Your 30-day plan
Courses end; habits start. This last lesson is a four-week plan that turns the previous thirteen lessons into defaults, one layer per week — each week adding a skill and keeping the previous ones in play. None of it needs more than the tools you already have and twenty minutes of intention a day. At the end of it you will not "know about" AI at work; you will simply work this way.
Week 1 — conversations and rewrites
The foundation layer: briefing and iterating (lessons 2–4). Every email over three sentences gets drafted or tone-checked through the assistant. Every request you make of it carries background and material, and gets pushed back on at least once — "shorter", "less formal", "fix point two" — until correcting the model feels as unremarkable as correcting your own typo. By Friday, write your standing brief and set it (lesson 10); you now know what context you kept repeating, because you spent a week repeating it.
Week 2 — summaries and meetings
The reading layer (lessons 5–6). Every document over five pages gets the triage pass — "one sentence: do I need to read this?" — before it gets your eyes. Every meeting you run gets a two-minute generated agenda; every meeting you attend, your fragments become minutes within the hour, with the "uncertainty goes under open questions" instruction every time. This is the week the reclaimed hours become obvious enough that someone asks what changed.
Week 3 — numbers and facts
The trust layer (lessons 7–8, 12). Bring in spreadsheet asks: one formula explained, one formula written, one pasted-table question — with the sheet verifying every total that travels. Make search-mode-plus-clicked-source your default for factual questions. And run the sycophancy drill once a day on anything you're deciding: "now argue the other side." Week 3 is where the checking habits stop feeling like overhead and start feeling like what competence means with these tools.
Week 4 — your system
The assembly layer (lessons 10, 13). Build your two or three projects with their standing instructions and files. Consolidate the ask library from three weeks of accumulated wins; prune what you didn't reuse. Set the sigh rule formally. And take stock honestly: which delegations earned trust, which stay on the always-verify list, roughly what a week costs you now versus a month ago. That calibration — knowing your map of where the tool is strong, not this course's generic one — is the actual graduation.
Where to go from here
Three good next moves, all free. If this course made you curious what is actually inside the machine — why it hallucinates, what a token is, how "it" finds its meaning — How AI Actually Works is the plain-English tour, built for exactly the reader who finished this one. If your work is content, audiences, or ads, AI for Marketing & Creators applies everything here to that world. And for staying current without the hype cycle, Snack is the daily habit: one AI story at a time, readable in the time the kettle takes. The tools will keep changing under all of us; the briefing, checking, and delegating skills you just built are the part that transfers.