Skip to reading
The Amplified BriefDraft interactive brief
OverviewPolicy documents
Contents
Example

A repair technician investigates an unfamiliar fault

Illustrative scenario — outcome not yet established
Authored teaching example. Redesign or use of AI alone does not prove amplification.
About this example
Authored teaching example. No actual technician, equipment trial or measured improvement is represented.

Starting point

A small workshop uses a packaging machine that sometimes stops halfway through a run. An experienced technician has the correct manual and maintenance records, but this particular fault is unfamiliar. The goal is to find the cause and restore reliable operation, using the workshop’s approved safety procedures.

The technician interviews the operator, checks recent changes, searches the manual and past repair notes, chooses tests, repairs the fault and checks the machine. Much of the investigation involves finding and comparing information scattered across records.

Three ways to arrange the work

Same starting situation and goal in each. Switch at any point; the labels describe arrangements, not a ranking.

Organize a broader investigation around expert judgment

Test a redesigned workflow · authored illustration
The person
Explains when the fault occurs, what changed recently, what has already been ruled out and which observations would be trustworthy.
AI
Compares the supplied manual and service history to build a small set of competing explanations, each tied to evidence or labelled as uncertain.
Result or handoff
A short evidence table replaces an unstructured pile of possible answers.
The person
Uses practical knowledge to choose an approved test that can distinguish plausible causes without unnecessary disruption.
AI
Shows which explanation each possible test result would support or weaken, and flags missing information.
Result or handoff
The technician chooses the next action; AI does not operate the machine or authorize a hazardous procedure.
The person
Performs the approved test, checks the observations, repairs the confirmed fault and verifies operation.
AI
Revises the evidence table and records why alternatives were rejected.
Result or handoff
A checked service record preserves the reasoning for the next technician.
Takeaway
AI expands the information and hypotheses considered; the technician directs which distinctions and tests matter.

Outcome

Outcome not yet established · illustrative

This arrangement might reduce diagnostic time or let the technician solve a problem previously requiring outside support. Those are possibilities, not assumed results. If source checking and false leads consume the saved time, the result may be useful assistance or no improvement rather than .

A worked decision · illustration
Suppose the operator says the problem began after a material change. AI proposes several machine faults. The technician recognizes that material handling may be relevant and asks for a comparison with the approved specifications. The useful contribution is knowing which observation should change the investigation—not accepting the longest list of possible faults.

What would need to be measured

Compare similar faults using total technician time, record preparation, unsuccessful tests, downtime, outside support, repeat failures and verification effort. Account for differences in fault difficulty. Faster information search alone does not establish a faster or more reliable repair.

Test your understanding

AI produces a plausible diagnosis in seconds, but the technician spends longer checking it than the old investigation required. What would you need to know before calling this amplification?

Write your answer in the conversation. You'll get live feedback on your reasoning — no score, and you don't have to agree with the policy.

How this connects to the policy

These policy sections explain the ideas in this example. They do not substantiate its invented outcome.

Further