The M-CaReM-AI Method: Defining Work for AI / Unit 5 MCM 125 review, 2 to 3 minutes
A Card Checked Six Ways
The idea
A worked example, constructed for this course, runs the Renew Loan card through the book's six checks, with an AI draft for comparison.
In practice
Before the team's own draft was finished, an analyst had asked an AI tool to draft the same card from the Loan's rough definition. The comparison is instructive. On thing clarity, both name the Loan. On change clarity, both say update. On property realism, the tool's draft adds a renewal reason field that the library has never collected, and leaves out the policy version entirely. On rule testability, the tool writes members may renew loans up to the allowed limit, which no tester can test without knowing where the limit lives. The team's card names the policy version as the source. On stakeholder recognition, the team shows its own card to two desk staff. One points out that renewals by phone are common for older members, so the card's location section gains the phone channel. On atomicity, the tool's draft also sends the confirmation and updates the member's history in the same card. The team's version keeps those apart. The analyst deletes the invented field, adds the policy version to the tool's draft, and keeps it as a reference for the next card.
Where people slip
The tool's draft failed property realism and testability in exactly the ways the book predicts: an invented field and a missing audit field. It also passed the two checks that matter least. The six checks turned a vague sense that the draft looked fine into specific, fixable faults. Kept as a reference, the corrected draft also shows the next analyst what the tool tends to get wrong with this library's things, which makes the next review faster.
Your place on the trail
Every lesson rests on earlier ones and opens later ones. Tap any to jump.
Unlocks
- Nothing yet depends on this.
Something wrong in this lesson? Flag it · Sources for The M-CaReM-AI Method: Defining Work for AI