Classify the output as a forecast or generated instruction: Before selecting evaluation and governance
First determine whether the system forecasts an event or generates technician-facing content.
The question
An energy company is evaluating a maintenance system that predicts which turbine will require service next month. A proposed change would replace that prediction with automatically drafted technician instructions. Before selecting evaluation and governance controls, what should be established about the proposed output?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Determine whether technicians prefer the drafted instructions to the existing maintenance forecast.Preference testing may inform usability after classification, but it does not establish whether the output is predictive or generated.
- Classify the output as a forecast or generated instruction. ✓A forecast estimates a future or unknown event; a generated instruction composes new content for a user. The distinction determines the relevant evaluation and governance questions.
- Measure predictive accuracy and writing quality before deciding which output category applies.Those metrics may be relevant after classification, but the category follows from what the system produces, not from its measured performance.
- Confirm that historical turbine records remain the principal data source for the revised system.Historical data can support both forecasting and content generation, so input provenance does not determine output type.
The trap
Classify what the system produces, not merely the data it uses or its performance score. How to remember it
First determine whether the system forecasts an event or generates technician-facing content.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding the Foundations of AI Governance questions
- Inspect reward traces across representative scenarios: Before approval, which evidence most directly tests →
- Trace predictions through staff outreach: Before approval, which evidence best determines whether review must →
- Calibration results plus agent override performance: Which evidence is most decision-relevant? →
- All 337 Understanding the Foundations of AI Governance questions →
Part of the Certsqill AIGP question bank · Understanding the Foundations of AI Governance ·
Every answer, right and wrong, comes with its own explanation.
