Answers are grounded in this paper and cite their sources.
This article argues that artificial intelligence (AI) can productively act as a co-worker in scholarly editing, not by replacing editorial judgement but by accelerating complex, repetitive tasks that depend on established forms of critical reasoning. Drawing on historical examples of digital scholarly editing, it situates generative AI within a longer trajectory of machine-assisted approaches and distinguishes computational systems that reveal patterns in textual data from those that propose interpretative decisions. Its principal case study is the C21 Editions Project, which experimented with generative AI and retrieval-augmented generation to produce a teaching edition of Chaucer’s The Pardoner’s Prologue and Tale. The experiments found that approximately 64% of AI-generated translation was satisfactory, while only 13% of AI-generated annotation and topic-summary text required editorial intervention when prompts were supported by curated knowledge resources. These findings demonstrate both the potential for reducing editorial labour and the limitations of generative AI, whose outputs remain dependent on expert prompting, source selection and critical evaluation. The article proposes a model of human–AI co-working in which editorial expertise shifts from the repetitive execution of established procedures towards methodological design, supervision and evaluation. It concludes that realising the potential of machine-assisted scholarly editing will require transparent, trustworthy and intuitive human–AI interaction tools and editorial workbenches.
No takes yet. Share an insight, caveat, or question.
Michael Pidd (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: