Skip to main content

Week 13: The AI Discovery Lab

Introduction to Generative AI

Throughout this course you have learned how to:

  • control computers
  • organize files
  • communicate with text and images
  • search for information
  • write simple programs
  • debug problems
  • understand systems

Now we explore something new and very powerful:

Artificial Intelligence (AI).

AI tools can:

  • write text
  • generate images
  • answer questions
  • help solve problems
  • help people create things faster

But there is something very important to understand.

AI is not magic.

AI does not truly “know” things the way people do. Instead, it generates responses based on patterns it has learned from large amounts of information.

The big idea this week:

AI is a tool that generates ideas, but humans must think about whether those ideas make sense.

Learning to use AI well means:

  • asking good questions
  • thinking critically about answers
  • using AI to help creativity rather than replace thinking

🔒 Safety Note

All AI interactions for ages 8–12 should be supervised by an adult. Do not let learners share personal information with AI tools, including full names, addresses, school names, passwords, or private photos. Remind learners that AI can sound confident and still be wrong — always check important facts with a trusted source. AI outputs should be reviewed by a human before being trusted or shared.

Facilitator Snapshot
  • You do not need to teach every bullet on the page. Use the learning goal and one or two activities for the session you are teaching today.
  • If time is short, teach one guided session well and leave the rest for later. The lessons are designed to stretch across the week.
  • The independent session works best after the learner has already explored the main idea with you once.
Communication Moment

Talking to an AI rewards the same skill as talking to people: clear requests. A vague prompt gets a vague answer; a specific one — "Explain ___ in two sentences for a 10-year-old" — gets something useful. Saying exactly what you want, and what you already tried, is good communication whether the listener is a person or a program. (More on the Communication Skills page.)

Problem Solving Moment

If an AI tool gives a confusing result, treat it like a test: change one thing in your prompt, observe what changed, and adjust. That try-observe-adjust loop is problem solving, not just guessing. (More on the Problem Solving Skills page.)

Facilitator Preparation

Before You Begin
  • Time needed: ~30–45 minutes per guided session; ~20–30 minutes for the independent session.
  • Devices needed: A computer or tablet with internet access.
  • Access note: Check provider terms, school or library policy, and family expectations before using any AI tool. In this curriculum, adult-managed access is the default: the adult opens the tool, the learner directs the questions, and saved examples or screenshots are a valid no-login fallback.
  • Ensure access to a generative AI tool (such as ChatGPT, Gemini, Copilot, or another conversational AI). Any conversational AI tool will work for this lesson — see Tool Alternatives for options.
  • Prepare a few simple prompts to demonstrate AI behavior.
  • Have paper or a whiteboard available for discussion.
  • Confirm the My Projects folder exists.
  • Set up a visual timer.
Teaching Mindset

This week should feel curious and exploratory, not technical.

The student should leave with three ideas:

  • AI can generate things
  • AI can make mistakes
  • humans must still think carefully
Coping Skill Moment

AI tools can sound very sure of themselves, which can feel impressive or convincing. Before believing or acting on an AI answer, run a quick fact-vs-story check: What do I actually know is true, and what is the tool just claiming? Slowing down keeps you in charge of the tool. (More on the Coping Skills for Digital Life page.)

Week at a Glance

Learner Goal

I can ask AI a question, notice when it sounds helpful or wrong, and keep my private information safe.

Materials

  • adult-managed AI tool or saved prompt-and-response examples
  • paper or note sheet for observations
  • a second source to compare against when needed

Quick Formative Check

Ask two questions before using any tool: "Should we type our full name into AI?" and "Can AI sound right but still be wrong?"

What Success Looks Like

  • The learner can name one privacy rule for using AI.
  • The learner can point out at least one part of an AI answer that should be checked, changed, or questioned.

Low-Tech / Offline Option

Use printed AI prompts and responses, then annotate what seems useful, questionable, private, or worth checking with another source.

AI Use Note

If live AI access is not appropriate in your setting, use saved examples and keep the focus on privacy, fact-checking, and judgment. The learning goal is not logging in. The learning goal is thinking clearly about outputs.


Guided Session 1

What Is AI?

Learning Goal

By the end of this session, the student can:

  • analyze AI output as pattern-based rather than human understanding
  • compare what AI does well with what people still need to decide or judge
  • create or extend an idea using AI as a brainstorming partner

Activities

1. Ask the Student What They Think AI Is

Start with an open question:

“What do you think artificial intelligence is?”

Let the student share their thoughts.

Many children imagine robots or super-smart machines. That’s a great starting point.

Explain that AI is really a computer system trained to recognize patterns and generate responses.


2. Demonstrate AI Generating Text

Open an AI tool.

Before typing, do a quick privacy pause together:

  • no full name
  • no address
  • no school name
  • no passwords
  • no private photos or identifying details

Ask it a simple question such as:

  • “Tell me three interesting facts about dogs.”
  • “Write a short story about a robot and a puppy.”

Read the response together.

Ask the student:

  • “How do you think the computer came up with that?”
  • “Do you think it knows those things or is it generating them?”

Explain that the AI is generating answers based on patterns it learned during training.


