// NORTH · The Logbook · Year Binder

FULL-YEAR LOGBOOK · SPARK.

Every module's teaching plan, ready to print as a single PDF and bind for the school year. One artifact a teacher can keep on their desk — works even if Wi-Fi goes down.

// Track

// Table of contents

CONTENTS.

01How Computers ThinkWeek 1 · 2 sessions
02Your First Lines of CodeWeeks 2–3 · 4 sessions
03Decisions & Loops · Build a Text GameWeeks 4–5 · 4 sessions
04Working with Data · Story & Song GeneratorsWeeks 6–7 · 4 sessions
05What is AI, Really?Weeks 8–9 · 4 sessions
06Build Your First AI ClassifierWeeks 10–11 · 4 sessions
07Ethics & ResponsibilityWeek 12 · 2 sessions
08Capstone: Ship Something You LoveWeeks 13–16 · 8 sessions

// Module 01 · SPARK

HOW COMPUTERS THINK.

Duration: Week 1 · 2 sessions (90 min)

// Learning objective

Understand what code is, what a program does, and how computers follow instructions.

// Standards alignment · AI4K12 5 Big Ideas

Bearing NE · Big Idea 2: Representation & Reasoning (How AI Thinks)

// Lessons

1	What is a Program?	45m
2	Instructions Matter (Order)	45m

// Teaching plan

Big idea

A program is an exact list of instructions a computer follows in order — and order matters.

Essential questions

What is a program?
Why do computers need 'exact' instructions when humans don't?
What happens when the order is wrong?

Warm-up

10 min

Sandwich algorithm

Have one student write 'how to make a peanut butter and jelly sandwich' on the board. Teacher follows the instructions LITERALLY (don't open the jar unless told, etc.). Class laughs and revises. Reveal: 'You just wrote your first program. The bugs you saw are the same bugs every coder hits.'

Mini-lecture

15 min

Computers don't guess

A program is a list of instructions in a specific order.
Computers do EXACTLY what you say — even if it's wrong.
The word for 'order matters' is sequence.
Python reads like English, which is why we use it.

Guided practice

15 min

Hello, World together

Open the in-platform Python playground on the projector. Type `print("Hello, World")` together. Run it. Now have students type their own name. Demo one bug (missing quote) on purpose and ask 'why?'

Independent practice

25 min

Self-introduction program

Students complete the in-platform exercise and mini-project ('morning routine program'). Walk the room. Hint at decomposition: 'how would you break the routine into smaller steps?'

Exit ticket

5 min

In one sentence: what is a program?
Give one example of a real-world program (don't say 'an app').
Why does order matter?

Vocabulary

program · Python · print() · string · sequence · bug

Materials

NORTH platform open on student devices (or one shared device per pair)
Projector for live coding demo
Print copy of this plan (no other handouts needed)

Differentiation

Support: Pair students. Have the more-fluent reader type while their partner narrates the step out loud.

Challenge: Ask them to write a 'morning routine' program where one wrong order produces visibly silly output. Share with the class.

Homework / extension

Find one program/app you used today and write down 3 instructions you think it follows in order.

Safety & ethics note

No student logins — they work in the public playground. Nothing they type is saved or shared.

// Facilitator notes

// Module 02 · SPARK

YOUR FIRST LINES OF CODE.

Duration: Weeks 2–3 · 4 sessions (180 min)

// Learning objective

Write working Python code: print, variables, simple math. No installation.

// Standards alignment · AI4K12 5 Big Ideas

Bearing NE · Big Idea 2: Representation & Reasoning (How AI Thinks)

// Lessons

1	Hello, World	45m
2	Variables Hold Things	45m
3	Math in Code	45m
4	Make the Computer Talk Back	45m

// Teaching plan

Big idea

Variables are labeled boxes that hold values. Naming them well is one of the most important skills in coding.

Essential questions

What is a variable, and why do we need them?
Why are good variable names so important?
What's the difference between text and a number in Python?

Warm-up

10 min

Empty backpack

Pull out 5 unlabeled lunch boxes. Ask: 'whose is whose?' Then label them. Discussion: when you have 100 things, labels stop being optional — that's a variable.

