Task Suitability Classification Exercise

Context

This is the first formal assessment. It tests your ability to classify work tasks on the AI reliability spectrum — the skill from Outcome 1. You'll classify 8 provided tasks and 3 of your own.

Instructions

Part A: Provided Tasks (8 tasks)

For each task, provide:

• Zone (High, Moderate, Low, Not Suitable)
• Primary failure mode (what would go wrong)
• Verification action (what you'd specifically check)

Before you start: predict your score — how many of the 8 do you think you'll get right? Write the number down. You'll compare after.

1. A project manager needs to convert meeting transcript bullet points into a formatted client status report.

2. A real estate agent wants to generate property listing descriptions from a features list (beds, baths, sqft, neighborhood).

3. An HR coordinator needs to calculate vacation accrual balances for 50 employees based on company policy and start dates.

4. A nonprofit director wants to write grant proposal language about their organization's impact, including specific outcome statistics from their last annual report.

5. A teacher wants to generate parent communication emails about upcoming standardized testing schedules and requirements.

6. A small business owner needs to draft social media posts announcing a seasonal promotion.

7. A compliance officer needs to identify which data privacy regulations apply to their company's new customer data collection form.

8. A sales representative wants to create a presentation summarizing a prospect's industry trends and challenges.

Part B: Your Tasks (3 tasks)

Select 3 tasks from your own AI Task Map (Lesson 2.3) or add new ones. For each, provide:

• Zone with rationale
• Primary failure mode
• Verification action

Your personal task classifications should not contradict the task's stated output type or verification requirements. (A task you described as "requires checking against legal sources" cannot be classified as High Reliability.)

Part C: Self-Assessment Comparison

Compare your predicted score (Part A) to your actual score. Answer: "Was I overconfident, underconfident, or about right? What does that tell me about my ability to judge AI reliability?"

Part D: Elaborative Interrogation

Pick the task you were least certain about. Answer: "Why was it hard to classify? What additional information would make you more confident?"

Bridge

You now have a tested framework for judging task suitability. Module 3 builds on this — once you know a task is in the Green or Yellow zone, you need to know how to prompt for good output and catch the failures in what comes back.