AIQ
AIQ
Part of the Smart vs. Intelligent lesson guide. Teaching a different grade? 🌈 Explorer (5–7) · 💻 Hacker (11–14) · ⚡ Architect (15–18)
Open with a genuine challenge. Write "47 × 823" on the board and ask a volunteer to solve it by hand while another student solves it on a calculator. The calculator wins by a landslide, usually before the first student finishes carrying a digit.
"Here's a puzzle: a calculator can solve math faster than any human alive. So is it smarter than us? 🤔 The answer might surprise you. Let's find out what 'smart' really means!"
Take a quick show of hands: who thinks the calculator is smarter than us right now? Don't resolve it — tell them the app is going to help settle the argument, and that the answer is more interesting than a simple yes or no.
Point out the four hook icons — 🧮 Calculator, 🧠 Your Brain, 🤖 AI, 🔍 Search — and frame the lesson: "We're about to compare all four of these and figure out what actually separates a calculator from real AI, and what your brain does that neither one can."
If your class already completed "AI Is Everywhere!" (lesson 1.1.1), call back to it briefly: "Last time, we learned to spot AI hiding in things you use every day. Today we go one level deeper — it's not just about whether something IS AI, it's about what actually separates something 'AI' from something that's just fast, or from what YOUR brain does. Same detective skills, sharper tools."
Work through the app's three scenes with the class, but push each example one level deeper than Explorer mode — ask "why" after each verdict, not just "AI or not AI."
Calculator and stopwatch: fast, but each one runs the exact same formula every single time, no matter how many times you use it. Contrast this with a search engine: "Google looks at billions of past searches to figure out what results actually help people — it's learning a pattern, not running one fixed formula." Same logic for a chatbot: "It has processed millions of real conversations and picked up patterns in how people talk, which is why it can hold a conversation that feels natural even though it's never truly 'thought' about anything."
A good check-for-understanding question here: "If I gave the calculator a MILLION more math problems to solve, would it get any better at math?" Students should land on no — it would just do the millionth problem exactly the same way as the first, because nothing about how it processes numbers changes. Then flip it: "If a search engine got a million more searches, would its results get better?" Yes — because that's data it can learn patterns from. That contrast, run twice with concrete numbers, is the whole lesson in miniature.
A Filipino example that lands well here: a jeepney's fare board versus a map app predicting how long your commute will take. "The fare board follows one fixed table printed by the government — it never changes based on anything. But a map app's traffic prediction has learned patterns from huge amounts of past trip data, which is why it can guess that EDSA will be worse at 5pm on a Friday than on a Sunday morning. Same basic job — give you a number — very different mechanism underneath."
Ask who has used Shazam. "It learned the 'fingerprint' of millions of songs, so now it can recognize almost any tune in seconds — that's a skill it built from data, not a rule someone programmed by hand." Compare to radio: "A radio just receives whatever signal is sent — zero learning involved." Then a face filter: "It had to learn what a human face generally looks like — where eyes, nose, and mouth usually sit — before it could reliably put a filter on yours." Compare to a flashlight: pure circuit, no intelligence needed.
Have students try to stump each other: in pairs, one student names a gadget or app, the other has ten seconds to argue AI or not-AI and give one reason. Encourage genuinely tricky picks — a dishwasher with a "smart" sensor that just measures water cloudiness with a fixed threshold is a good trap, since "smart" is in the marketing name but nothing about it is learning from data.
Push a little further with Shazam specifically, since it's a strong example for this age group: "Shazam has never heard the exact song you play for it recorded in a noisy classroom before — but it can still usually recognize it. Why do you think that is?" Guide toward: it learned the general pattern of what that song sounds like, not just one perfect recording — which is why it can recognize a slightly muffled or noisy version too.
This is the section worth the most class time. Ask directly: "Can AI feel proud of a good grade the way you can?" Guide toward: no — AI has no inner experience; whatever confident-sounding text it produces is not evidence of feeling anything. Ask: "If an AI draws you a picture, whose idea is it?" Guide toward: it's recombining patterns from images it has seen, not having a truly original idea the way a person does. Then flip it: "But AI CAN look through millions of pieces of data in seconds — something no human brain can do. So each side has a real strength."
Read the app's summary together: "Smart ≠ intelligent! Calculators are fast but can't learn. AI learns from data and improves over time. But your brain still wins at creativity, emotions, and understanding the world!" Make sure students can restate this in their own words before moving on — that "≠" sign is the whole lesson in one symbol.
A useful closing frame for the activity, since this age group is old enough to hold two ideas at once without oversimplifying: AI is genuinely impressive at what it does — recognizing songs, ranking search results, holding a conversation — and it is also genuinely limited in ways that matter. Neither half of that sentence cancels the other out, and that balance is worth modeling explicitly rather than letting the class land on either "AI is basically magic" or "AI is nothing special."
These work well as a think-pair-share: give students 30 seconds to think alone, a minute to compare answers with a partner, then take a few pairs' answers for the whole class. That structure gets more honest reasoning out of this age group than cold-calling.
The music-app question tends to spark real disagreement — some students will argue a non-learning recommender is "not really trying," others that it's still technically doing a job correctly. Let both sides make their case; there's no single right answer here, and the disagreement itself is doing the teaching. The goal is for students to notice that "learning from data" is doing real work in the definition, not that they all reach the same verdict.
Students take this on their own device after the activity discussion. Once everyone's submitted, review all four out loud — Builder-mode students respond well to being asked WHY the wrong options are tempting, not just which option is correct, so each explanation below calls that out.
Close the loop on the opening challenge: "So is the calculator smarter than us? Now you know the real answer — it's faster at math, but it's not intelligent, because it can't learn. Intelligence isn't about speed. It's about being able to learn and adapt — and on top of that, your brain does things, like truly feeling and creating, that no AI does at all."
It's worth naming the badge this lesson unlocks — "Brain Explorer" 🧠 — as a nice bookend: "You earned that badge today because you explored what your OWN brain can do that even a learning machine can't. That's worth being proud of, in a way no AI ever will be."
Extension activity (15–20 minutes): Split the class into pairs. Give each pair two minutes to list as many "fast but not learning" tools around their own home or barangay as they can — a jeepney fare board, a rice cooker timer, a sari-sari store's calculator, an ATM. Then two more minutes to list "learns from data" examples — Facebook's feed, a mobile game that adjusts difficulty, a translation app. Have pairs trade lists and challenge each other: for every item, can you explain in one sentence WHY it belongs on that list? The goal is defending the reasoning, not just naming examples.
To stretch a full class period, add a short debate round afterward: pick one item that different pairs disagreed on (a mobile game's difficulty setting is a reliably good one — some students will say it's just a fixed rule like "harder every 5 levels," others that it's adjusting based on how the player performs) and have the class argue both sides for two minutes before you weigh in. Resist settling ambiguous real-world cases too quickly — a lot of real apps genuinely mix both a fixed rule and a learned component, and getting comfortable with "it depends, here's how I'd check" is a more useful takeaway than a clean yes/no every time.