Learning roadmap
Product Management in the AI Era
AI builds in a weekend what used to take a quarter. The PM edge now: pick the right thing, spec it clearly, and tell the story that gets a yes.
5 stages~32 h
That leaves 5 stages, ~32 hours of focused work. You can clear it in any order, but the stages compound — try them top-to-bottom your first time.
- Conceptsuser-interviewsjobs-to-be-doneai-research-synthesisproblem-statementsevidence-vs-opinionPractice quizzesLab task
Build a single-page 'Discovery Board': paste raw interview quotes, tag each with a pain point, then drag them into theme columns — with a counter showing how much evidence each theme has.
Build it in the LabCheckpointYou can turn 10 raw customer conversations into 3 problem statements, each backed by at least 5 pieces of evidence.
What you'll be able to do
- Run a discovery sprint that turns 10 customer conversations into 3 validated problem statements.
- Write a spec an AI coding agent can build from on the first pass — acceptance criteria and edge cases included.
- Define a metric tree that tells you within a week whether a launch actually worked.
- Rank 10 competing ideas and name exactly what you're saying no to — and why.
- Defend a roadmap in front of executives with a story, not a spreadsheet.
Or try a related path