Ages 6–7

Ethical Inventors

A clever machine still needs a good heart to guide it.

Children learn what a "thinking machine" really is — a tool that follows instructions and learns from examples — and grapple with the big question: who is responsible for what it does? They solve multi-step problems and design fair rules for their own inventions.

👨‍👩‍👧 For parents: Your child meets technology and AI the right way — as powerful tools that a kind, responsible human must guide. They learn how machines "learn", who is responsible when one errs, and to ask "should we?" not just "can we?". This is digital wisdom for the generation that will shape AI, not be shaped by it.
STEM & LogicEthics & Values (Akhlaq)CognitiveInquiry-basedSocratic dialogueProject-basedUnplugged coding

10 · 2 · 45weeks · sessions/week · min/session

📅 Session plan📝 Observation log🖨 Home cards🃏 Card decks

Learning objectives

  • Explain in their own words that a machine learns from examples we give it. STEM & Logic · understand
  • Argue who is responsible when a machine makes a mistake, with a reason. Ethics & Values (Akhlaq) · evaluate
  • Break a hard problem into smaller steps (decomposition). Cognitive · analyze

Modules

What Does a Machine Know?

🎮 Play it on screen

Big idea: A machine only knows what we teach it.

🔬 Why this works: Constructionist AI literacy (CSforAll; MIT): children grasp machine learning by training a model themselves with examples. Seeing the robot inherit their own mistakes makes the abstract idea "machines learn from human data" concrete and personal.

Responsibility / trust (Amanah)Humans carried the Amanah — the trust. A tool has no Amanah; the person using it does. (Surah Al-Ahzab 33:72)

Teach the Sorting RobotUnplugged coding · 20m

Children "train" a pretend robot (a friend) by showing examples of apples vs oranges, then test it.

▶ Show me how
  1. 🔬 Experiment Show the robot 5 apples and 5 oranges. Now give it a tricky one — what does it guess, and why?
    Facilitator cue: When it errs, ask: "Did the robot do wrong, or did we teach it badly?" This is the core idea.
    Support: Start with obvious cards before the tricky red orange.
    Stretch: What examples would you add so it never gets fooled again?
Train Another RobotUnplugged coding · 18m

Children give a pretend robot matched examples, then test its guesses.

▶ Show me how
  1. 🔬 Experiment Teach the robot with clear examples, then give it a hard one.
    Facilitator cue: When it errs, ask "did it fail, or did we teach it too little?"
    Support: Use only obvious examples first.
    Stretch: What example would fix its mistake?
  2. 💬 Discuss The robot only knows what we showed it. Is that fair to the robot?
    Facilitator cue: Lead toward "the human is responsible for the teaching".
    Support: Compare: a tool follows; a person chooses.
    Stretch: What should we always check before trusting its guess?

Who Is Responsible?

🎮 Play it on screen

Big idea: Tools obey; people choose.

🔬 Why this works: Moral-reasoning research (Kohlberg; restorative practice): children reason most deeply about responsibility through a concrete dilemma they care about. Separating the tool from the human choice plants the root of accountability — and of human agency over machines.

The Courtroom of KindnessSocratic dialogue · 18m

A short scenario: a delivery robot dropped a neighbor's gift. The class reasons out who should fix it.

▶ Show me how
  1. 💬 Discuss The robot dropped it — but who told the robot what to do? Who should say sorry?
    Facilitator cue: Resist naming the answer; let them trace the chain from robot back to person.
    Support: Walk it back one step at a time: who pressed go? who built it?
    Stretch: Is more than one person responsible? Share the blame fairly.
  2. 🌙 Reflect Could we make a rule so it does not happen again? Write our fair rule.
    Facilitator cue: A good rule is short and testable; ask "how would we know it was followed?"
    Support: Offer a frame: "The robot must always ___ before ___."
    Stretch: Write a second rule for what to do if the first one fails.
Whose Feelings?Socratic dialogue · 16m

Children weigh how each person in a mishap feels before deciding what is fair.

▶ Show me how
  1. 🌙 Reflect How does the neighbour feel? How does the robot’s owner feel?
    Facilitator cue: Hold both feelings at once; fairness needs every side.
    Support: Take one person at a time.
    Stretch: Could a fair fix make everyone feel better?
  2. 💬 Discuss Saying sorry can mend feelings. Who should say it, and how?
    Facilitator cue: A real sorry names the harm and offers to fix it.
    Support: Practise the words of a kind apology.
    Stretch: Is sorry enough, or is a fair fix needed too?

Invent It Right

🎮 Play it on screen

Big idea: A good invention helps people and harms none.

🔬 Why this works: Decomposition is a core computational-thinking skill (Wing); engineering-design pedagogy adds an ethics constraint from the start. Naming "who must never be harmed" before building teaches that good design is responsible by design.

Design a Helpful MachineProject-based · 25m

Teams decompose a real problem (e.g. remembering to water plants) and sketch a machine with a fairness rule.

