The M-CaReM-AI Method: Defining Work for AI / Unit 8 MCM 199 review, 2 to 3 minutes
The Change Plan
The idea
A worked example, constructed for this course, turns the library's change-impact walk into a plan, and compares it with an AI tool's suggested impact list.
In practice
The analyst writes one line per affected item, with an owner and an order. First, publish the new Loan Policy Version, owned by finance. Next, update the Create Loan and Renew Loan cards and the overdue notice card, owned by circulation. Then change the overdue job and confirm the council export, owned by the systems administrator. Next come the account page and a note on the monthly report. The tests follow, with renewals covered for the first time. Then the manual and training. Last is the chatbot's material. Finally, update the Loan Pack and its matrix. Before the workshop, the analyst had given an AI tool the matrix and the change, and asked for likely impacts. Comparing the two now, the tool's list had found the cards, the overdue job and the notices. It also proposed changing the loans table to add a period field, which was not needed. And it missed both the council export and the chatbot. The analyst keeps the tool's list as a starting point next time, and records that it missed the council export, which was not obvious from field names.
Where people slip
The tool's list was useful and wrong in both directions: it proposed one unnecessary change and missed two needed ones. That is exactly the book's description of an impact scout. The matrix, walked row by row by people who own each row, was the version the library acted on. The tool's list saved time as a first draft; the matrix decided what was true.
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