Choose your track

The situation

You have a build that works and a workspace that holds your standing instructions. The next question is not what to type. It is which part of your job to point all of it at.

Three people reach this unit in the same week. A director of manufacturing whose real recurring decision is which two of eleven proposed pilots get budget. An operations VP who spends her week getting four hundred people to change how they close out a shift. A learning lead asked to write the memo saying where the company should start. Their badges say manufacturing, operations, learning. The decision each of them owns says something else.

Pick by the badge and you will build the wrong artifact very well.

The principle

Which AI work you should learn is set by the decision you own, not by the title on your badge.

Ask one question: what call lands on your desk repeatedly, and what goes wrong when you get it wrong? If it is where scarce budget and attention go first, you belong in Leadership. If it is whether several hundred people are still working differently a month from now, you belong in L&D and People. If it is what to do about a reading from a machine, a route, or a line, you belong in Physical and Industrial Operations.

Each track is three lessons and ends in one page you keep.

Leadership — a one-page first-move brief

For the person who picks where AI goes first and then defends that pick to people who have sat through demos before. Taught by Leah Davis, who led change programs at a global strategy consultancy across consumer goods and industrial manufacturing, and built the training function inside each one.

L&D and People — a rollout plan

For the person accountable for whether a workforce is still using the thing after the session ends. Taught by Johann Marquez, a product and engineering leader who turns one team's working practice into training the rest of an organisation can actually run.

Physical and Industrial Operations — a signal-to-decision map

For the person whose work has weight, a location, and a shift schedule, where being wrong costs downtime rather than a rewrite. Taught by Tony Toteda, who ran product and supply-chain operations for hardware launches and real-time systems at scale, and has trained both executives and front-line teams.

The tracks are independent. Nothing in one assumes you have read another, and plenty of readers own two of these decisions. Take them in any order. The capstone in Unit 6 is graded against the task you named in Unit 0, not against your track — so start with the page you would have had to write this quarter anyway, and treat it as a deliverable for your job rather than as coursework.

Try it

Before you open a track, finish this sentence in writing: "Every week or month I decide ______, and when I get it wrong, ______ happens."

You are done when the first blank names a decision rather than a task, and the second is something you could describe to someone outside your function in one line. If you wrote "I run the weekly report", that is a task — go up one level, name who acts on it and what they act on. Then match it to the three descriptions above. If two fit, take the one whose artifact is already overdue.