Study finds AI tools boost scientific output but limit research diversity

A new study of 68 million scientific papers reveals that while AI tools help young scientists advance their careers, they also limit the diversity of research topics. Researchers found that AI-assisted papers are more frequently cited but focus on narrower subjects. The study, conducted by the University of Chicago and Tsinghua University, shows that scientists using AI publish 67% more papers and are 32% more likely to lead research teams. However, their work covers 5% less topical ground compared to non-AI research. The findings suggest that AI is leading to repetitive ideas and less interaction among research papers. The authors recommend changes to funding and incentive systems to encourage broader exploration in scientific research.