Opus 4.8 Beats Human Experts as Peer Reviewers in a Harvard Study

“As academics, we need to figure out how to adapt the practice of science to AI, fast.”

By every measure, Opus 4.8 beats human experts as peer reviewers, @emollick reports, in a study of 124 Harvard social science papers. The papers range from statistical analyses to formal models to qualitative work.

“This is not going to be the only area in which AIs beat humans,” he adds, linking the paper.