“The noise in human peer reviews is crucial because it arises from the variability in human experience and practical knowledge — two features that no model’s training corpus can fully capture,” writes Giorgio F. Gilestro in Nature about the introduction of AI peer reviews. #Academicsky 🧪