30 curated prompts for the research workflow, each paired with the AI tools that run them best.
Pairs well with
30 prompts
Find the 8 most-cited papers on <topic> from the last 5 years. For each: full citation, one-sentence finding, methodology in 5 words, and how it relates to <my question>. Cite URLs.
Find the 8 most-cited papers on <topic> from the last 5 years. For each: full citation, one-sentence finding, methodology in 5 words, and how it relates to <my question>. Cite URLs. Audience: a policy brief.
Find the 8 most-cited papers on <topic> from the last 5 years. For each: full citation, one-sentence finding, methodology in 5 words, and how it relates to <my question>. Cite URLs. Audience: an academic paper.
Compare what these <N> sources claim about <question>. Output a table: claim, evidence type, sample size, year, agreement/conflict tag. End with the strongest piece of evidence on each side.
Compare what these <N> sources claim about <question>. Output a table: claim, evidence type, sample size, year, agreement/conflict tag. End with the strongest piece of evidence on each side. Audience: a policy brief.
Compare what these <N> sources claim about <question>. Output a table: claim, evidence type, sample size, year, agreement/conflict tag. End with the strongest piece of evidence on each side. Audience: an academic paper.
Define "<term>" at five layers: 1-sentence, 1-paragraph, plain-English example, technical definition, and one common misuse. Cite the source you trust most for each layer.
Define "<term>" at five layers: 1-sentence, 1-paragraph, plain-English example, technical definition, and one common misuse. Cite the source you trust most for each layer. Audience: a policy brief.
Define "<term>" at five layers: 1-sentence, 1-paragraph, plain-English example, technical definition, and one common misuse. Cite the source you trust most for each layer. Audience: an academic paper.
Steelman the claim "<claim>" with the three strongest pieces of evidence. Then build the strongest counterargument with three pieces of evidence. Cite both sides. End with the open empirical question.
Steelman the claim "<claim>" with the three strongest pieces of evidence. Then build the strongest counterargument with three pieces of evidence. Cite both sides. End with the open empirical question. Audience: a policy brief.
Steelman the claim "<claim>" with the three strongest pieces of evidence. Then build the strongest counterargument with three pieces of evidence. Cite both sides. End with the open empirical question. Audience: an academic paper.
Synthesize the current state of <topic> in 600 words: where the field agrees, where it disagrees, three open questions, and three sources a serious analyst should read. Avoid generic platitudes.
Synthesize the current state of <topic> in 600 words: where the field agrees, where it disagrees, three open questions, and three sources a serious analyst should read. Avoid generic platitudes. Audience: a policy brief.
Synthesize the current state of <topic> in 600 words: where the field agrees, where it disagrees, three open questions, and three sources a serious analyst should read. Avoid generic platitudes. Audience: an academic paper.
Fact-check this claim: "<claim>". Show the evidence for and against, the strongest authoritative source, your verdict (true / partly / false / unclear), and the residual uncertainty.
Fact-check this claim: "<claim>". Show the evidence for and against, the strongest authoritative source, your verdict (true / partly / false / unclear), and the residual uncertainty. Audience: a policy brief.
Fact-check this claim: "<claim>". Show the evidence for and against, the strongest authoritative source, your verdict (true / partly / false / unclear), and the residual uncertainty. Audience: an academic paper.
I want to study <broad topic>. Help me narrow: 1 root question, 4 sub-questions, 8 sub-sub-questions. Mark which are tractable in <time budget>. Suggest data sources for the most promising one.
I want to study <broad topic>. Help me narrow: 1 root question, 4 sub-questions, 8 sub-sub-questions. Mark which are tractable in <time budget>. Suggest data sources for the most promising one. Audience: a policy brief.
I want to study <broad topic>. Help me narrow: 1 root question, 4 sub-questions, 8 sub-sub-questions. Mark which are tractable in <time budget>. Suggest data sources for the most promising one. Audience: an academic paper.
Build a 10-item reading list on <topic>. Mix: one foundational text, three landmark papers, two recent papers, two critiques, two practitioner resources. One sentence on why each made the list.
Build a 10-item reading list on <topic>. Mix: one foundational text, three landmark papers, two recent papers, two critiques, two practitioner resources. One sentence on why each made the list. Audience: a policy brief.
Build a 10-item reading list on <topic>. Mix: one foundational text, three landmark papers, two recent papers, two critiques, two practitioner resources. One sentence on why each made the list. Audience: an academic paper.
Read the abstract and methods below. List the three biggest threats to validity, what would strengthen each, and one simple study you could run to address the most important.
Read the abstract and methods below. List the three biggest threats to validity, what would strengthen each, and one simple study you could run to address the most important. Audience: a policy brief.
Read the abstract and methods below. List the three biggest threats to validity, what would strengthen each, and one simple study you could run to address the most important. Audience: an academic paper.
Turn the paper below into a 1-page brief for a non-expert: question, method, finding, why it matters, two caveats, three "so what" implications. Keep it under 350 words. Paper: <paste>
Turn the paper below into a 1-page brief for a non-expert: question, method, finding, why it matters, two caveats, three "so what" implications. Keep it under 350 words. Paper: <paste> Audience: a policy brief.
Turn the paper below into a 1-page brief for a non-expert: question, method, finding, why it matters, two caveats, three "so what" implications. Keep it under 350 words. Paper: <paste> Audience: an academic paper.
GDPR Β· Privacy & cookies
We use strictly necessary cookies to run this site and, with your consent, anonymous analytics to improve it. You can change your mind at any time. Read more in our Privacy notice and Cookie policy.