Lesson 1, Topic 1

In Progress

Unit 1 – Teacher Guide

Unit 1 Teacher Guide - Introduction to AI | Knowledge Pillars · AWS AI Explorers

📐

Unit 1 At a Glance

Lessons: 1.1 - 1.4

Learning Goals

2.5 hrs

Est. Student Time

Quiz Questions (5 per lesson)

Learn

Phase of the Arc

Unit 3

Where the Project Lives

#	Title	Core Ideas	Interactive Elements	Est. Time
1.1	Why AI Matters	AI in daily life User to builder Challenges AI creates Unequal access	Flip-card vocabulary Knowledge Check (5 Q) Activity: A Challenge Worth Solving	~40 min
1.2	Fundamentals of AI and ML	AI vs ordinary software How machines learn Garbage in, garbage out Narrow AI	Flip-card vocabulary Knowledge Check (5 Q) Activity: Spot the AI	~35 min
1.3	Fundamentals of Generative AI	How GenAI creates Answers vs building Hallucinations Output types	Flip-card vocabulary Knowledge Check (5 Q) Activity: Try and Verify	~35 min
1.4	Responsible AI	Privacy, transparency, attribution, honesty Risks Verification Representation	Flip-card vocabulary Knowledge Check (5 Q) Activity: Responsible-Builder Habits	~40 min

📚 Lesson-by-Lesson Resources

🎯

Lesson Purpose

Lesson 1.1 Purpose: Students start from what they already know. The goal is to shift their mindset from AI as a thing that answers questions to AI as a tool they can build with. Surface the challenges AI creates (misinformation, bias, privacy, unequal access) honestly, and have every student leave with a real community challenge they care about, which becomes the seed of their Unit 3 project.

📖

Learning Objectives

📖 Students will know:

AI is already part of their daily lives (feeds, recommendations, voice, chat)
The difference between using AI and building with AI
The real challenges AI creates
That access to AI and representation are not equal

⚡ Students will be able to:

Name several places AI shows up in their day
Explain what it means to build with AI
List risks AI creates
Write one community challenge they might tackle

🔑

Quiz Answer Key: Lesson 1.1

Q1. Which of these is an example of AI you probably already use?

A feed or app that learns what you like and recommends more of it

Recommendation systems learn patterns from your behavior, which makes them AI. Calendars, calculators, and switches follow fixed rules.

Watch for: Common wrong - A/C/D: fixed-rule devices are not AI. Ask: does it learn from data?

Q2. What does it mean to move 'from user to builder'?

To put AI inside a tool you design that helps real people

This reframes the whole program: students are not just consumers of AI, they will create with it in Unit 3.

Watch for: Common wrong - A: only asking questions is using AI, not building.

Q3. Which of these is a real challenge that AI creates?

Misinformation and bias

Responsible builders know the problems, not just the benefits. Misinformation and bias are central.

Watch for: Common wrong - A/C/D: these overstate AI as harmless; they are not true.

Q4. 'Access to AI is not equal' mostly means:

Some communities, languages, and cultures are better represented and served than others

Access and representation vary; this sets up the representation work in Lesson 1.4 and in the build.

Watch for: Common wrong - A/B: cost and language support are not equal.

Q5. What is a strong first step toward a project?

Notice a real problem in your school or community that you care about

A real, cared-about problem makes a stronger project, and no coding is needed to start.

Watch for: Common wrong - A/B/D: complexity, copying, and waiting for coding skills all weaken the start.

💬

Discussion Prompts: Lesson 1.1

AI you did not notice

Name a moment today when AI made a decision for you without you thinking about it. Did it help, or did it nudge you somewhere?

Look for: Look for students connecting everyday tech (feeds, autocomplete, maps) to the idea that AI is making predictions about them. Strong responses notice both benefit and influence.

User or builder

Think of a problem at school. What would it look like to build something that helps, versus just asking AI about it?

Look for: Look for the shift from consuming answers to designing a tool for others. Encourage concrete, local problems.

