To generate effective responses from AI, prompts should be clear, specific, and provide sufficient context. Refining the prompt may be necessary to achieve optimal results.
Principles of Effective Prompting:
This taxonomy provides a structured overview of generative AI prompt types, drawing on categories identified by Grammarly (2023) and instructional approaches adapted from Mascellino (2023). Each prompt type is defined by its functional role and illustrated with an example, supporting the development of effective prompts across domains such as education, content creation, and human–AI interaction.
Prompt Type | Description | Example | |
Completion |
|
“Complete this sentence: Prompt engineering is…” | |
Question / Informational |
|
“What are the key principles and techniques used in prompt engineering to optimize AI model outputs?” | |
Instructional |
|
“Write a step-by-step guide on effectively employing prompt engineering strategies when working with AI systems.” | |
Classification | Prompts that require the AI to categorize information based on specified criteria or into predetermined classes. | “Are the reviews of the Dune movie in major publications generally positive, negative, or mixed?” | |
Comparison | Prompts that involve evaluating similarities and differences between two or more items, often with an emphasis on performance or suitability. | “Compare the effectiveness of prompt engineering in enhancing AI model performance versus using unstructured prompts.” | |
Creative | Prompts designed to elicit original, imaginative, or artistic content, such as stories, poems, advertisements, or speculative scenarios. | “Imagine a future where prompt engineering has revolutionized artificial intelligence. Write a short story describing this transformation.” | |
Dialogue | Prompts that request the generation of a conversation between two or more individuals, often to illustrate multiple perspectives or simulate discussion. | “Generate a conversation between two AI researchers discussing the significance of prompt engineering in improving model transparency.” | |
Interactive / Role-play | Prompts that simulate a contextual interaction or scenario, often requiring the AI to assume a specific role or persona. | “Pretend you’re a hiring manager at a software company, and I am interviewing for a QA position. Conduct the interview.” | |
Summarization | Prompts that ask the model to condense a larger body of text or information into a shorter, more concise summary while preserving the core ideas. | “Summarize the benefits and challenges associated with prompt engineering in three concise points.” | |
Translation | Prompts that involve converting text from one language into another while maintaining semantic accuracy and tone. | “Translate the term ‘prompt engineering’ into French.” | |
Persona / Format-based | Prompts that include meta-instructions about the AI’s role, intended audience, or desired output format, shaping the model’s tone, style, and structure. | “You are a technical writer. Create a bullet-point guide explaining prompt engineering to non-specialist stakeholders.” |
Bibliography
Grammarly. (2023, July 11). Generative AI prompts: What they are and how to use them. Grammarly. https://www.grammarly.com/blog/ai/generative-ai-prompts/
Adapted from Mascellino, A. (2023, September 2). How to write AI prompts: ChatGPT, Bard, Bing & more [+examples]. Techopedia. https://www.techopedia.com/how-to-write-ai-prompts