Too Slow for the AI Age? Building a Dynamic Research Continuum

Academic publishing is misaligned with the tempo of contemporary artificial intelligence. Large models alter capabilities, risks and regulatory questions on a quarterly cycle, while journal review and publication often take years. This tempo gap displaces epistemic authority from public, peer-reviewed knowledge to corporate laboratories, private platforms and informal commentary. The paper argues that closing this gap requires redesigning the research pipeline itself. It proposes a Dynamic Research Continuum (DRC): a continuous, versioned publication architecture that integrates preprints, open and AI-assisted review, and “living” synthesis articles into a single visible knowledge loop. Unlike piecemeal open-science reforms, DRC treats access, review and synthesis as one lifecycle explicitly designed for high-velocity AI domains. The architecture is operationalized through the Iterative Human–AI Co-Creation (IHACC) framework, which specifies how human researchers and AI systems cycle through scoping, drafting, critique and updating while preserving judgement, transparency and epistemic inclusion. Using a historical reconstruction of journal publishing, a diagnosis of current challenges in AI-relevant fields and an analysis of reform options, the paper argues that timely, trustworthy and openly updated research has become a precondition for legitimate governance in the AI age.