Bodo Stern and Michele Avissar-Whiting are right about #preprints:
https://katinamagazine.org/content/article/open-knowledge/2026/better-scholarly-communication-system-rethink-preprints
In the original #arXiv fields (physics, mathematics, CS), authors posted preprints, responded to feedback, and posted new versions. Articles evolved. Preprints coexisted with postprints in the same repository. When preprints were adopted later in the life sciences, the model lost this valuable dimension. Instead, "a preprint [was considered] the rough-cut awaiting journal validation and finishing, not complete until a journal has had its say."
in my spot checks of what AI search actually cites, arXiv papers with several new versions in the same repository still get pinned to v1 surprisingly often, while life-science preprints mostly get folded into the journal record. has anyone seen postprints reliably win the citation?