Quo vadis, Clio? Assessing Use and Abuse of AI Workflows in Historical Research

This project proposes to explore a new avenue for the methodological advancement of historical research by implementing and critically assessing the massive integration of AI tools into scholarly workflows, aiming at the work environment of ‘classical’ historians who have no particular background in Digital History.

Since the release of ChatGPT the latest, we all have become aware of the massive disruptive potential of generative Artificial Intelligence (AI) in general and Large Language Models (LLMs) in particular. We all have made our own experiences with both the amazing capabilities and with equally amazing shortcomings of these new technologies.

What we need now is a systematic Technology Assessment (Technikfolgenabschätzung) and the development of new best practices in dealing with these new technologies. While AI tools can facilitate various aspects of research, they also pose risks, such as producing biased results or encouraging approaches that violate the principles of good academic practice. The project aims to promote critical reflection on these technologies and provide practical insights for their responsible use in the field of history.

The project is built around an experimental historical case study, which will be dissected into discrete research steps and workflows. For each step, the project team evaluates where AI tools can be implemented, assesses the advantages and challenges they bring, and considers their compliance with scholarly standards.

Special attention is given to identifying potentially abusive uses of AI, such as shortcuts that compromise academic integrity or generate misleading interpretations. Moreover, the project seeks to establish principles and guidelines for employing AI in ways that align with the highest standards of good scholarly practice.

Quo vadis, Clio is a highly unconventional and interdisciplinary research project that combines the revision of ‘classical’ methods of historical research with the introduction of AI workflows and, above all, a Technology Assessment approach. The transformative potential of this project lies in its ability to advance informed methodological innovation within the historical discipline. It aims to broaden the scope of critical engagement with AI technologies beyond specialists in the field of Digital History, addressing the implications for historical research as a whole.