For Parents, Kindergarten & Early Primary Educators

Raising Curious Minds in an AI World

Clear, non-technical conversation guides, classroom-tested unplugged lesson plans, browser puzzles, and printable activities to help kids aged 4–8 understand AI safely.

10 lesson plans21 browser puzzles7 printable sheets20 discussion cardsFree, no login
Conversation Framework

The 11 Big Questions Kids Ask About AI 💬

Children naturally wonder if robots are alive or if computers can think. Here are honest, developmentally appropriate answers you can share word-for-word.

What to say to your child:

"Robot Miles is like a really smart toaster or a bicycle with a screen! It runs on electricity and computer code. It doesn’t have a heartbeat, it doesn’t have a tummy, and it doesn’t dream when you close the laptop. It cannot feel happy or sad, even if it uses a smiley emoji!"

💡 Helpful Analogy:Think of a music box: it plays a lovely tune when wound up, but it doesn’t know what music is.
🎯 Follow-Up Question:Ask your child: What things can you do that a computer can never do? (Give hugs, taste ice cream, run in the grass).

Healthy Digital Habits

4 Screen-Time Principles for Early AI 🌿

How to nurture healthy computational curiosity without falling into passive screen addiction.

Active Creation over Passive Consumption

Look for tools where children build, prompt, and test hypotheses rather than passively watching infinite video feeds.

Practical Tip: Encourage kids to treat AI like digital clay or LEGO blocks: prompt it with curious questions, challenge its answers, and test its boundaries.

Zero Login, Zero Telemetry

Young children should never be required to have an account, enter an email, or submit biometric voiceprints.

Practical Tip: Mile Coder is designed to be 100% ephemeral: no cookies, no tracking, and everything runs in the local browser.

Co-Viewing & Conversational Exploration

AI tools for young children are most effective when experienced alongside an engaged grown-up.

Practical Tip: Take turns asking questions to Robot Miles. Pause after each reply: "Does that match what we saw at the zoo last week?"

Demystifying the "Magic"

Children naturally anthropomorphize machines. Helping them understand the math keeps technology grounded.

Practical Tip: Replace phrases like "the computer knows" with "the computer counted" or "the computer was trained on photos".
Jargon, Translated

The AI Words Grown-Ups Keep Hearing 📖

No computer science background needed. Each entry gives you a plain-English definition — and the version that actually works on a five-year-old.

Algorithm
A fixed sequence of steps for completing a task. A recipe is an algorithm; so is long division. Nothing about the word implies intelligence.
Say it to a child: "It is a list of steps, like the instructions for making a sandwich."
Model
The file that comes out of training — a very large set of numbers capturing patterns found in the data. When people say "the AI", they usually mean the model.
Say it to a child: "It is the part that did all the practising, saved into a file."
Training data
The examples a model was shown while learning. Its strengths and its blind spots both come directly from here.
Say it to a child: "It is the pile of examples we showed the computer so it could learn."
Token
The chunk a model actually reads. Roughly a short word or part of a longer one — "unhappiness" might arrive as three tokens.
Say it to a child: "Computers chop words into little pieces before reading them."
Large language model (LLM)
A model trained to predict the next token across an enormous amount of text. Everything it produces — answers, poems, code — comes out of that single prediction step, repeated.
Say it to a child: "It is a really good guesser that has read a huge number of sentences."
Prompt
The text you give a model. It is not a command so much as a starting point the model continues from — which is why phrasing changes the answer so much.
Say it to a child: "It is what you ask the robot."
Hallucination
A confident, fluent, entirely invented answer. Not a bug being fixed soon — a direct consequence of a system that predicts plausible text rather than retrieving facts.
Say it to a child: "Sometimes the robot daydreams and says something that is not true, without meaning to."
Bias
Systematic unfairness that a model inherits from gaps or skews in its training data. It is a data problem far more often than a code problem.
Say it to a child: "If we only show the robot black cats, it will not believe an orange cat is a cat."
Machine learning
Building a system by showing it examples rather than writing rules for it. Nobody programmed "what a cat looks like" — the model worked it out from photographs.
Say it to a child: "Instead of telling the computer the rules, we show it lots of examples."
Reinforcement learning
Learning by trial and reward. The system tries something, receives a score, and adjusts — the same loop as training a puppy with treats.
Say it to a child: "The computer keeps trying, and we say 'yes, warmer!' when it gets closer."
Computer vision
Getting meaning out of images. The model never sees a picture as we do — only a grid of numbers, one per pixel.
Say it to a child: "To a computer, a photo is just thousands of tiny coloured dots."
AI agent
A model given tools and a goal, allowed to take several steps on its own — look something up, use the result, decide what to do next.
Say it to a child: "A helper that plans a job and does it one step at a time."
Human in the loop
A design where the machine proposes and a person decides. The standard safeguard wherever a wrong answer would actually matter.
Say it to a child: "The computer helps, but a person always makes the final choice."
On-device
Computation that happens in the browser or on the phone itself, with nothing sent to a server. Nothing to intercept, store, or leak.
Say it to a child: "Everything happens right here on our own computer. Nothing gets sent away."
Ready to Explore?

Play the 16 AI Missions with Your Child or Class

Step-by-step interactive games explaining prediction, tokens, agents, and robot training.