🕵️ Mission #002: The AI Decoder

Mission Level: AI Decoder

Goal: Level up from Investigator (finding AI) to Decoder (understanding AI).
In this mission, you’ll take one AI system and decode what kind of AI it really is.

Output: Create a 7-slide Investigation Report based on the challenges below.

Tasks

📋 Task 1: Target Identification (The Subject)

The Mission: Review your previous “AI Investigator” field notes. You need a specific subject for this deep dive.

The Action: Choose your favourite AI example found in your home or school (e.g., Netflix, Alexa, Roomba, FaceID).

Slide 1:

  • [ ] Create a Title Slide with your Agent Name and your chosen “AI Subject.”

📋 Task 2: The Vocabulary Decode (Buzzwords)

The Mission: To understand the technology, you must speak the language. Research the terms below.

  • Algorithm
  • Data
  • Bias
  • Prompt
  • Hallucination
  • Generative AI
  • Machine Learning

Slide 2:

  • [ ] Define these terms in your own words.

📋 Task 3: The Brain Size Battle

The Concept: AI brains come in two sizes:

  1. LLM (Large Language Model): Trained on lots of information and can do many different tasks
  2. SLM (Small Language Model): Trained for one specific task. It is faster, cheaper and more focused

The Mission: Compare a big brain AI with a small brain AI.

  • The Big Question: “You are on a long camping trip in the woods with NO Wi-Fi. Which AI model (LLM or SLM) do you need to help you write a scary story?”

Slide 3:

  • [ ] Explain the difference between LLM and SLM
  • [ ] Give one example each of LLM and SLM
  • [ ] The Answer: Which one would you use on the camping trip?

📋 Task 4: The Generative Artist (Lab 1)

The Concept: Generative AI creates new things from scratch using your “Prompts.”

The Mission: Go to AutoDraw (or a text-to-image tool). Start drawing a “soccer ball.” Watch the bar at the top guess what you mean. Click the AI suggestions to finish your art.

Slide 4:

  • [ ] Paste your final artwork and the “Prompt” (idea) you had in your head.

📋 Task 5: The Prediction Engine (Lab 2)

The Concept: Most AI today is “Narrow AI” (ANI). It tries to guess what you are doing based on patterns.

The Mission: Go to Quick, Draw! (Google). Play 2 rounds of the game. Pay attention to when it guesses right. How did it know? (Hint: Did it learn from other people’s drawings?)

Slide 5:

  • [ ] Report your score. Answer: How did the AI guess your drawings?

📋 Task 6: The Classification Protocol – Types Of AI

The Mission: Decide which “Team” your AI Subject belongs to.

Team 1: The Predictors (The Guessers)

  • They look at the past to guess the future.
  • Example: Netflix says, “You watched Batman, so you will like Spiderman.”

Team 2: The Generators (The Creators)

  • They make brand new things that never existed before.
  • Example: You ask for a “cat eating pizza” and it draws it.

Look at your Subject from Task 1. Is it a Predictor or a Generator?

Slide 6:

  • [ ] Write down the Team Name and one sentence explaining why. (e.g., “Netflix is a Predictor because it guesses what I want to watch.”)

📋 Task 7: The Reality Check

The Concept: There is a big difference between “Smart at one thing” and “Smart like a Human.”

  1. ANI (The Specialist): Artificial Narrow Intelligence.
  • Like a calculator or a chess bot. It is a genius at ONE specific job, but if you ask a chess bot to make toast, it fails.
  • Status: Real. We use this every day.
  1. AGI (The Super Brain): Artificial General Intelligence.
  • Like a human brain. It can learn to cook, drive, write poetry and tell jokes all at once.
  • Status: Sci-Fi (Does not exist yet).

Look at your Subject from Task 1. Is it ANI or AGI? (Hint: Can your Roomba also write your homework?)

Slide 7:

  • [ ] Define ANI vs. AGI in your own words.
  • [ ] Is your subject ANI or AGI? Why?

Checklist for Completion

  • [ ] Slide 1: Title Slide (Your Subject).
  • [ ] Slide 3: Definitions.
  • [ ] Slide 3: LLM vs. SLM + The “Phone” Question.
  • [ ] Slide 4: AutoDraw Art + Your Prompt.
  • [ ] Slide 5: Quick, Draw! results + How it works.
  • [ ] Slide 6: Types of AI + Subject Classification.
  • [ ] Slide 7: The Reality Check

Sharing your investigation

This mission works best when the Investigation Report is shared – with a parent, teacher or friend. Explaining what you learned helps you notice gaps in your own thinking and strengthens your understanding.

If you’d like to share what you discovered with us, ask a parent or teacher to submit your final 8-slide Investigation Report using the link below.

Submissions help improve future missions and understand how learners are thinking about AI. There are no right or wrong answers.

Ask a question

If this mission raised a question you’re still thinking about, ask it here:

Ask a question

This mission is part of the Learning to Think About AI series. The content is designed as a guided experience.

© Kanika Aggarwal. Licensed under Creative Commons BY-NC-ND 4.0.