Skip to Main Content

Need Support?

Prompting in Gen AI

A prompt is a set of instructions used to ask an AI tool to perform a task and produce an output. The more specific, explicit, and concise the instructions are, the more successful the resulting outputs will be.

Prompt Examples

Tips for Successful Prompts

A successful prompt (Crabtree, 2024) is comprised of several characteristics:

  • Instruction: Clear and explicit instructions tell the AI tool what action(s) you want it to perform.
  • Context: This information helps the AI tool understand your request's broader scenario or background and helps better tailor and frame its response.
  • Input data: data is the information you provide to the AI tool for processing. It could be a paragraph or a set of numbers, for example.
  • Output indicator: This part of your prompt tells the AI tool how to respond to your request. It might be a stylistic direction or citation style, for example.

Some other tips and tricks to consider trying (University of Calgary Libraries and Cultural Resources, 2023):

  • Provide an example that you have already found or created.
  • Prompt the AI tool to act as an expert using wording like "Guide me…" or "Act as the expert…"
  • Divide significant questions or topics into smaller chunks.
  • Create new chats for unrelated questions or prompts for the AI tool.
  • Prompt the AI tool to ask questions to clarify your desired output.

Image Prompting

How good are your prompting skills? Try it with this Say What You See activity from Google Arts and Culture Experiments.

CLEAR Framework

The  CLEAR Framework consists of five principles that help users “learn to navigate and develop AI-generated content more effectively” and develop critical thinking skills (Lo, 2023).

  • Concise: Be brief and clear in your prompt, using precise language.
  • Logical: Structure prompts coherently to reveal relationships between concepts.
  • Explicit: Provide clear instructions and expectations of output, including format, scope, and content, to reduce irrelevant responses.
  • Adaptive: Be flexible and adjust the prompt as needed.
  • Reflective: Evaluate the performance of your AI prompts and adjust your approach if necessary.

Dive Deeper into Prompting