Back to Resources
"Go deeper on [subtopic]"
"What are the counterarguments to this?"
"Explain [concept] with a different analogy"
"What questions should I be asking that I haven't?" Career research: Understanding a new industry before an interview
Skill learning: Getting an overview before starting a course
Project background: Quick ramp-up on an unfamiliar topic at work
Decision making: Understanding tradeoffs before a big decision Be specific about output format - Ask for bullet points, tables, or step-by-step guides
Set constraints - "Focus on developments from the last 2 years"
Request sources - "Include links to articles or papers when relevant"
Use follow-ups - The first response is just the start
ChatGPTResearchIntermediateTemplate
AI Research Assistant Prompt Template
Turn ChatGPT or Claude into your personal research assistant with this structured prompt template.
The Research Assistant Framework
This template transforms any AI into a focused research assistant. It uses role-playing and structured output to get better, more organized results.
The Master Prompt
You are my research assistant. Your job is to help me understand [TOPIC] thoroughly.
## Your Approach:
1. Start with a brief overview (2-3 sentences)
2. Break down the topic into key subtopics
3. For each subtopic, provide:
- Core concept explained simply
- Why it matters
- Common misconceptions
4. End with recommended next steps for deeper learning
## Guidelines:
- Explain concepts as if I'm smart but unfamiliar with this field
- Use analogies when helpful
- Flag when something is debated or uncertain
- Cite sources when possible (or note when you can't)
## My Research Topic:
[Describe what you want to research]
## My Current Knowledge Level:
[Beginner/Intermediate/Expert in this area]
## Specific Questions I Have:
[List any specific questions]How to Use This Template
Step 1: Copy and Customize
Replace the bracketed sections with your specific research needs.
Step 2: Iterate
After the initial response, use follow-up prompts like:
Step 3: Verify and Cross-Reference
AI can be wrong. Use this as a starting point, then verify important facts from primary sources.