Mini-lecture

15 min

Boxes with labels

A variable is a labeled box. The `=` puts a value INTO the box.
Variable names use lowercase with underscores: `first_name`, not `firstName` or `FirstName`.
Numbers don't need quotes. Text does.
You can change what's in the box — read `score = score + 10` as 'new score = old score + 10'.

Guided practice

20 min

Student ID card together

On the projector, build a 'student ID card' with variables: name, grade, favorite_class, fun_fact. Print each with a label. Have a student come up and edit their own.

Independent practice

25 min

Build your own ID card

Students complete the in-platform mini-project. Walk the room. Compliment good variable names out loud — this is how the value of naming gets internalized early.

Exit ticket

5 min

Why is `student_age` a better variable name than `x`?
What does `score = score + 10` mean?
Which is text and which is a number: `"42"` vs `42`?

Vocabulary

variable · assignment · identifier · int · str · concatenation

Materials

NORTH platform open on student devices
Projector
Optional: 5 unlabeled boxes for the warm-up demo

Differentiation

Support: Provide a fill-in-the-blank starter with variable names already created.

Challenge: Add a 'days_until_birthday' variable. Have them compute it from `current_year` and `birth_year`.

Homework / extension

Find 3 places at home that 'use' variables (e.g., a price tag, a thermostat, a score on a game).

Safety & ethics note

No student data leaves the playground. Encourage fictional 'ID cards' if students are uncomfortable using real info.

// Facilitator notes

// Module 03 · SPARK

DECISIONS & LOOPS · BUILD A TEXT GAME.

Duration: Weeks 4–5 · 4 sessions (195 min)

// Learning objective

Use if/else and loops to make programs that adapt and repeat — and ship your first text-adventure game.

// Standards alignment · AI4K12 5 Big Ideas

Bearing NE · Big Idea 2: Representation & Reasoning (How AI Thinks)

// Lessons

1	If This, Then That	45m
2	Comparing Things	45m
3	Doing Things Over and Over	45m
4	Mini-Project: Choose-Your-Own Adventure	60m

// Teaching plan

Big idea

Programs that just run top-to-bottom are boring. Real programs make decisions (if) and repeat themselves (loops) — those two ideas power every app on Earth.

Essential questions

When does code branch into different paths?
How can one program treat 1 user and 1 million users the same way?
What's the difference between checking equality and assigning a value?

Warm-up

10 min

Choose-your-own-line

Two lines at the door: 'pizza' or 'taco'. Ask each student. Then add a 3rd option, then a 4th. That's elif/elif/elif/else. Visceral demo.

Mini-lecture

20 min

If/else and for loops

`if`, `elif`, `else` let your code branch.
`==` asks 'equal?' — `=` puts a value IN. Don't mix them up.
A `for` loop walks through a list and runs the same code for each item.
Loops + decisions are how AI 'trains' — they loop over data millions of times.

Guided practice

20 min

Movie rating + attendance counter

Build the movie-rating if/else together on the projector. Then build the attendance counter loop. Run both with sample data. Predict-then-run for each.

Independent practice

30 min

Movie rating + attendance mini-projects

Students complete both in-platform mini-projects. Walk the room — watch for `=` vs `==` bugs and praise students who catch their own.

Exit ticket

5 min

What does `if score >= 70:` mean?
What's the difference between `=` and `==`?
Write the FIRST and LAST lines of a for loop that prints every name in a list called `students`.

Vocabulary

conditional · boolean · comparison · indentation · loop · iterate · range()

Materials

NORTH platform on student devices
Projector
Two paper signs ('pizza', 'taco') for warm-up

Differentiation

Support: Pre-fill the if-else skeleton. Students just fill in the condition and body.

Challenge: Add a `with_parent` override on the movie-rating project, or rewrite the attendance counter using `enumerate()`.

Homework / extension

Write the if/else logic (in English, not code) for whether a friend can come over: must have permission AND finished homework AND it's before 8pm.

Safety & ethics note

Movie ratings touch on age — be sensitive if any student is anxious about being 'too young' for things. Frame as 'the rule', not personal.

// Facilitator notes

// Module 04 · SPARK

WORKING WITH DATA · STORY & SONG GENERATORS.

Duration: Weeks 6–7 · 4 sessions (195 min)

// Learning objective

Lists, dictionaries, and reading real data — the foundation of every game, creative tool, or dataset.

