Scientists started publishing more papers after ChatGPT arrived — but many added little value, according to the Cornell Chronicle. A new Cornell study found AI tools sharply boosted output while overwhelming reviewers with polished but mediocre research.
Researchers analyzed over 2 million preprints and found scientists using large language models posted up to 50% more papers. Non-native English speakers saw the biggest gains, especially researchers based in Asia, narrowing long-standing language barriers.
But there’s a catch. AI-written papers with strong writing often failed peer review, breaking the usual link between clarity and quality. Editors, funders and universities now struggle to judge real scientific merit as productivity soars.
FingerLakes1.com is the region’s leading all-digital news publication. The company was founded in 1998 and has been keeping residents informed for more than two decades. Have a lead? Send it to [email protected].