Human Ideas, Artificial Intelligence Processing: A Proposal for Redesigning Scientific Production and Scholarly Publishing

Scientific publishing still consumes substantial researcher time in formatting, resubmission, repetitive data entry, and other tasks that do not themselves create knowledge. I propose a human–AI division of scientific labour in which humans retain responsibility for questions, hypotheses, interpretation, ethics, and final judgement, while artificial intelligence undertakes literature processing, data organisation, drafting, verification support, formatting, and submission mechanics. The aim is not to replace scientific judgement but to protect scarce human cognitive time. A four-stage Human Idea–AI Processing–AI Verification–Human Judgement (HAIV) model and an AI-native submission protocol are proposed.