// Standards alignment · AI4K12 5 Big Ideas

Bearing N · Big Idea 1: Perception (How AI Sees)
Bearing NE · Big Idea 2: Representation & Reasoning (How AI Thinks)

// Lessons

1	Lists Hold Many Things	45m
2	Dictionaries Map Things	45m
3	Looking at Real Data	45m
4	Mini-Project: Mad-Libs Story Generator	60m

// Teaching plan

Big idea

Real programs don't just hold one value at a time — they juggle lists and maps of values. Lists keep order; dictionaries label things by name.

Essential questions

When does order matter, and when does a label matter more?
What shape does real data come in?
How do you compute one number from many?

Warm-up

10 min

Backpack vs. yearbook

Hold up a backpack and a yearbook. Backpack = list (order: zipper pouch, then book, then snack). Yearbook = dictionary (look up by NAME). When does each shape make life easier?

Mini-lecture

15 min

Lists are ordered. Index starts at 0.
Dictionaries map a key to a value. No order — you look things up by name.
Real data is almost always a LIST OF DICTIONARIES (a spreadsheet in code).
The most common pattern: loop → accumulate → compute.

Guided practice

20 min

Class survey on the projector

Build a list of dicts together: 5 imaginary students with name + minutes-read-yesterday. Loop, sum, divide. Predict-then-run.

Independent practice

30 min

Class survey mini-project

Students run a real survey of 10 classmates and code up the dict + 3 stats. Walk the room — common bug: forgetting that survey strings need int() before math.

Exit ticket

5 min

When use a list vs. a dictionary?
What index is the 1st item in a list?
Pattern to compute an average from a list of dicts?

Vocabulary

list · index · append · dictionary · key · value · items() · len()

Materials

NORTH platform on student devices
Projector
Optional: paper survey sheets for the homework

Differentiation

Support: Provide the survey dict pre-filled. Student writes only the average-calculation loop.

Challenge: Add a 'which student read most?' query that returns the name, not just the number.

Homework / extension

Run a real 10-person survey on something you actually care about (favorite lunch, hours of sleep). Bring the responses to next class.

Safety & ethics note

Use first names + last initials only. Never collect emails, phone numbers, or addresses in a class survey.

// Facilitator notes

// Module 05 · SPARK

WHAT IS AI, REALLY?.

Duration: Weeks 8–9 · 4 sessions (180 min)

// Learning objective

AI demystified — pattern matching, training data, why models get things wrong.

// Standards alignment · AI4K12 5 Big Ideas

Bearing N · Big Idea 1: Perception (How AI Sees)
Bearing E · Big Idea 3: Learning (How AI Learns)
Bearing S · Big Idea 5: Societal Impact (AI in the Real World)

// Lessons

1	AI vs. Magic	45m
2	Patterns Everywhere	45m
3	What Training Data Is	45m
4	Why AI Gets Things Wrong	45m

// External tools (optional)

Google Teachable Machine: Free, browser-based — students train a classifier with their webcam. Optional in-class demo.

// Teaching plan

Big idea

AI isn't magic. It's a pattern-finder. If the data is biased, the AI will be biased — and that's why responsible AI work starts with the data, not the model.

Essential questions

What does an AI actually do?
Where do AI 'mistakes' come from?
Who is responsible when an AI is wrong?

Warm-up

10 min

Pattern guessing game

Show 5 examples: 'cat', 'dog', 'hamster' → PET. 'tiger', 'shark' → NOT PET. Ask: 'is goldfish a pet?' Vote. Then ask: 'how did you decide?' Reveal — that's exactly what AI does.

Mini-lecture

20 min

Training data → model → prediction

Every AI follows the same pipeline: data → train → predict.
The model only knows what was in the training data.
Biased data → biased model. (Show the Amazon resume story.)
Watch the 3Blue1Brown video together — it's the best visual ever made of how a neural network works.

Guided practice

20 min

Tiny AI in 6 lines

Build the in-platform tiny AI (positive/negative word classifier) on the projector. Then have one student suggest a word — predict, then run.

Independent practice

25 min

Pet-vs-not-pet classifier

Students complete the in-platform mini-project. Test it on words NOT in the data. Discuss: what do you do when the AI doesn't know?

