How Computers Think

Learning Objectives

This lesson is the second stop in World 1 (What Is AI?), right after students learn to spot AI in everyday life. Having seen that AI is everywhere, they now need an accurate mental model of what a computer actually is — because every later lesson (algorithms, pattern learning, generative AI) builds on top of it. By the end of the lesson, depending on age band, a student should be able to:

• Explain, in their own words, that a computer only ever works with two signals — on and off, written as 1 and 0 — and that everything that appears on a screen or comes out of a speaker is really a very long string of those two signals underneath.
• Tell the difference between fast and smart: a computer can run billions of calculations a second, but it cannot imagine something new, feel an emotion, or understand what a word actually means the way a person does.
• Recognize the Input → Process → Output pattern and identify it in an everyday device — a phone camera, a microwave, a game controller, a search box.
• (Grades 6 and up) Connect binary signals to logic gates and transistors, and explain that a modern phone or laptop chip contains billions of these simple on/off switches working together.
• (Grades 9–12) Articulate why adding more speed and memory to a computer has not, by itself, produced genuine understanding — using the von Neumann architecture and the "frame problem" as concrete reference points.

Notice the shape of that progression: every age band is being taught the same underlying truth — computation is not comprehension — just at a different level of technical precision. A five-year-old who can tell you "computers are fast, not smart" and a seventeen-year-old who can explain the von Neumann bottleneck have both fully met the goal of this lesson for their grade. Neither needs to borrow vocabulary from the other age band to have understood it correctly.

Teacher Background

You don't need a computer science degree to teach this lesson well — you need one clear idea, which is that a computer, at the very bottom of everything it does, is flipping enormous numbers of tiny electrical switches on and off. Those switches are called transistors, and a modern phone chip has billions of them. Each switch can only be in one of two states at any instant: on or off. Programmers write software that treats "on" as the number 1 and "off" as the number 0, and everything the computer ever does — running an app, showing a photo, playing a game — is built out of enormous patterns of 1s and 0s. This is called binary, and it is the one idea the whole lesson exists to plant.

It helps to walk through a concrete example before class. The letter "A" is stored inside a computer as the byte 01000001. A digital photo is a grid of tiny colored squares (pixels), and each pixel is stored as a few numbers describing its color. A song is thousands of numbers per second describing the shape of the sound wave. In every case, the computer itself never "sees" a letter, a picture, or a song — it only ever handles numbers, and the translation into something a human recognizes as text, an image, or music happens in the software and the screen or speaker, not in the computer's "understanding." This is the single most useful sentence you can give students at any age: the computer sees numbers, not meaning.

Worth being explicit with yourself, if not necessarily with the youngest students, about how this connects to the physical hardware: a single transistor doesn't do anything interesting on its own — it's just a switch. What makes a computer powerful is combining billions of them into small logical building blocks (called logic gates) that can compare, add, and route signals, and then combining those into a processor that repeats a fetch-execute cycle at blistering speed. None of that combining ever introduces a step where the machine has to "know what it's doing" in any meaningful sense — it's switches, all the way down, following rules that a human designed in advance.

The second half of the lesson introduces the Input → Process → Output cycle: something goes in (a tap, a typed word, a photo taken), the computer does something with it following its instructions, and a result comes out (a sound, a picture, a number on screen). This single pattern, repeated millions of times a second, is the shape of every computer program that has ever been written — a calculator, a search engine, a game, and every AI system in the app. Students will meet it again immediately in the next lesson, "Step-by-Step Instructions," where the "Process" step gets broken down into the actual steps of an algorithm.

Alongside all this speed, the lesson deliberately spends real time on what computers are not good at: imagining something that has never existed, feeling an emotion rather than detecting one on a face, and grasping the meaning behind a joke or a metaphor. This contrast — blazing calculation speed paired with zero genuine understanding — is the thread that the older age bands pick up again explicitly, in the language of "computation is symbol manipulation, not comprehension."

Materials & Prep

One device per student (or per pair) — a phone, tablet, or classroom computer — with a browser and an internet connection to load the lesson at aiq.ph. No installation, download, or student account is required; a child can complete the lesson anonymously, and their progress saves on that device automatically. There is nothing else to prepare for the in-app portion.

For the classroom delivery on top of the app, a few optional extras help but are not required:

• A real light switch, a flashlight, or even the classroom lights, for the on/off demo.
• A whiteboard or chart paper to sketch the Input → Process → Output diagram as a class.
• For Hacker and Architect groups only: scratch paper for the logic-gate or von Neumann exercises in those age pages' extension activities.

The lesson itself runs five to ten minutes inside the app for any age band, so plan the rest of a normal class period around the delivery script on the relevant age page rather than around additional app content — there isn't more app content to show. Each age page's Discussion and Wrap-Up & Extension sections are what turn this into a full lesson rather than a five-minute activity.

Common Misconceptions

One thread worth watching for across every age band: students will naturally reach for anthropomorphizing language ("the computer knows," "it wants to," "it thinks that") because that's how we normally talk about anything that behaves purposefully. You don't need to police every sentence — that would make the lesson feel like a grammar exercise — but it's worth gently offering the more accurate phrasing ("it's been given instructions for," "it was trained to react to") often enough that students start to notice the difference themselves by the end of the unit. That habit of precise language is, in miniature, the whole point of the "AI Foundations" big idea this lesson belongs to.

Pick your grade's script