Research
Scientists who use AI propose less distinctive ideas for federal research grants
PsyPost · By Eric W. Dolan ·
The Brief
- A new study published in PNAS finds that scientists who use generative AI to help write federal grant applications propose research topics that more closely resemble recently funded projects.
- The study refers to National Institutes of Health submissions, but the source excerpt is cut off before its findings are complete.
- The source describes the result as less distinctive research ideas, not as a measure of grant success or quality.
Why it matters
If AI-assisted proposals converge on topics already funded, the diversity of federally supported research could narrow, which matters for researchers planning proposals and for funders judging novelty.
Who’s affected
- researchers applying for federal research grants
- research funders and grant reviewers
- policymakers interested in how AI affects science funding
- #generative ai
- #research funding
- #grant writing
- #nih
- #scientific publishing
Scientists who use generative artificial intelligence to help write federal grant applications tend to propose research topics that more closely resemble recently funded projects, according to a new study published in PNAS . At the National Institutes of Health, submissions with…