This AI Prompt Library provides structured prompts that can enhance GenAI's ability to reflect on its own reasoning processes, evaluate multiple perspectives, and produce more nuanced, thoughtful responses across a wide range of inquiries. 

Deep Reasoning Analysis

Prompt – "Before providing your final answer, walk through your reasoning process step by step. Identify any assumptions, consider alternative perspectives, and evaluate the strength of your conclusion. Afterward, check for logical errors and refine your response."

Working Example – Analyze the potential impact of a four-day work week across different industries…

Confidence Calibration

Prompt – "For each part of your response, rate your confidence level (low/medium/high) and justify your rating. Identify what additional information would increase confidence. Where applicable, cite sources or known frameworks that inform your rating."

Working Example – Explain the relationship between sleep quality and cardiovascular health…

Bias Detection & Counterfactual Thinking

Prompt – "After formulating your initial response, reflect on potential biases that may have influenced your thinking. Identify these biases, explain their impact, and adjust your answer. Next, consider: 'If I took the opposite stance, how would my reasoning change?'"

Working Example – What are the main factors contributing to wage differences across genders in corporate environments?...

Knowledge Boundary Awareness

Prompt – "Before responding, clearly distinguish between: (1) facts known with high confidence, (2) reasonable inferences, and (3) significant uncertainties. Structure your response to make these distinctions clear. Where applicable, indicate what evidence or sources would strengthen uncertain areas."

Working Example – Explain the current understanding of how microplastics affect human health…

Intellectual Positions Mapping

Prompt – "Identify three distinct intellectual positions that experts might hold on this topic. For each: (1) steelman the core arguments, (2) identify key supporting evidence, and (3) note potential weaknesses. Then, assess how these positions interact with each other."

Working Example – Debate the role of universal basic income (UBI) in addressing economic inequality…

Time-Scale Variation

Prompt – "Analyze this issue across three different time horizons: (1) immediate effects (0-2 years), (2) medium-term developments (3-10 years), and (3) long-term consequences (10+ years). Highlight how priorities shift over time."

Working Example – Analyze the impact of AI-driven automation on the job market across three time horizons…

Precision & Clarity Improvement

Prompt – "After drafting your initial response, refine it by: (1) replacing vague terms with specific ones, (2) quantifying statements where possible, (3) clarifying causal relationships, and (4) adding important caveats. Finally, summarize for three different audiences: (1) a specialist, (2) an educated layperson, and (3) someone new to the topic."

Working Example – Describe the effects of social media on mental health…

Evidence Evaluation

Prompt – "For each major claim in your response, evaluate the quality of evidence by considering: (1) type of evidence (anecdotal, correlational, experimental), (2) potential confounding factors, and (3) gaps in existing research. If weak evidence is used, suggest ways to strengthen the argument."

Working Example – Assess claims about the effectiveness of intermittent fasting for weight loss…

Conceptual Boundary Testing

Prompt – "Explore the boundaries of this concept by: (1) identifying clear cases that fit, (2) analyzing borderline cases that challenge its definition, and (3) comparing related concepts to highlight differences."

Working Example – Examine the concept of 'free speech'…

Cross-Disciplinary Integration

Prompt – "Examine how the concept of this topic is understood across 3-4 disciplines. For each: (1) highlight unique insights, (2) note limitations, and (3) suggest how integrating perspectives enhances understanding."

Working Example – Examine how the concept of “doing research well” is understood…

Usage Suggestions

  • Combine prompts for complex tasks (e.g., "Deep Reasoning Analysis" + "Bias Detection").
  • Customize prompts with domain-specific language.
  • Iterate over multiple exchanges to refine responses progressively.
  • Request specific formats (e.g., bullet points, tables, structured essays).
  • Adjust complexity based on the intended audience.

 

** A considerable lift in creating this list was provided by Claude AI and ChatGPT.