Exit ticket

5 min

In your own words: what does an AI actually do?
If you trained an AI on only photos of golden retrievers, what would happen if you showed it a poodle?
Whose fault is it when an AI gets something wrong?

Vocabulary

AI · model · training · prediction · training data · bias · pattern matching

Materials

NORTH platform on student devices
Projector with audio for 3Blue1Brown video
Optional: Google Teachable Machine open for a live demo (browser only)

Differentiation

Support: Provide the dictionary half-filled. Students add 3 more entries and test.

Challenge: Have them build a 3-category classifier (positive/negative/neutral) and discuss when 'neutral' is ambiguous.

Homework / extension

Find one news story about an AI that got something wrong. Be ready to share: what data do you think it was trained on, and why did that cause the problem?

Safety & ethics note

Be careful with examples — avoid biases that affect students in the room (race, gender, ability). Use the Amazon resume story as a SAFE distant example.

// Facilitator notes

// Module 06 · SPARK

BUILD YOUR FIRST AI CLASSIFIER.

Duration: Weeks 10–11 · 4 sessions (180 min)

// Learning objective

Write Python that classifies things based on rules — the foundation of every model, game AI, and recommender.

// Standards alignment · AI4K12 5 Big Ideas

Bearing NE · Big Idea 2: Representation & Reasoning (How AI Thinks)
Bearing E · Big Idea 3: Learning (How AI Learns)

// Lessons

1	Rule-Based Classifier	45m
2	Making the Rules Better	45m
3	Testing What You Built	45m
4	Where Rules Break Down	45m

// Teaching plan

Big idea

Before machine learning, you can already build a classifier with if/else rules — and you'll understand WHY ML exists when you hit the limits of rules.

Essential questions

Can you classify without machine learning?
What's a false positive? A false negative?
When do hand-written rules stop working?

Warm-up

10 min

Spam or not spam

Project 6 example messages. Have the class vote spam or not for each. Now ask: 'what rules did your brain use?' Capture them on the board. Reveal — you just wrote a rule-based classifier.

Mini-lecture

15 min

Classifying = picking a label for a thing.
Rules-first is a real engineering choice. Gmail STILL uses rules alongside ML.
Two ways to be wrong: false positive (cried wolf) and false negative (missed the wolf).
Test → fail → fix is the engineering loop, forever.

Guided practice

20 min

Build is_spam() together

On the projector, build the 3-rule spam classifier. Run it on 4 test cases. Find a false positive together. Add a rule. Re-run. Celebrate.

Independent practice

30 min

Positive/negative review classifier

Students complete the in-platform mini-project: a sentiment classifier using +/- word scoring. Walk the room — push them to test on edge cases ('not good', sarcasm).

Exit ticket

5 min

What's a false positive?
When do rules start to break?
Why is testing as important as writing the rules?

Vocabulary

classifier · rule · false positive · false negative · accuracy · edge case

Materials

NORTH platform
Projector
Whiteboard for capturing student-generated rules

Differentiation

Support: Provide 3 positive + 3 negative words pre-loaded. Student writes the scoring loop only.

Challenge: Handle negation ('not great' should NOT be positive). Discuss why this is genuinely hard.

Homework / extension

Find one real message in your texts/email that you'd classify as 'spam' and one as 'not spam'. Write down 2 rules that would tell them apart.

Safety & ethics note

Don't have students paste real emails/messages into shared screens. Use generic examples.

// Facilitator notes

// Module 07 · SPARK

ETHICS & RESPONSIBILITY.

Duration: Week 12 · 2 sessions (90 min)

// Learning objective

When AI helps, when it hurts, and who decides — for games, creative tools, and community apps.

// Standards alignment · AI4K12 5 Big Ideas

Bearing S · Big Idea 5: Societal Impact (AI in the Real World)

// Lessons

1	When AI is Unfair	45m
2	What You Owe Your Users	45m

// Teaching plan

Big idea

AI doesn't get to choose what it learns — the people who build it do. With that power comes a responsibility to test fairly and tell the truth.

Essential questions

What does 'fair' actually mean when you measure it?
Whose fault is it when an AI hurts someone?
What can you, personally, do about it?

