ทันเอไอ

Nature Study Shows AI Accelerates Individual Researchers' Careers, but Narrows Overall Scientific Scope

A major study finds that while AI tools accelerate scientists' success, they lead to a worrying contraction in the overall scope of new knowledge discovery.

A significant study published in the world-renowned journal Nature reveals the dual impact of artificial intelligence (AI) on the scientific community. It points out that AI has become a powerful tool helping individual researchers achieve career success more rapidly, but simultaneously narrows and reduces the diversity of the overall direction of scientific inquiry.

The research team, led by James Evans, a sociologist from the University of Chicago, analyzed over 310,957 research papers across six natural science fields (biology, medicine, chemistry, physics, materials science, and geology), using a Google language model to identify studies that utilized AI. The study found that researchers using AI published 3.02 times more papers than those who didn't, and their work received 4.85 times more citations. Furthermore, they became lead researchers an average of 1.37 years faster than their peers.

However, looking at the scientific community as a whole revealed negative impacts. The adoption of AI led to a 4.63% reduction in the number of research topics studied and a 22% decrease in interactions among researchers. The study authors concluded that "AI tools appear to automate existing fields rather than explore new ones," creating a paradox: "expanding individual scientists' impact while contracting the overall reach of science."

James Evans summarized this phenomenon as "You have this conflict between individual incentives and science as a whole". Interestingly, this pattern has consistently appeared from the early days of machine learning and deep learning to the current era of generative AI, reflecting the structural challenges facing science in the age of AI.