Skip to Main Content

AI Prompt Engineering

This guide provides information and resources for helping you to understand how to conceive and execute effective prompts to elicit optimal results from artificial intelligence tools. The guide is co-created and co-owned by Brian Quinn and Erin Burns.

Defining Prompt Engineering

Prompt engineering is designing specific instructions or queries, called prompts, to guide AI models' behavior and output. (Alice Ivey, Cointelegraph)

Prompt Engineering Frameworks

CLEAR Framework for AI Prompts:

Concise-use clear simple language, prioritize critical information, be succinct, remove irrelevant details (don’t want to exhaust token limits of AI in its response)

Logical-structure information in order, establish context and relationships, ensure logical flowing narrative

Explicit-define instructions and format, specify length, sources, terminology, set tone, reading level restrictions, provide examples

Adaptive-be flexible, rephrase and restructure, adjust setting and parameters, try different approaches, look at setting and parameters, try different approaches

Reflective-carefully evaluate A.I. responses, identify areas for improvement, use insights to further refine strategies for engagement

Leo S. Lo,
The CLEAR path: A framework for enhancing information literacy through prompt engineering,
The Journal of Academic Librarianship,
Volume 49, Issue 4,
2023
https://doi.org/10.1016/j.acalib.2023.102720 

Prompt Engineering Tips and Best Practices

When prompting AI tools:

1. Be Specific

2. State Your Intention

3. Break Down Bigger Tasks into Smaller Tasks

4. Specify the Output

5. Provide Examples

6. Experiment with Phrasing

7. Ask Follow Up Questions (Iteration)

 

 

AI Prompt Tricks

If unsure of what prompt to use, AI prompt engineers will ask an AI chatbot "Please generate a prompt that I can use with an AI to provide me with...." followed by whatever outcome they are seeking to obtain from the AI.

Prompt Engineering Resources

Prompt Engineering Research