CROODI mascotCROODI
Get startedHow it worksFree missionsPricingSafety
Back to home

The AI Innovation Factory · Ages 12–16

AI for teens (ages 12–16)

6 chapters and 60 levels of comic missions with Nova, each ending in a hands-on CROODI Lab. Parent-managed, safe AI learning with no open chat for kids.

Play a free mission Start your family account
  • Build Nova, the AI model for their indie game studio team
  • Real technical terms, used correctly: mechanism first, then analogy
  • Real cases with dates: deepfakes, scams, bias and AI rules
  • Treats teens as capable, never preachy

The full course

Every chapter and level for ages 12–16, and the question each one answers.

Chapter 1 · earns the Trainer rank

What Is AI?

  1. L1Defining a Moving TargetWhat is AI?
  2. L2The Invisible StackWhere is AI in my apps?
  3. L3Rules Hit a WallWhy did machine learning win?
  4. L4Sensors to SignalsHow does AI perceive the world?
  5. L5Data Is the FuelWhat does AI actually learn from?
  6. L6Algorithm vs ModelWhat's the difference?
  7. L7Trees, Graphs and LogicCan AI reason with explicit knowledge?
  8. L8Search and PlanningHow does AI plan ahead?
  9. L9Fluent, Narrow, Not ConsciousHow smart is AI really?
  10. L10Our Studio's AI PolicyHow will we use AI?

Chapter 2 · earns the Builder rank

How AI Learns

  1. L1The Supervised PipelineHow does supervised learning work?
  2. L2Ground Truth Is HumanWho labels the data, and how good are labels?
  3. L3Features and RepresentationsWhat does the model actually 'see'?
  4. L4Generalise, Don't MemoriseHow do we know it works on new data?
  5. L5Neurons, Weights, ActivationsWhat is a neural network mathematically?
  6. L6Rolling DownhillHow does a network learn?
  7. L7Why Deep Learning Took OffWhat changed in 2012?
  8. L8Unsupervised: Clusters and RecommendersHow does AI find structure without labels?
  9. L9Learning by RewardHow does AI learn to play?
  10. L10Audit and ShipIs Nova fair and ready to launch?

Chapter 3 · earns the Word Wizard rank

How Chatbots and Generative AI Work

  1. L1From ELIZA to LLMsWhat makes modern chatbots different?
  2. L2Tokens and ProbabilitiesHow does an LLM write?
  3. L3Attention Is the TrickHow does it keep track of meaning?
  4. L4Pretraining at ScaleWhat does training a frontier model take?
  5. L5From Predictor to AssistantHow does it become helpful and safe?
  6. L6Context, Memory, RetrievalWhat does it know right now?
  7. L7Why Hallucinations HappenWhy do LLMs make things up?
  8. L8Diffusion and GeneratorsHow are AI images made?
  9. L9Reasoning Models and AgentsWhat can AI do on its own?
  10. L10System DiagramCan you explain the whole system?

Chapter 4 · earns the Prompt Pro rank

Prompt Engineering

  1. L1Prompts as ProgramsWhat is a prompt, technically?
  2. L2Constraints and CriteriaHow do I get exactly what I want?
  3. L3System vs User PromptsWho controls the AI's behaviour?
  4. L4Few-Shot PromptingHow do examples steer outputs?
  5. L5Evaluate, Then IterateHow do I know a prompt is good?
  6. L6Reasoning on RequestHow do I get better reasoning?
  7. L7Tutor, Not GhostwriterHow do I use AI for school with integrity?
  8. L8Creative WorkflowsHow do creators use AI well?
  9. L9Prompt InjectionHow do people break AI systems?
  10. L10Ship the NPCIs the NPC robust?

Chapter 5 · earns the Guardian rank

Smart and Safe with AI

  1. L1Your Data, Their ModelWhat happens to what I type into AI?
  2. L2Synthetic MediaHow do I verify a photo or video?
  3. L3Social Engineering 2.0How does AI supercharge scams?
  4. L4Parasocial MachinesWhy do AI companions feel so real?
  5. L5Lateral ReadingHow do professionals fact-check?
  6. L6Engagement MachinesWhy are feeds so hard to put down?
  7. L7Rules of the RoadWhat rules apply to teens and AI?
  8. L8Influence OperationsWho is trying to sway me, and how?
  9. L9Cognitive OffloadingDoes AI make us smarter or lazier?
  10. L10The Safety PlaybookWhat's our team's plan?

Chapter 6 · earns the AI Pro rank

AI and Our World

  1. L1Evidence for ImpactIs AI actually helping?
  2. L2Fairness Is ContestedWhose fairness?
  3. L3Work, RemixedWhat happens to jobs?
  4. L4The FootprintWhat does AI cost the planet?
  5. L5Training Data and CopyrightIs training on the internet fair?
  6. L6Governing AIWho should make the rules?
  7. L7AccountabilityWho is responsible when AI fails?
  8. L8Safety at the FrontierHow do we keep powerful AI on track?
  9. L9Open QuestionsWhat don't we know yet?
  10. L10The PitchWhat would you build?

How every level works

Three comic missions, then they build it for real.

Each level answers one big question, like “How does AI see a picture?” Every mission ends with a quick game and a Boss quiz worth up to three stars.

  1. 1

    Mission 1

    Discover

    Meet the idea in a comic. Your child makes the call on every page.

  2. 2

    Mission 2

    Use

    Spot the same idea at school, at home and in the apps they already use.

  3. 3

    Mission 3

    Fix

    The sidekick gets it wrong. Your child works out why and fixes it.

  4. 4

    Level finale

    CROODI Lab

    Build and test a tiny AI model by hand. The sidekick earns a new part.

🎮 Quick game⭐ Boss quiz: up to 3 stars⚡ XP for every star📘 One new AI word per page

Start Chapter 1 with Nova.

Try a mission free. No sign-in needed.

Play a free mission See plans