⚠️

Common Misconceptions: Lesson 1.1

❌

Misconception

"If I use AI, I am cheating or not really doing the work."

✅

The reality

Using AI as a tool you direct, check, and are honest about is a real skill. The work is in the judgment, design, and verification.

❌

Misconception

"AI is fair and neutral because it is a computer."

✅

The reality

AI learns from human data, so it can carry human bias. Fairness is something builders have to check for.

🎯

Lesson Purpose

Lesson 1.2 Purpose: Give students a plain-language mental model: AI learns patterns from data rather than following only hand-written rules. Keep it concrete with everyday examples. The most important takeaways are that data quality shapes AI quality (garbage in, garbage out) and that all of today's AI is narrow, not human-like.

📖

Learning Objectives

📖 Students will know:

What AI is and how it differs from ordinary software
How machines learn from data
That data quality shapes AI quality
That today's AI is narrow, not general

⚡ Students will be able to:

Contrast AI with rule-based software
Give a daily-life example of learning from data
Explain garbage in, garbage out
Name a common misconception about AI

🔑

Quiz Answer Key: Lesson 1.2

Q1. What is the main difference between AI and ordinary software?

AI learns patterns from data instead of following only fixed rules

This is the core concept of the lesson. Ordinary software follows rules; AI learns patterns.

Watch for: Common wrong - C: AI is not just a faster calculator; it learns.

Q2. How does a machine 'learn'?

It finds patterns across many examples in data

Learning means generalizing from many examples, not hand-coding rules or random guessing.

Watch for: Common wrong - A: hand-written rules describe ordinary software, not learning.

Q3. What does 'garbage in, garbage out' mean for AI?

If the training data is biased or poor, the AI's results will be too

Data quality directly drives output quality, which connects to bias and representation.

Watch for: Common wrong - A: it is not about physical waste; it is about data quality.

Q4. Which statement about today's AI is TRUE?

It is narrow: great at specific tasks, not general

All current AI is narrow. There is no general, human-like AI today.

Watch for: Common wrong - A/B: AI does not think like a human and is not general.

Q5. Which is a common misconception about AI?

That AI understands things the way people do

AI predicts from patterns; it does not understand like a person. The other options are true statements.

Watch for: Common wrong - A/B/D: these are all accurate, so they are not misconceptions.

💬

Discussion Prompts: Lesson 1.2

Rules vs learning

Describe something a computer does with fixed rules and something that seems to learn about you. How can you tell them apart?

Look for: Look for the distinction between deterministic, rule-based behavior and behavior that adapts based on data.

Garbage in, garbage out

If an AI only ever saw examples from one group of people, how might it behave for everyone else?

Look for: Look for students connecting biased or narrow data to unfair or wrong results, previewing representation in Lesson 1.4.

⚠️

Common Misconceptions: Lesson 1.2

❌

Misconception

"AI understands what it is saying."

✅

The reality

AI predicts likely patterns; it has no understanding. That is why it can produce fluent but wrong output.

❌

Misconception

"More AI always means smarter results."

✅

The reality

Results depend on data quality and the task. Poor data produces poor results no matter how advanced the model.

🎯

Lesson Purpose

Lesson 1.3 Purpose: Introduce the kind of AI students will build with. Land two ideas hard: generative AI predicts new content (so it can be confidently wrong, a hallucination), and there is a difference between getting answers and building with AI. Verification is the habit that carries into every build.

📖

Learning Objectives

📖 Students will know:

How generative AI creates content from patterns
The difference between getting answers and building
That outputs can be wrong and need verification
The output types of generative AI

⚡ Students will be able to:

Explain prediction in plain terms
Tell apart asking and building
Describe how to verify an output
Name generative AI output types

🔑

Quiz Answer Key: Lesson 1.3

Q1. Generative AI creates new content by:

Predicting based on patterns it learned from training data

Generative AI predicts the most likely next token; it does not copy a stored file.

