So far, there is no interdisciplinary and comparative study which has examined the use of machine learning in different scientific fields, encompassing historical, ethnographic and informatic perspectives. "In this way, the ubiquitous claim of an AI revolution can be critically examined through concrete and detailed analyses,” says Prof. Dr. Jens Schröter of the University of Bonn’s Media Cultural Studies Department, who is also a member of the University’s Transdisciplinary Research Area “Individuals, Institutions and Societies.” What distinguishes this research project, he says, is that it examines exactly what role AI-based data analyses have in relation to existing methodological traditions.
The team is analyzing concrete “focus projects” in different disciplines (film studies, sociology, and geosciences) that span the spectrum of humanities, social sciences, and natural sciences. The goal is to examine how AI methods are used to conduct research in different disciplines. The researchers explore the potentials, but also the limitations, risks, and ambivalences of AI-based methods. The project brings together the expertise of media studies, history of science, and computer science to study the use of specifically data-intensive methods in current academic research.
Historical, media studies, and computer science perspectives
Prof. Dr. Anna Echterhölter from the Institute of History at the University of Vienna is producing an in-depth study of sociology in the history of science and examining the use of data and the definition of classification criteria in historical statistical procedures.
Prof. Dr. Schröter from the University of Bonn is conducting a media archaeological study of the specific technological procedures of AI used in the selected focus projects. Which technologies are used exactly? How have these developed historically? What specific strengths and weaknesses do they bring to the table?
Privatdozent Dr. Andreas Sudmann, currently a visiting professor at the University of Regensburg and future coordinator of the project, is investigating with a media ethnographic approach how the technology is actually used in the focus projects. The methodical use of AI technologies will be observed and analyzed decidedly as media practices, i.e. with a view to processes of storing, generating, processing, distributing and presenting information.
Prof. Dr. Alexander Waibel from the Institute of Anthropomatics and Robotics at the Karlsruhe Institute of Technology is developing a machine learning-based AI system for analyzing and summarizing scientific texts and lectures, which themselves deal with the use of AI-based methods in various fields.
Publications and conferences are planned
The research results are expected to result in publications, including a monograph and an edited volume, to be published in 2026. In addition, two international conferences and several workshops are planned to discuss the results. The computer science group will build and continuously maintain a database on "AI meta-research" to collect the data generated by the project and make it accessible together with the developed text analysis system.
Website of the project: https://howisaichangingscience.eu
Volkswagen Foundation press release: https://www.volkswagenstiftung.de/aktuelles-presse/aktuelles/k%C3%BCnstliche-intelligenz-%E2%80%93-ihre-auswirkungen-auf-die-gesellschaft-von-morgen-rd-10-mio-euro-f%C3%B6rderung
Contact:
Prof. Dr. Jens Schröter
Department of Media Studies
University of Bonn
E-mail: schroeter@uni-bonn.de