3. Generate Something Creative

Try a fun prompt together:

Examples:

  • “Invent a silly animal that lives on the moon.”
  • “Describe a robot that helps take care of dogs.”

Explain that AI can help generate ideas.

But the human decides whether those ideas are good.


Reflection Questions

  • “What evidence makes AI seem powerful, and what evidence shows its limits?”
    • Sentence starter: “AI seems powerful because… but it is limited because…”
  • “How is AI pattern-making different from human understanding?”
    • Sentence starter: “AI finds patterns, but a person can…”
  • “When would you trust AI to help you brainstorm, and when would you question it more carefully?”
    • Sentence starter: “I would trust AI to help with… but I would double-check when…”

Guided Session 2

When AI Gets Things Wrong

Learning Goal

By the end of this session, the student can:

  • evaluate whether an AI response is sensible, useful, or questionable
  • verify AI information by comparing it with another source or their own reasoning
  • justify why AI should support thinking instead of replacing it

Activities

1. Ask AI a Tricky Question

Try asking the AI something slightly unusual or playful.

Examples:

  • “Can dogs breathe on the moon?”
  • “How many legs does a dragon have?”
  • “What is the fastest flying sandwich?”

Discuss the response.

Ask:

“Does that answer make sense?”

Explain that AI sometimes guesses based on patterns.


2. Compare AI to Searching the Internet

Explain that using AI is similar to searching online.

Both require us to ask:

  • Who made this information?
  • Does it make sense?
  • Should we check another source?

Add one more question:

  • Is there anything here we should reject, rewrite, or keep private?

AI is simply another tool for exploring ideas.

Connect to something they already know:

"Remember in Week 7 when we learned that not every search result is trying to help you? And in Week 8 when we learned that polished doesn't mean true? The same rules apply to AI. Just because the AI writes a confident-sounding answer doesn't mean it's correct. Always check important facts."

For a refresher on evaluating online information, see Digital Habits & Safety.


3. Improve an AI Idea

Ask the AI to generate something creative.

Example:

Invent a new kind of robot pet.

Then ask the student:

  • What do you like about that idea?
  • What would you change?
  • How could you make it better?

Explain that creativity often works this way.

AI gives a starting point, and people improve it.


Reflection Questions

  • “What helped you decide whether an AI answer was sensible or questionable?”
    • Sentence starter: “I could tell the answer was… because…”
  • “How can AI support creativity without making the important decisions for us?”
    • Sentence starter: “AI can help by… but the person still needs to…”
  • “Why is human judgment still necessary even when AI sounds confident?”
    • Sentence starter: “Even when AI sounds sure, a person should… because…”
  • “What information should stay private when using AI, and what should we double-check before sharing?”
    • Sentence starter: “I would keep … private, and I would check … before sharing it.”

Independent Session

AI Idea Explorer

Instruction

Use AI to explore an idea you are curious about, but evaluate the result instead of just accepting it.

After reading the AI response, ask yourself:

  • Does this make sense?
  • What parts seem useful?
  • What parts need to be checked, changed, or improved?

Then create your own version by writing, drawing, or expanding the idea into something better.

Be ready to explain what you kept, what you changed, and why.

Save any drawings or writing in:

My Projects → Experiments


Skills Reinforced

  • analyzing AI as a tool rather than a thinker
  • asking stronger questions to test or extend ideas
  • evaluating generated information for usefulness and accuracy
  • developing ideas creatively with human judgment

Setup

  • AI tool open (adult-supervised account)
  • drawing or writing tool available
  • access to My Projects → Experiments
  • visual timer

🔄 Simplify or Extend

Simplify:

  • Limit the session to just 2–3 AI prompts and focus on the question: “Does this answer make sense?”
  • Use a pre-written list of prompts instead of asking the learner to create their own.
  • Focus on one big idea: AI generates based on patterns, not understanding.

Extend:

  • Ask the learner to create a “Fact vs. Fiction” chart — test the AI with questions they already know the answers to and record what it gets right and wrong.
  • Have the learner compare AI responses to the same question asked twice — do the answers change?
  • Challenge the learner to write a short guide: “Three things to check before trusting an AI answer.”

💾 Save This Week’s Artifact

Save your AI interaction notes, screenshots of prompts and responses, or written observations about AI accuracy to your portfolio folder (My Projects → Experiments). Include at least one example where you identified something the AI got wrong or something you improved. Label it with the date.


Check for Understanding

By the end of this week, look for evidence that the learner can:

  • Explain that AI generates responses based on patterns, not true understanding.
  • Give an example of something AI got wrong or something that didn’t make sense.
  • Describe why AI answers should be checked with a trusted source before being trusted.
  • Explain that AI is a tool that helps people think, not a replacement for thinking.
  • Identify at least one way AI could be helpful and one way it could be misleading.
  • Use phrases like "I need to check that" or "that doesn't sound right" when reviewing AI output.

Vocabulary This Week

Artificial Intelligence (AI)Generative AIPromptPatternOutputVerifyHallucination (when AI confidently states something false)
See the Glossary for definitions.

Preview of Next Week

Next week, students use AI as a creative partner — designing and refining prompts to improve AI responses and discovering how to stay in control of the thinking while AI handles the generating.