Science thrives on curiosity, doubt, and the courage to ask questions that challenge existing knowledge. By fostering this mindset, scientists unlock new avenues of exploration and discovery. Albert Einstein wrote: “The important thing is not to stop questioning. Curiosity has its own reason for existing.” And then this wonderful statement from Nobel Laureate Richard Feynman, “I would rather have questions that can’t be answered than answers that can’t be questioned.”
Kokoro begins with asking questions that are important to you and your organization. We offer the following Module, excerpted from our course on Resilient Community Leadership, to help you learn and perfect the Art of Prompt Engineering.
The INNER QUADRANT is the most obvious, with primary axes of:
- LEFT to RIGHT being PROBLEM <> SOLUTION axis, articulated in different ways
- UP and DOWN being GENERAL <> SPECIFIC axis articulation
The OUTER QUADRANT is the less obvious, with primary axes of:
- LEFT to RIGHT being PRETEXT <> SPECIFIC ACTION axis, articulated in different frameworks
- UP and DOWN being PRINCIPLES <> PRACTICES axis articulation
Often I will ask questions that focus on inner quadrant and then move on into outer quadrant, e.g.
INNER:
- L/R: What are the likely challenges or causes of ‘A’, and what are the likely results of, or solutions to, ‘A’?
- U/D: What are the likely examples or elements of ‘A’, and what are the likely issues that ‘A’ is about?
OUTER:
- L/R: What is the historical precedence of ‘A’, and formulate an action plan and task list to respond to ‘A’.
- U/D: What are key ideas and big picture of ‘A’, and provide me with a list of supporting resources and use cases of ‘A’
OpenAI Prompt Engineering Recommendations Prompt engineering – OpenAI API
Strategy: Write clear instructions
- Tactic: Include details in your query to get more relevant answers
- Tactic: Ask the model to adopt a persona
- Tactic: Use delimiters to clearly indicate distinct parts of the input
- Tactic: Specify the steps required to complete a task
- Tactic: Provide examples
- Tactic: Specify the desired length of the output
Strategy: Provide reference text
- Tactic: Instruct the model to answer using a reference text
- Tactic: Instruct the model to answer with citations from a reference text
Strategy: Split complex tasks into simpler subtasks
- Tactic: Use intent classification to identify the most relevant instructions for a user query
- Tactic: For dialogue applications that require very long conversations, summarize or filter previous dialogue
- Tactic: Summarize long documents piecewise and construct a full summary recursively
Strategy: Give models time to “think”
- Tactic: Instruct the model to work out its own solution before rushing to a conclusion
- Tactic: Use inner monologue or a sequence of queries to hide the model’s reasoning process
- Tactic: Ask the model if it missed anything on previous passes