Interview Prompting

Interview Prompting involves designing a prompt that encourages the language model to ask clarifying questions to better understand the user’s needs or the desired output.

This technique is particularly useful when the initial request is vague or when additional information would help produce a more tailored response. It creates a more interactive experience, similar to a conversation with a helpful expert.

Pros:

• Useful when the user’s needs or the desired output are not fully defined
• The model’s questions can help clarify requirements
• Guides the interaction towards a more useful outcome
• Creates a more conversational and interactive experience

Cons:

• Can feel less direct and may require multiple turns
• The model’s questions might not always be relevant or helpful
• Takes more time than a single-prompt approach
• May not be necessary for straightforward requests

Example:

Prompt:

I need help planning a weekend trip.

Possible Model Question:

Great! To help me plan, could you tell me what kind of destination you're interested in (e.g., city, beach, mountains) and what your approximate budget is?

Interview prompting works best for:

• Complex requests with undefined parameters
• When the user isn’t sure exactly what they need
• Tasks that benefit from customization
• Planning and advisory scenarios
• When gathering requirements is an important first step

For a more structured approach to prompt design across various use cases, prompt frameworks provide reusable templates.