Teaching "AI Is Everywhere!" to Builder mode (ages 8–10)

Hook & Warm-Up

Builder mode still has a mascot — Axiom — but the tone shifts from the wide-eyed wonder of Explorer mode toward something closer to a friendly challenge. Open with a straight-up claim and let students push back on it. Read the app's hook line as written, in Axiom's voice:

"Hey there! I bet you used AI at least 5 times today and didn't even know it. Ready to explore where AI is hiding?"

Then challenge the class directly: "I'm betting five things. Anyone want to bet me it's fewer than that?" Give them 60 seconds to jot down (or call out) anything from today they think might count — their alarm, a game, their phone. Don't confirm or deny answers yet; tell them the app is about to check their guesses.

Quick framing before you open the lesson: "Today we're building a simple test you can use on ANY piece of technology for the rest of your life: does it just do one fixed thing, or does it learn and get better from data? That's the whole question."

This works well as a two-minute private write followed by a public tally — have students write their guessed count on a sticky note or scrap paper before anyone shares out loud, so louder students don't anchor everyone else's guess. Collect a quick show of hands for "more than 5," "exactly 5," and "fewer than 5" before opening the app, and revisit that tally at the end of the lesson once the real answer is much higher than most students expected.

Main Activity

Open the lesson and work through its three time-of-day scenes as a class, but push a level deeper than just "AI or not" — for each item, also ask what data it would need to learn from. This "what data" question is the single most useful habit to build at this age: it turns a yes/no label into a small piece of real reasoning, and it's a question students can keep asking about any new technology they meet for the rest of their lives, long after they've forgotten the specific examples from today.

Morning 🌅: ⏰ Smart Alarm (AI — it understands spoken commands, which needs a huge dataset of recorded speech to learn from), 🚿 Shower (not AI — "pipes and valves, no AI needed," straight from the app), 📱 Face Unlock (AI — trained on many photos of faces so it can tell yours apart from anyone else's), 💡 Light Switch (not AI — a plain circuit).

School 🏫: 📝 Spell Check (AI — learned the difference between "their" and "there" from huge amounts of real writing), 📚 Paper Book (not AI — ink on paper), 🎮 Game Levels (AI — adjusts difficulty by tracking how well you're doing), ✂️ Scissors (not AI — two blades and a pivot).

Evening 🌆: 🎵 Song Picks (AI — Spotify's system finds patterns in millions of listening habits), 📺 Video Ideas (AI — YouTube predicts what you'll watch next), 🚲 Bicycle (not AI — "pure mechanics, YOU are the smart part," per the app), 🗺️ GPS Maps (AI — predicts traffic to route you around it).

Land on the app's own summary line and unpack it word by word: "AI means a computer that learns from experience — like how you learned the stove is hot. AI does something similar, but with data!" Ask: "What's the 'stove' for a music app?" Guide toward: the millions of songs people have played and skipped is its "stove" — its experience. Push one step further if the class is following well: "You only need to touch a hot stove once to learn the lesson. Why does an AI system usually need thousands or millions of examples instead of just one?" There's no need for a precise technical answer here — the point students should land on is that a single example is too easy to be a coincidence, while a pattern that holds across huge numbers of examples is much more likely to be real and reliable.

Have students do the in-app practice round (ten more items) individually, then run a quick show-of-hands check on any they disagreed on as a class. The round mixes clear cases with genuinely tricky ones on purpose: Siri/Alexa, Photo Filters, Smart Thermostat, Netflix Picks, Auto-Translate, and Spam Filter are all AI, while a Broom, Paper Clip, Metal Key, and Ruler are not. Smart Thermostat and Auto-Translate are the ones worth pausing on, since they're less obviously "smart-looking" than a chatbot — a thermostat is easy to mistake for "just a dial," but it's learning your temperature preferences and schedule over time, and a translation app isn't looking words up in a dictionary but predicting the most likely translation from patterns across millions of translated sentence pairs.

Discussion

• "Was your guess from the warm-up right — did you actually use AI more or fewer than 5 times today?"
• "Pick one AI example from today's lesson. What do you think the 'data' behind it actually looked like — what was it trained on?"
• "Two students use the same music app but get completely different song recommendations. What does that tell you about how the app works?" (It's learning from each listener's individual history, not applying one fixed rule to everyone — a good check on whether the "learns from data" idea has really landed.)
• "Why do you think a calculator isn't AI, even though it's definitely 'smart' at math?"
• "Can you think of something that LOOKS high-tech but isn't AI, and something that looks ordinary but secretly is?"

Let students argue with each other on these, not just answer you — the "looks high-tech but isn't AI" question especially tends to produce genuine disagreement (a digital thermometer is high-tech and precise but isn't AI; a basic weather app just displaying a forecast number isn't AI either, even though the forecast model behind it, somewhere on a server, likely is). You don't need to resolve every edge case perfectly; the goal is students practicing the reasoning, not memorizing a complete list.

Quiz Walkthrough

Notice that all four quiz questions in this age band circle back to the same single test in different clothes — that repetition is deliberate, both in the app's design and in this walkthrough. By the fourth question, most students should be answering from the underlying rule rather than pattern-matching against the exact wording of the first question they saw.

Wrap-Up & Extension

Close with: "You now have a real test for spotting AI: does it just repeat the same fixed action, or does it learn from data and change its behavior based on that? Keep using that test — you'll start noticing AI in places you never thought to look."

It's worth also naming, briefly, where this lesson stops. Today was entirely about recognizing AI, not judging whether any particular use of it is good or bad, fair or unfair, or something students should be worried about. Those questions matter and the curriculum comes back to them directly later (privacy, fairness, and misinformation each get their own lessons) — but layering that in today would blur the one clean skill this lesson is building. If a student raises a worry ("is it bad that YouTube tracks what I watch?"), it's fine to say plainly: "That's a really good question, and we're going to spend a whole lesson on exactly that later — for today, let's just get good at spotting where AI is."

Extension activity — App Audit: In pairs, have students list every app on a family phone or tablet (or a printed list you provide of common apps: YouTube, Waze, Duolingo, a camera app, a calculator app, a weather app, a games app). For each one, they decide "AI" or "not AI" and write one sentence justifying it using the learned-from-data test. Bring the class back together and debate any apps pairs disagreed on — weather apps are a genuinely interesting edge case (the forecast model itself is AI-adjacent, but the app just displaying a number isn't). This turns the 10–15 minute core lesson into a full period with a built-in class discussion.

To make the audit even more concrete, have each pair pick their single most confident "AI" verdict and their single most confident "not AI" verdict, and write both on the board under two headings as the class works. By the end you'll have a growing, visible list the whole class built together — a good artifact to leave up on the wall for the next few lessons in World 1, since several of them (Lesson 1.1.2 "Smart vs. Intelligent" especially) build directly on the same distinction.