Part of the How Computers Hear lesson guide. Teaching a different grade? 🔧 Builder (8–10) · 💻 Hacker (11–14) · ⚡ Architect (15–18)
Gather students close, phones and tablets down for now. Say the app's own opening line slowly, with big wonder in your voice, as if you're sharing a secret:
"When you say 'Hey Siri!' — how does the phone hear you? 🎤 It doesn't have ears! It turns your voice into a PICTURE of sound!"
Stop right there and let "a PICTURE of sound" sit for a second — most 5–7 year-olds have never heard of such a thing, and that confusion is exactly what you want. Then ask:
"Everybody cup your hands around your ears like this — pretend they're big elephant ears! Now, does a phone have ears like that?"
Let them giggle and say no. Then say: "You're right, no ears at all! So today we're going to be Sound Detectives — 🎤 — and find out the phone's secret trick for 'hearing' without ears."
Before opening the app, do one more warm-up so the idea of "a picture of sound" isn't totally new when it shows up on screen. Have everyone close their eyes and stay perfectly quiet for five seconds, then say: "Open your eyes! Even with your eyes closed, your ears were still picking up sound waves — maybe the air conditioner, maybe someone shuffling their feet. A microphone is always doing that same job, even when nobody's talking." This primes them for the idea that sound is happening constantly, even when it's invisible and quiet.
Open the lesson together on the shared screen. It has three parts — Sound Waves, Speech Recognition, and Sound AI — walk through each like a mini story, reading the facts out loud in your own simple words rather than word-for-word. At this age, the goal isn't for students to remember terms like "spectrogram" — it's for them to walk away with one solid, correct idea: a computer doesn't hear like a person, it turns sound into a picture and finds patterns in the picture. Every part below is really just a different flavor of that same idea, so feel free to slow down, repeat, and use your own classroom examples rather than rushing through all three parts.
Part 1 — Sound Waves 🌊. Say: "When you talk, invisible wiggles zoom through the air — like ripples in a pond, but you can't see them! Those wiggles are called sound waves." Have everyone clap once and say "the clap made a wiggle fly all the way to your ears!" Then explain the microphone: "A microphone is like a tiny ear for the computer. It catches the wiggle and turns it into a special picture called a spectrogram — that's just a fancy word for 'a picture of sound.'" Show the spectrogram example on screen if the lesson displays one, and say "see how it's not words, it's just shapes? That's what the computer 'sees' instead of hearing." Point out one more small but important idea from this part: the computer doesn't just take one look, it checks the sound super, super fast — 44,100 tiny little checks every single second! Say: "That's faster than you could blink a hundred times. It's checking that fast so it doesn't miss even a tiny piece of your voice."
Part 2 — Speech Recognition 🗣️. Say: "Your phone has a tiny, sleepy helper inside it that only wakes up for one job: listening for its wake-up name, like 'Hey Siri!'" Do a quick game — say a bunch of random words plus "Hey Siri" hidden in the middle, and have kids "wake up" (sit up tall) only when they hear the real wake word. Play it two or three times so even the kids who miss it the first round get another try — this game is doing real work, because it's the exact behavior the tiny on-device helper performs thousands of times a day without ever getting tired. Then say: "Once it wakes up, a bigger helper — usually far away on a big computer — figures out the rest of what you said, chops it into tiny sound pieces, and matches them to words it already knows." Try one very simple phoneme example together: say "hello" slowly, breaking it into its chunks — "huh... eh... l... oh" — and explain that this is close to what the bigger helper does, just much, much faster and with way more words to choose from.
Part 3 — Sound AI 🎵. Ask: "Has anyone heard a song and wondered 'what song is this?' and used an app to find out?" Explain Shazam simply: "It doesn't listen to the words like you do — it makes a special sound fingerprint, like your actual fingerprint, and checks it against millions of songs super fast!" Mention the bird app (Merlin) the same way — "it learned thousands of bird songs so it can guess which bird is singing" — and noise-canceling headphones — "they listen to the noisy world outside and make an opposite sound to cancel it out, like erasing noise with more noise!" If any student wears noise-canceling headphones or has seen a parent's, ask them to describe what it feels like when you turn the feature on — most describe it as the world going suddenly "quiet" or "muffled," which is exactly the canceling sound doing its job.
Close the guided part by reading the app's own summary line: "Computers can't hear like you! They turn sounds into pictures called spectrograms, then look at the picture to figure out what you said." Then let students finish the short in-app matching activity on their own or in pairs, walking around to help any student stuck on a word they don't recognize yet.
Before moving to the quiz, do a very quick standing review game if there's time: call out an item ("a book," "a voice assistant," "a chair," "Shazam," "scissors," "noise-canceling headphones") and have students stand up if they think it's a "sound helper" (uses AI to work with sound) or stay seated if it isn't. This gets everyone moving after sitting through the guided part, and gives you a fast, visible read on who's ready for the quiz and who needs one more example walked through together.
Bring the class back together in a circle or on the rug for this part. There are no wrong answers here — the goal is getting students to talk through the idea in their own words, which is a better test of understanding at this age than a written quiz alone.
Close with: "You're official Sound Detectives now! 🎤 Every time you say 'Hey Siri' or use an app like Shazam, you'll know the real secret — no ears, no magic, just a picture of sound and a very good guesser." Hand out or draw a simple detective badge sticker if your class enjoys a physical reward for finishing a lesson — the app's own badge for this lesson is called "Sound Detective," so it's a nice, easy tie-in if you want to make a paper version for the wall.
Extension activity — "Guess the Sound Picture": Make a few very simple, exaggerated "sound wave" drawings on the board — one tall and spiky (a shout), one long and flat and low (a hum), one with lots of short bumps close together (clapping). Make each actual sound yourself (or play a short clip) and have students point to which drawing matches. Then flip it — point to a drawing and have a volunteer make the matching sound. This turns the abstract "sound becomes a picture" idea into something they physically do, and stretches the 10–15 minute core lesson into a full 25–30 minute period. If time allows, add a second round where pairs of students take turns being the "sound" and the "detective" — one makes a sound with their voice or hands behind a book or folder so it can't be seen, and their partner tries to draw a quick shape that matches what they heard, then compares it to the actual sound-picture drawings on the board.