Cornell study finds scientists using ChatGPT publish up to 50% more papers than they did before using AI
A Cornell University study reveals that scientists using ChatGPT and other large language models (LLMs) publish up to 50% more papers than before adopting AI. Non-native English speakers benefit the most, posting up to 89% more. The study analysed over 2 million preprints from ARXIV, BIORXIV and the Social Science Research Network (SSRN). While AI improves productivity and research discovery, it also poses challenges in assessing the scientific quality of AI-assisted papers.
A Cornell University study reveals that scientists using ChatGPT and other large language models (LLMs) publish up to 50% more papers than before adopting AI. Non-native English speakers benefit the most, posting up to 89% more. The study analysed over 2 million preprints from ARXIV, BIORXIV and the Social Science Research Network (SSRN). While AI improves productivity and research discovery, it also poses challenges in assessing the scientific quality of AI-assisted papers.
