home! Smart speakers, phones, TVs, watches — they all use AI to learn your habits and help you. But lots of things work just fine without AI too!" Ask pairs to quickly total their AI vs. not-AI tally across all three rooms (it should come out 6 AI, 6 not-AI) and compare it to their opening prediction from the hook.

The app then moves into a short solo practice round: each student sorts eight of the twelve objects again (smart speaker, toaster, smart TV, couch, smartwatch, power strip, game console, desk lamp) without a partner to lean on. Let this happen quietly and individually — it's the moment where you can actually see whether the test transferred from the paired discussion to independent judgment, which matters more for your sense of the class's understanding than the discussion itself did.

Discussion

These work well as a quick round-robin — call on a few different pairs per question rather than always taking the first hand up, so the conversation reflects more than one squad's thinking. Five to seven minutes is a reasonable target; if a question sparks unusually strong debate, it's fine to let it run a little longer at the expense of one of the others.

• "Why do you think a toaster and a spoon don't need AI, even though people use them just as often as a smart speaker?" — Look for an answer that mentions "fixed job" or "always the same," not just "they're simple."
• "Was there an object in the hunt that surprised you — something you didn't expect to be AI, or didn't expect to NOT be AI?" — The game console and the couch are the two most common answers here.
• "Think about school. Is there anything here — a smartboard, a tablet, a library search system — that might use AI the same way these home objects do?" — This is where students start applying the lesson outside the twelve examples they were given, which is the real goal.
• "If you had to explain the difference between AI and not-AI to your little cousin in one sentence, what would you say?" — A good check for whether a student's understanding is solid enough to teach, not just recognize.
• "Was your prediction from the start of class close to the real number? Why do you think it's easy to underestimate how much AI is around you?" — Tie this back to the idea that AI tends to hide inside ordinary objects rather than announcing itself.

Quiz Walkthrough

Wrap-Up & Extension

Wrap with the tally comparison: "You just found six AI helpers hiding in three rooms of a house you've never even seen. Imagine what you'd find hunting through your OWN house tonight." Award or acknowledge the "AI Hunter" badge they've earned in-app, and let a couple of pairs share whether their opening prediction from the hook landed close to six, or was way off in either direction.

Extension activity (fills the rest of the period): Give students (individually or in their squads) a blank two-column chart labeled "Uses AI" and "Doesn't Use AI." For homework or as a take-home challenge, have them list at least five real objects from their own home under each column, with a one-sentence reason for each. The next class, have squads compare lists and debate any object they disagreed on — this surfaces genuinely tricky edge cases (a rice cooker with a "smart" setting, a karaoke machine that scores your singing) and reinforces the core test rather than memorized examples.

If a squad disagrees about a borderline object and can't settle it, that's a good outcome, not a problem to fix on the spot — write the disputed object on the board and tell them you'll revisit it once the class has covered more about how AI actually works in later lessons. It's more useful for students to notice a genuinely unclear case than to be handed a confident answer for everything, especially since some real products really do sit in a gray zone (a "smart" appliance with one basic sensor and no real learning behind it, marketed with the word "smart" mostly for sales purposes).

For classes that finish quickly, a fast follow-up: have each squad pick their single best "gotcha" example — the one object they think would trip up the rest of the class — and present it in one sentence. Vote as a class on whether it's AI before the presenting squad reveals their answer. This tends to produce a livelier, more memorable close than simply reviewing the homework chart line by line.