70 Best AI Research Prompts
for Claude & Perplexity

Give me a structured overview of [TOPIC]. I'm starting from [MY CURRENT KNOWLEDGE LEVEL]. Cover: 1. What is it? (1-2 sentence definition that doesn't assume prior knowledge) 2. Why does it matter? (the actual stakes — who cares and why) 3. Key players (people, companies, institutions that matter) 4. Current state (where things stand as of now) 5. The main debate or controversy (where experts disagree) 6. What changed recently (last 6-12 months) 7. Where to go deeper (best primary sources, not just Wikipedia) Be specific. Name names, cite numbers. No "it's complicated" hedging without explanation.Copy

Steelman both sides of the debate on [TOPIC/CLAIM]. For each side: - The strongest version of the argument (not the most common, the strongest) - The best evidence supporting it - Who holds this position and why they're credible - What would need to be true for this side to be right Then: What do the best experts who've genuinely changed their minds say? What evidence caused them to update? What's the crux — if you could resolve one factual question, which position would be vindicated? Don't write a "both sides have valid points" summary. Push toward what the weight of evidence actually suggests.Copy

Audit the quality of evidence behind the claim: "[CLAIM]" For this claim: 1. What type of evidence exists? (RCTs, observational studies, expert opinion, anecdote, correlation) 2. How strong is the best evidence? (sample size, methodology quality, replication) 3. What's the evidence against the claim? 4. Publication bias: is this the kind of claim where negative results don't get published? 5. Who funded the key studies? (potential conflicts of interest) 6. What do the systematic reviews or meta-analyses say (if any exist)? 7. Bottom line: strong / moderate / weak / speculative — and why If you're uncertain about specific studies, say so clearly rather than confabulating citations.Copy

I'm going to give you [NUMBER] sources on [TOPIC]. Synthesize them into a coherent picture. For each source, note: - Source identifier (Author/Publication/Date) - Main argument or finding - Key data points or evidence - How it relates to or conflicts with the other sources Then synthesize: - What do all sources agree on? - Where do they disagree, and what explains the disagreement (methodology, timeframe, definition differences)? - What's the strongest overall conclusion supported across sources? - What's the biggest gap — what question do these sources collectively leave unanswered? [PASTE SOURCES BELOW]Copy

These two sources say contradictory things about [TOPIC]. Help me figure out why and which is more credible. Source A: [PASTE OR SUMMARIZE] Source B: [PASTE OR SUMMARIZE] Analyze the conflict: 1. Are they actually contradicting each other, or talking about different things (different time periods, populations, definitions)? 2. What's the methodological quality of each? 3. What are the incentives or biases of each source? 4. If this is a genuine factual contradiction, what would it take to resolve it? 5. Which source should I update toward, and how much? Give me a credence level: how confident should I be in Source A vs. B after this analysis?Copy

Help me build a literature review on [RESEARCH QUESTION]. Organize what I give you by: **Theoretical frameworks:** The major conceptual approaches scholars use to study this **Empirical findings:** What the data actually shows, organized by finding type (not by paper) **Methodological debates:** Disagreements about how to study this **Consensus vs. contested:** What's settled, what's still debated **Research gaps:** What hasn't been studied, and why it matters **Landmark papers:** The 3-5 papers that changed the field or are most cited Output format: flowing prose organized by theme, not a list of paper summaries. Cite as [Author, Year] inline. [PASTE YOUR PAPERS / SOURCES BELOW]Copy

Research [COMPETITOR NAME] in depth. I'm trying to understand them well enough to compete against them. Cover: 1. Business model (how they make money, pricing structure) 2. Customer segments (who they target, who their biggest customers are) 3. Key strengths (what they genuinely do well) 4. Weaknesses and complaints (what customers say negatively — check reviews, Reddit, Twitter) 5. Recent moves (product launches, fundraising, hiring, acquisitions in last 12 months) 6. Messaging and positioning (what they say about themselves, their brand voice) 7. Where they're vulnerable (segments or problems they don't serve well) Be specific — I need actual data points, not general observations.Copy

Give me a trend report on [INDUSTRY] for [YEAR]. Cover: 1. Macro forces driving change (technology, regulation, demographics, economics) 2. Where money is flowing (what's being funded, what's being bought) 3. What's declining or being disrupted 4. Emerging behaviors or demand signals 5. The biggest debate in the industry right now 6. What industry insiders are worried about that outsiders don't know yet 7. A 3-year outlook: what will look different about this industry in 2028? Prioritize signals over noise. Name specific companies, products, and numbers where possible.Copy

Fact-check this claim: "[CLAIM]" I need to know: 1. Is this claim true, false, misleading, partly true, or unverifiable? 2. What's the primary evidence for or against it? 3. Is this a case where the literal claim is true but the implication is misleading? 4. What's the original source of this claim? (especially if it's being shared secondhand) 5. How confident should I be in this verdict? (high / medium / low) 6. What would change the verdict if new information came in? If you can't verify this claim directly, tell me what I would need to look up to verify it myself.Copy

Sanity-check this statistic: "[NUMBER/STAT]" Questions to address: 1. Does the scale of this number make intuitive sense? (Fermi estimation) 2. What's the source, and how reliable is it? (methodology issues, who funded it, how old) 3. What's the exact population or scope this number applies to — and am I misapplying it? 4. How does this compare to related numbers I can triangulate against? 5. Is this the right metric for what I'm trying to understand, or is there a better one? 6. Is this number being cherry-picked (selected because it makes a point) vs. representative? Give me a confidence level: should I use this stat, use it with caveats, or discard it?Copy

Convert my research notes into an executive summary for [AUDIENCE — e.g., board / CEO / client]. My research notes: [PASTE NOTES] The exec summary should: - Open with the single most important finding (not the background) - Use a maximum of 5 bullet points, each 1-2 sentences - End with a clear recommendation or next step - Be under 300 words total For this audience specifically: - Lead with [BUSINESS/FINANCIAL/STRATEGIC] implications - Avoid: [JARGON OR DETAIL THEY DON'T CARE ABOUT] - They will ask: [ANTICIPATED QUESTION] — preempt itCopy

I'm going to share my research on [TOPIC]. After reading it, identify what's missing. My research: [PASTE YOUR NOTES / SOURCES] Find: 1. Questions my research raises but doesn't answer 2. Counterarguments or counterevidence I haven't addressed 3. Assumptions I'm making that I haven't validated 4. Key stakeholders or perspectives not represented 5. Data I should have but don't 6. The single most important thing missing that would change my conclusion if I found it For each gap: rate how critical it is (must-fill / nice-to-have / out-of-scope), and suggest the fastest way to fill it.Copy