Watch for: Common wrong - A: it generates, it does not retrieve one file.

Q2. What is the difference between getting answers and building?

Getting answers is asking; building puts AI inside a tool you design

This reinforces the program's core arc toward the Unit 3 build.

Watch for: Common wrong - C: building takes design thinking; it is not effort-free.

Q3. An AI gives a confident answer that turns out to be false. This is called:

A hallucination

Hallucinations are confident, false outputs. Naming them helps students stay skeptical.

Watch for: Common wrong - C: a syntax error is a coding error, not a false statement.

Q4. What is the best response to an important AI answer?

Verify it against a reliable source

Confidence is not correctness. Verification is the key habit.

Watch for: Common wrong - A: sounding sure does not make it true.

Q5. Which is a real output type of generative AI?

Text, images, and code

Generative AI spans multiple modalities, which is why it is so flexible.

Watch for: Common wrong - A/C: it is not limited to a single output type.

💬

Discussion Prompts: Lesson 1.3

Fluent but wrong

Why might an AI answer sound completely convincing and still be false? How would you catch it?

Look for: Look for understanding that prediction produces fluent text regardless of truth, and that verification against sources is the fix.

Answer or build

Pick a task. What would it look like to just ask AI, versus build a small tool around it for your class?

Look for: Look for a concrete reframing from one-off answers to a reusable tool that helps others.

⚠️

Common Misconceptions: Lesson 1.3

❌

Misconception

"If the AI sounds confident, it is probably right."

✅

The reality

Confidence and correctness are unrelated for generative AI. Always verify what matters.

❌

Misconception

"Generative AI looks up the answer like a search engine."

✅

The reality

It predicts new content from patterns; it is not retrieving a single correct source.

🎯

Lesson Purpose

Lesson 1.4 Purpose: This is the lesson the educator leads most directly. Make responsible AI practical: privacy, transparency, attribution, honesty, plus knowing the real risks and the habit of verification. Close on representation: who an AI serves and who it leaves out. Remind students they will apply all of this during the Unit 3 build.

📖

Learning Objectives

📖 Students will know:

Responsible practices: privacy, transparency, attribution, honesty
The risks and threats of generative AI
Why verifying outputs matters
What representation means in AI

⚡ Students will be able to:

List responsible building habits
Name real risks of generative AI
Explain why verification matters
Ask who an AI serves and who it excludes

🔑

Quiz Answer Key: Lesson 1.4

Q1. Which is a responsible practice when building with AI?

Be transparent and give attribution

Transparency and attribution are core to trustworthy work.

Watch for: Common wrong - C: never enter personal or sensitive data into AI tools.

Q2. Which is a real risk of generative AI?

Deepfakes and misinformation

These are concrete, current risks students should be able to name.

Watch for: Common wrong - B/C/D: AI does not automatically protect privacy or remove bias.

Q3. Why does verifying AI outputs matter?

Because AI can hallucinate, spread misinformation, and be unfair

Verification protects the people who use what students build.

Watch for: Common wrong - A: verification is essential, not optional.

Q4. 'Representation' in AI is mostly about:

Whose data, languages, and perspectives are included or missing

Representation determines who an AI serves well and who it leaves out.

Watch for: Common wrong - A/C/D: representation is not about hardware or speed.

Q5. When you build, a good question to ask is:

Who does this serve, and who might it exclude

This question operationalizes responsible building during the project.

Watch for: Common wrong - C: hiding sources is the opposite of responsible practice.

💬

Discussion Prompts: Lesson 1.4

Who is left out

Think of an app you use. Who does it work well for, and who might struggle with it? How would you fix that?

Look for: Look for students applying the representation idea to a real product and proposing inclusive improvements.

Your pledge

What responsible habits will you commit to when you build your Unit 3 project?

Look for: Look for specific, checkable habits: protect privacy, verify outputs, disclose AI use, consider who is served.

