Principles of AI ›  Syllabus

Course Syllabus

How this class runs day to day, how your grade is built, and how we use AI, in this class and on the way to the capstone.

Principles of AI Home

How This Course Runs

This is an in-person course: 60 minutes a day, five days a week, for 11 weeks (55 class periods). Every unit follows the same arc, and the guidance fades as you go:

🧩 Activities (guided)
Short, tightly guided builds and experiments that teach one skill at a time, like building a language model from dice. Low-stakes: this is where you're allowed to be bad at something on the way to getting good.
🛠️ Projects (apply it)
Multi-day builds where you apply what the activities taught to something real. In Unit 1 the targets are tightly specified; in Unit 2 you choose your own tasks and tools.
🎯 The Problem (open-ended)
In Unit 3 you serve a real client with an AI-use guide. There's no answer key, and you defend your choices.

The Problem is the end game of the whole course. Every activity and project in Units 1 and 2 is practice for it.

How Your Grade Is Built

Your grade splits between Formative work (Activities, where you practice) and Summative work (Projects, plus the Problem and final check that make up your final exam, where you show what you can do). Projects carry most of the weight.

8%  Activities
17 guided activities
72%  Projects
6 projects, 2 per unit
20%  Problem & Final Check
Capstone Problem + no-device final (the final exam)
Formative — Activities

Activities are where you try things, break things, and learn. They count for a small share of your grade (8%), and they're graded on effort and thinking, not on getting a perfect result the first time.

Summative — Projects, Problem & Final Check

Projects apply a skill to something real and make up most of your grade (72%). The Problem and the no-device final check together are your final exam (20%).

How Points Work

Here's what each kind of work is worth, as a share of your final grade (set by school grading guidelines):

ActivityFormative

Guided and low-stakes. Graded on completion and quality of thinking. This is where you're allowed to be wrong on the way to getting good.

≈0.5% each
ProjectSummative

Multi-day builds with a clear target: a classifier, a study, a career brief. Each ends with a short no-device check so your own understanding counts.

≈12% each
Problem + Final CheckSummative

Your final exam. The capstone Problem has a real client and no answer key, and the no-device final checks that the thinking is yours.

20% total
In short: each Project is worth about 12% of your grade, and the Problem and final check together are worth 20%. Your projects are where the grade is earned, and the final exam is where you prove the thinking is yours.

Course Structure

Principles of AI runs 11 weeks in three units. Each unit follows the same shape: Activities lead to a Project, then more Activities lead to another Project. Unit 3 adds the capstone Problem at the end.

Weeks 1–416 periods
Unit 1 — Inside the Machine
Weeks 4–717 periods
Unit 2 — Working With AI
Weeks 7–1122 periods
Unit 3 — AI in the World (Capstone)

A Few Ground Rules

Activities can be redone anytime before the unit closes. Projects can be revised until the following class session.

Keep an AI use log and an evidence ledger. Note when you used AI, what you asked, what you checked, and where it helped, hurt, or stood in for a person. Unit 3's conclusions come from your own entries.

Some work is no-device: the last day of each project and the final check. That's not a trap. It's how you find out what you can actually explain on your own.

Use AI to learn, not to skip learning. Where an assignment allows it, you can use it, and you check what it gives you. If you turn in something you can't explain, that's treated as plagiarism.

Principles of AI ·  In person · 60 min/day · 11 weeks
🤖 A tool that makes you better, not a replacement for you