Prompt Engineering Techniques AIE Practice Questions (Page 7)

Questions 31–35

1. 31expert · hard

   A data science team uses an AI to generate Python code for data cleaning. They have a set of test cases that the code must pass. The initial prompt 'Write code to clean the data' produces code that fails on missing values. The team wants to systematically improve the prompt. Which approach is most effective?

   • AAsk the AI to write more robust code without specifying what is missing.
   • BProvide a few examples of clean data and ask the AI to mimic the pattern.
   • CRun the generated code against the test cases, identify the specific failure (missing values), and refine the prompt to explicitly mention handling missing values, then retest.
   • DSwitch to a different AI model that is better at code generation.
   Reveal Solution Discussion Select an answer first
2. 32application · medium

   A project manager is evaluating whether an AI-generated project plan meets requirements. The plan must include milestones, resource assignments, and risks. The AI output includes all three but the milestones are not specific. How should the manager evaluate the prompt's effectiveness?

   • AAccept the plan and manually add specific dates later.
   • BRate the prompt as partially effective and refine it to require specific dates and deliverables for each milestone.
   • CReject the prompt entirely and start over with a different AI tool.
   • DConsider the prompt effective because all required sections are present.
   Reveal Solution Discussion Select an answer first
4. 33application · medium

   A customer support team uses an AI chatbot to handle refund requests. The prompt currently says: 'If the customer is unhappy, offer a refund.' This has led to inappropriate refunds for minor issues. How should the prompt be refined to reduce undesired outcomes?

   • AChange the prompt to 'Always offer a refund if the customer is upset' to ensure customer satisfaction.
   • BRemove the refund instruction entirely and let the AI decide based on conversation context.
   • CKeep the prompt but add 'be conservative' at the end.
   • DAdd a policy section that defines eligible refund conditions and instruct the AI to ask clarifying questions when criteria are unclear.
   Reveal Solution Discussion Select an answer first
5. 34application · medium

   A financial analyst wants an AI to summarize a quarterly report. The summary must be under 200 words and include a bullet list of key metrics. Which prompt best achieves this?

   • AGive me the main points of the report.
   • BProvide a summary of the quarterly report in under 200 words, using bullet points for key metrics.
   • CWrite a detailed analysis of the report.
   • DSummarize the quarterly report.
   Reveal Solution Discussion Select an answer first
6. 35application · medium

   A content writer uses an AI to generate blog intros. The first prompt 'Write an intro about cloud computing' yields a generic paragraph. After adding 'make it engaging', the intro is still not specific. What is the most effective next step in iterative refinement?

   • AAdd more adjectives like 'exciting' and 'innovative' to the prompt.
   • BReplace 'engaging' with specific criteria such as 'start with a surprising statistic, mention a common challenge, and end with a question'.
   • CUse the first generic intro and move on.
   • DAsk the AI to 'write like a professional blogger'.
   Reveal Solution Discussion Select an answer first