Warm-up

10 min

The two doctors

Tell the class: 'I'm thinking of a doctor.' Ask each student to draw what they pictured (30 sec). Reveal: research shows kids draw mostly men. Discuss WHY — not a moral lesson, a pattern lesson. Same thing happens inside AI.

Mini-lecture

20 min

Bias = mistakes that fall on one group more than another.
'High accuracy' isn't enough — accuracy for WHOM?
Three commitments of an honest AI builder: tell the truth, test fairly, make it stoppable.
Walk through one real example: ProPublica's COMPAS investigation (age-appropriate, distant).

Guided practice

15 min

Audit a fake model together

Project a tiny made-up dataset of accuracy by group (girls 95%, boys 80%). Ask: 'is this model fair? what would you change?' Discuss as a class — there's no single right answer.

Independent practice

25 min

Group reflection + 1-page commitment

In small groups, students write their own 'AI builder's pledge' — what they promise their community when they build something. Share out loud.

Exit ticket

5 min

Define bias in one sentence.
Why is overall accuracy not enough?
Name one of the three commitments.

Vocabulary

bias · fairness · audit · harm · accountability · transparency

Materials

Paper for drawings + pledges
Projector for example dataset

Differentiation

Support: Pair students. One narrates, the other writes the pledge.

Challenge: Have them apply the pledge to a SPECIFIC AI they've used (TikTok feed, Photomath). What would change?

Homework / extension

Find one news story about an AI that hurt someone. Be ready to share who was harmed and why.

Safety & ethics note

Avoid bias examples that touch students in the room. Stick to historical / distant cases. Frame as 'how engineers solve this,' not 'who's to blame.'

// Facilitator notes

// Module 08 · SPARK

CAPSTONE: SHIP SOMETHING YOU LOVE.

Duration: Weeks 13–16 · 8 sessions (405 min)

// Learning objective

Build, document, and present a Python program you're proud of — a game, a creative tool, or a community helper. Your choice.

// Standards alignment · AI4K12 5 Big Ideas

Bearing S · Big Idea 5: Societal Impact (AI in the Real World)
Bearing SE · Big Idea 4: Natural Interaction (How AI Talks With Us)
Bearing E · Big Idea 3: Learning (How AI Learns)

// Lessons

1	Pick Your Lane & Your Project	60m
2	Build Sessions (4×)	240m
3	Writing the README	45m
4	Project Showcase	60m

// External tools (optional)

GitHub (instructor-managed): Instructor creates a class organization; capstone code is pushed there for showcase.

// Teaching plan

Big idea

The capstone is your first piece of real engineering — small enough to finish, real enough to matter, specific enough to a person you actually know.

Essential questions

Who is this for, by name?
How will you know it works?
What's the smallest version you can finish?

Warm-up

10 min

Steel-thread the demo

Show them last cohort's strongest capstone (or one from the public showcase). Ask: 'who do you think they built it for? What would 'success' have looked like to that person?'

Mini-lecture

15 min

Three rules: real person, real outcome, finishable in 4 sessions.
Write the one-pager BEFORE you write any code.
README is your project's front door — it gets read more than your code.
Demos that succeed: 30-sec story, 60-sec live demo, 60-sec next steps.

Guided practice

25 min

Live scoping

Pick one student volunteer. Project their (rough) idea. Use the 3-question template to tighten it down to something doable. Class watches the editing process — that's the lesson.

Independent practice

240 min

4 build sessions

Across 4 sessions: (1) scope + one-pager, (2) first running version, (3) iterate based on user test, (4) README + demo prep. Instructor does a 5-min 1:1 each session.

Exit ticket

5 min

Who SPECIFICALLY does your project help?
What's the one sentence that describes it?
What does 'done' look like for you?

Vocabulary

capstone · scope · README · MVP · demo · user test

Materials

NORTH platform
Projector
Print copy of the one-pager template

Differentiation

Support: Provide a list of 5 'safe' project shapes (calculator, translator, classifier, helper bot). Student picks one + names the person.

Challenge: Push them to user-test with the real person and revise based on actual feedback.

Homework / extension

Talk to the person you're building for. Ask 3 questions. Bring notes.

Safety & ethics note

Scan capstone topics for sensitive ones (mental health, immigration, undocumented status). Route those through the ethics module BEFORE building. No real names, faces, or PII of the 'someone you're helping' in the code or README.

// Facilitator notes