Open Source AI: Month One — The Book That Installs Itself

An executable book. Your AI assistant reads it with you and teaches it — one 20-minute lesson a day for 30 days, from "what's a repo" to your first merged contribution and a tiny AI project of your own.

Most books about AI sit on your hard drive and judge you. This one works for a living.

Open the book's folder in Claude Code (or Cursor, or any agentic coding assistant) and say "start the book." From that moment your AI is the instructor: it greets you by day, teaches each lesson conversationally, runs every exercise with you on your real machine and your real accounts, verifies the day is actually done, and keeps your streak. You don't read about running a local model — you run one.

The 30 days:

• Week 1 — Get your bearings. The ten repos that anchor open-source AI, how to read any AI repo like a native, the feed and communities worth your attention, and the start of your own AI-operated workflow.
• Week 2 — Speak the language. Git without fear, open source vs. open weights in plain English (the license decides who owns your work at 2 a.m.), your first local model on your own hardware, Hugging Face literacy, quantization demystified.
• Week 3 — First contributions. Finding real good-first-issues live, docs PRs as the front door, your first fork → branch → PR with the AI pair-driving, and how to handle review, silence, and your second contribution solo.
• Week 4 — Build and become visible. Your GitHub profile as a portfolio, building in public, shipping your own tiny AI tool with a real release and a real license, and where Month Two goes.

What you get:

• The full 30-lesson course as a git repo — the book IS the software
• The reading editions (169-page PDF + single-file HTML) for humans who like pages
• The verified lists: the First 10 repos, who to follow, where the conversations are — every link checked live at build time
• The complete AI-workflow setup the book installs: starter CLAUDE.md, personal wiki, session logging — yours to keep and grow
• Lifetime updates: the AI ecosystem rots fast — every refreshed edition is yours free

Who it's for: smart, busy people who keep hearing "open source AI" and want in but don't know the door. No CS degree needed. If you can install software and hold 20 minutes a day, the course holds the rest.

Who it's not for: anyone expecting passive reading. The book makes you do things. That's the point.

Honest requirements: a computer you can install software on — any OS; Linux is the home-field recommendation, Windows and Mac carry every lesson fine (8GB+ RAM comfortable, ~20GB free disk for Week 2's local models) — a GitHub account (Day 0 helps you make one), and an AI coding assistant (Claude Code is the reference; any agentic tool works. The book's README installs one with you in five minutes).

Instant digital delivery — your download links arrive at checkout and by email.

Where the money goes: every dollar this store makes buys GPUs and funds independent open-source AI & robotics research. You are not just buying merch — you are funding the commons. Open source must win.