Peer-reviewed preprints and the Publish-Review-Curate model (www.coalition-s.org)

The emergence of peer-reviewed preprints and the Publish-Review-Curate (PRC) model marks a significant shift in the scientific publication landscape, addressing ongoing criticisms of traditional publishing methods. These models support authors in sharing their research more openly and transparently. Authors publish preprints on open platforms, which are then subjected to formal review by specialized services, with the reviews made publicly accessible. The PRC model adds a curation step, where selected articles are given visibility through journals or platforms, enhancing their recognition. This approach is exemplified by eLife, which, since 2023, has abandoned accept/reject decisions in favor of providing editorial assessments and public reviews alongside preprints. For the AI and ML community, this evolving model is crucial as it promises to democratize access to research findings, facilitating faster dissemination and collaboration. However, significant ambiguities surrounding the concepts of peer review and curation need to be carefully navigated. Critics argue that peer reviews are often misinterpreted as validations, potentially misleading readers about the research's integrity. Furthermore, while curation is generally positive, it does not inherently equate to validation, which can cause confusion. By advocating for binary validation within the PRC framework, such as clear accept/reject decisions post-review, the scientific community can enhance both the clarity and reliability of published research, ultimately enriching the rigor of scientific discourse.