▶ Show me how
  1. 🎯 Challenge Break the problem into 3 small steps your machine must do. Then add one rule to keep it kind.
    Facilitator cue: Push for the "kindness rule" — every team must name a person their machine must never harm.
    Support: Do the first step together to show what "small" means.
    Stretch: Add an "if it is unsure, then ask a human" rule.
Add a Kindness RuleProject-based · 20m

Children sketch a helpful machine and design one rule to keep it kind.

▶ Show me how
  1. 🎨 Create Design a machine that helps — then add one rule so it never harms anyone.
    Facilitator cue: Insist every design names a person it must never harm.
    Support: Decide what it helps with first, then add the rule.
    Stretch: Add a rule: "if unsure, ask a human".
  2. 🌙 Reflect Show your machine and its kindness rule. Why did you choose it?
    Facilitator cue: Value the reasoning behind the rule, not the drawing.
    Support: Say the rule in one short sentence.
    Stretch: What rule would you add if it could talk?

Should We, Even If We Can?

🎮 Play it on screen

Big idea: The human stays in charge of the machine — always.

🔬 Why this works: Ethical inquiry (P4C) applied to technology and AI literacy: children learn to separate "can we?" from "should we?" and to question a tool's authority. This is the direct lesson against being controlled by machines.

The Inventor's PauseSocratic dialogue · 22m

A clever tech could do harm; children reason out whether it should be used.

▶ Show me how
  1. 💬 Discuss A machine could do your friend's homework for them. Could it? Should it? Why not?
    Facilitator cue: Draw out the difference between capability and rightness. Let them wrestle with it.
    Support: Separate the two questions on paper: a "can" column and a "should" column.
    Stretch: When could the same machine be a good helper instead?
  2. 🎯 Challenge Write one rule for your invention that says when it must STOP — and ask a human.
    Facilitator cue: Insist the stop rule hands the decision back to a human, not another machine.
    Support: Finish the sentence: "It must STOP when ___."
    Stretch: Add: "If it is unsure, the human decides, not the machine."
  3. 🌙 Reflect Who should always decide the important things — the machine, or the person? Why the person?
    Facilitator cue: Let the reasons, not the grown-up, settle it; echo the strongest one back.
    Support: Offer the contrast: a machine has no Amanah; a person does.
    Stretch: Name one important thing you would never let a machine decide.
Can vs ShouldSocratic dialogue · 18m

Children sort actions into "a machine can" and "a machine should", and reason about the gap.

▶ Show me how
  1. 🎯 Challenge A machine CAN do this — but SHOULD it? Sort each card, and say why.
    Facilitator cue: Draw out the difference between capability and rightness.
    Support: Separate "can" and "should" on paper first.
    Stretch: Find one card that is "can" but never "should".
  2. 🌙 Reflect Who should always decide the important things — the machine or a person? Why?
    Facilitator cue: Let the reasons settle it; the human stays in charge.
    Support: A machine has no Amanah; a person does.
    Stretch: Name one thing you would never let a machine decide.

Leadership we plant

  • 🌱 Asks "should we?" — not only "can we?".
  • 🌱 Keeps the human in charge of the machine.
  • 🌱 Designs so that no one is harmed (la darar).

Research foundations

Philosophy for Children (P4C) — ethical inquiry
Moral reasoning grows through structured dialogue around real dilemmas.
In practice: Courtroom-of-kindness scenarios and the "should we?" pause.
AI literacy — "humans in control" (CSforAll / MIT-style)
Children should understand machines learn from human-given examples and stay human-governed.
In practice: Training a pretend sorting robot, then ruling on its mistakes.

🏡 Try at home

Should We, Even If We Can? · 10 min

At dinner, pose one "could vs should" question ("A robot could do homework — should it? Why not?"). Let your child reason it out; there is no single right answer.

Teach the Machine · 12 min

Play "robot": your child gives you example after example to "train" you to sort fruits, then tests you with a tricky one. Ask: did the robot fail, or did we teach it badly?

Standards alignment

ISTE Standards for Students
Digital Citizen & Computational Thinker
Responsible technology use and how learning systems work.
UNESCO — AI Competency Framework for Students
Human-centred AI & AI ethics
Keeping humans in control and reasoning about AI’s impact.
CASEL — Social & Emotional Learning
Responsible decision-making
Weighing consequences and acting ethically.

Value anchors

  • Trust & responsibility (Amanah)
  • Doing no harm (La darar)There should be neither harm nor reciprocating harm — our inventions follow this rule too. (Sunan Ibn Majah)

Everything you’ll need (home or school)

  • Fruit picture cards, scenario cards, design paper, markers
  • Picture cards of fruits (some tricky: a red orange!)
  • Picture cards to sort into two groups (animals vs plants)
  • A short "who is responsible?" scenario
  • Design paper and markers
  • Action cards (do homework, water plants, decide who is right)
🖨 Printable checklist

Assessment — portfolio

  • emerging: States that machines follow instructions.
  • developing: Explains learning-from-examples and decomposes a problem.
  • confident: Defends who is responsible and proposes a fairness rule.

Future skills

AI literacyDigital ethicsComputational thinkingProblem solvingLogical reasoning
Qubtan · قُبطان