Artificial intelligence is giving individual scientists a leg up—boosting paper output, citations and career progression—while potentially narrowing the terrain of scientific inquiry, according to an analysis in Nature. Reviewing 41 million publications, with about 311,000 using AI methods, researchers found AI adopters publish more and are cited more often but cluster in established domains rather than venturing into new fields. The findings suggest AI may automate what’s already popular, concentrating resources and attention and risking less disruptive breakthroughs. Funding agencies and journals may need new incentives and guardrails to counterbalance these systemic effects, even as labs continue to reap personal productivity gains from AI.
Related articles:
Large teams develop and small teams disrupt science and technology
NIH Prohibition on the Use of Generative AI for Peer Review
ChatGPT: five priorities for research
Self-driving laboratories for materials discovery
