AI speeds up science by 60x — and researchers are publishing worse papers
a new study models how llms impact research, finding that lower time costs lead scientists to prioritize volume over depth, potentially degrading scientific quality.
researchers from princeton and the university of washington published a theoretical analysis arguing that llms are likely to degrade scientific quality despite increasing efficiency. modeling research behavior as an optimal foraging problem, they found that in two of three scenarios—specifically when ai aids early ideation or final publication—scientists will sacrifice the voluntary 'deep dive' phases of work to produce a higher volume of projects. the authors conclude that as labor-augmenting tools increase the opportunity cost of time, researchers will produce 'more, less well' rather than better work.