At its core, effective prompting requires specificity, context and end product.
For example: Ineffective prompt: "Please give me a topic for my environmental law paper."
Effective prompt: "Act as a law student. You are writing a research paper for Environmental Law and need help coming up with a topic. You are interested in climate change. Provide a list of ten topics related to climate change." Continue to revise and refine your query with the LLM ("What about a New York City topic?")
The ineffective prompt will give you one topic that may not work for you at all. The effective prompt gives you more specific answers in a particular context.
A Legal Research Prompting Guide and Generative AI System Comparison Exercise, by Sarah Gotschall. This article encourages using the “RICE” method in your prompt engineering, an easy shortcut to remember.
Mastering AI the 5 Ps of Prompting, by Susan Guthrie. A method of prompting engineering that uses 5 P words (Prime, Persona, Privacy, Product, Polish) to develop an effective prompt.
Mastering AI Prompts for Legal Professionals Practical Strategies and Tools, by Jennifer L. Wondracek. A detailed explanation of prompting, and a discussion about how all gen AI products are not the same; nor do they all have the same capabilities.
Mastering ChatGPT: How to Craft Effective Prompts (Full Guide + Examples)
Best practices for prompting with the OpenAI API