Kids Guide

Typing letters on the keyboard is called input. The computer calculates with what it receives — that's processing. Then it shows the result on the screen — that's output.

AI works the same way. You input a question, the AI calculates, and it outputs an answer. That's why the first step to talking with AI is getting friendly with the keyboard.

Put both index fingers on F and J. Those two keys have small bumps, so you can find them with your fingertips without looking.

It's fine to be slow at first. Typing correctly comes before typing fast. A little practice every day helps your fingers remember where to go.

AI talks so smoothly that it can feel like a person. But AI isn't a living thing — it's a computer program that people built.

AI doesn't get hungry, happy, or sad. It just calculates very fast, based on a huge amount of text. Both people and AI can be wrong, so people check important decisions and take responsibility for them.

If you show a computer lots and lots of cat photos and dog photos, it calculates with numbers to find a rule (a pattern) that tells them apart. This is called learning.

After learning, it tries to guess whether a brand-new photo shows a cat or a dog. If it gets many wrong, people collect more examples and train it again. The examples it learns from matter a lot.

A chatbot AI doesn't write a long answer all at once. For each "next word" it calculates scores, picks one, adds it, and then scores the word after that.

Usually a high-scoring word is picked, but sometimes a lower-scoring one gets picked too. That's why the same question can get slightly different answers. Tap the bars below to finish the sentence.

AI sometimes makes answers that sound right but are wrong. Experts call this a "hallucination." It doesn't mean the machine is seeing things — it means a sentence it calculated doesn't match the facts.

So for important things like homework, health, and safety, always check once more with a book, a teacher, or a parent. Catching a wrong answer is a great skill too.

A fuzzy question gets a fuzzy answer. If you say what you're curious about, who will read the answer, and what shape you want it in, it's easier to get a helpful answer.

Even a good question can get a wrong answer. So after you get an answer, remember the checking steps from Lesson 6.

What you type to an AI travels over the internet to another computer (a server). So never type information that could identify you.

If something scary or sad happens, tell a trusted adult like a parent or teacher first — not an AI. An AI can't protect you in real life.

A program runs in a fixed order of steps. A flowchart is a picture of that order. Drawing a flowchart before you build something makes it clear what you're going to make.

Below is the flowchart for a number-guessing game. Play the game next to it and the box you're in lights up. If you guess wrong, follow the arrow back to the choosing box. Going around like this is called a loop.

Vibe coding means describing what you want to make in words, and the AI writes the code. You can simply describe the flowchart from Lesson 9.

It doesn't have to be perfect the first time. Run it, then fix it by pointing out odd things one at a time. The AI wrote the code, but you're the one who decides what to make and checks that it works.