⚠️

Common Misconceptions: Lesson 1.4

❌

Misconception

"Responsible AI is just a warning at the start that you can skip."

✅

The reality

Responsible AI is a habit applied in every step: checking bias, verifying outputs, and considering who is served.

❌

Misconception

"If AI made it, no one is responsible for the result."

✅

The reality

A human is always responsible. You review, verify, and decide before anything you build goes out.

📚 Program Alignment

AWS AI Explorers

Student Journey · Part 1: Introduction to AI

Module 1.1 Why AI Matters
Module 1.2 Fundamentals of AI and ML
Module 1.3 Fundamentals of Generative AI
Module 1.4 Responsible AI

This is the Learn phase of the learn, build, present arc. It sets up Unit 2 (Explore) and the PartyRock build in Unit 3.

Responsible AI

Woven Through Every Lesson

Formally introduced in Lesson 1.4, led by the educator
Bias, privacy, and verification raised in 1.1 to 1.3 in context
Carried into the Unit 3 build: check bias, verify outputs, ask who is served

🎯 Differentiation Strategies

Unit 1 serves learners with no AI background and learners who already use AI heavily. These strategies keep all of them engaged without changing the objectives.

🌱 Scaffolded Support

For students who need more support

Do the "Spot the AI" activity as a class before independent work
Provide a printed key-terms sheet from the Unit Overview flip cards
Sentence starters for the "challenge worth solving" activity
Pair students for the verify activity in Lesson 1.3

📘 Core Pathway

For most students

All four lessons in order with the Knowledge Check after each
One written response to a discussion prompt per lesson
Each student records a community challenge for the Unit 3 build
Review any quiz question scored below 3 of 5

🚀 Extension Challenges

For students ready for more

Research a real example of AI bias or a deepfake and present how it happened
Compare two AI tools on the same prompt and document the differences
Draft a one-page "responsible AI guide" for a younger student
Sketch two features for their future PartyRock app

📅 Pacing Guide

Unit 1 fits a single camp day or two shorter sessions. Adjust to your schedule.

Lesson	Self-Paced (Online)	Camp / Cohort	Key Checkpoint
1.1 Why AI Matters	~40 min async	40 min: discussion + challenge activity	Quiz 1.1 done; a community challenge written
1.2 AI & ML	~35 min async	35 min: Spot the AI activity	Quiz 1.2 done
1.3 Generative AI	~35 min async	35 min: try-and-verify with a live tool	Quiz 1.3 done
1.4 Responsible AI	~40 min async	40 min: educator-led responsible-AI discussion	Quiz 1.4 done; responsible-builder habits written

📌 Facilitator tip: Lesson 1.4 is best run as a live discussion, not solo reading. Use the discussion prompts and the misconception table to surface real student beliefs before they build in Unit 3.

❓ Facilitator FAQ

Do students need any AI or coding background for Unit 1?+

No. Unit 1 assumes no prior AI knowledge and no coding. It starts from the AI students already use every day and builds the concepts they need before the no-code PartyRock build in Unit 3.

Why am I, the educator, asked to lead the responsible-AI discussion in 1.4?+

Responsible AI is judgment, not facts to memorize, so a live, human-led conversation works best. The AWS AI Explorers model has the educator lead these segments. Use the discussion prompts and misconception table in the Lesson 1.4 tab as your script.

Should students use live AI tools during Unit 1?+

Where you have tool access, yes, especially for the Lesson 1.3 try-and-verify activity. If access is limited, the lessons stand on their own and tools become essential in Units 2 and 3.

What should every student leave Unit 1 with?+

A written community challenge they care about (from Lesson 1.1) and a short list of responsible-builder habits (from Lesson 1.4). Both feed directly into the Unit 3 project, so collect and keep them.

How do the quizzes work?+

Each lesson has a 5-question Knowledge Check that gives instant feedback and a score. Full answer keys with explanations and common wrong answers are in the lesson tabs above. Use scores below 3 of 5 as a signal to revisit a concept, not as a grade.