AI-PAST will develop a modular platform that seamlessly integrates generative AI with historians’ existing research workflows while respecting historical methodology. By encoding historians’ contextual understanding and interpretive approaches directly into AI systems, our platform will not only transform historical research practices but also advance AI development through a humanities-driven embrace of heterogeneity, contextual understanding, and multilingual analysis of historical sources.

Some fifty corpora of tweets and web material gathered between 2013 and 2021, each on an event that seemed, that day, to deserve a study. A few became articles. The others sit on a backup drive, in folders whose names are the only remaining trace of the research that was intended.

The aim of the project “Historical Cultures, Epistemology and Methodology of History in the Digital Age” (CulturHist) is to examine the implicit aspects of the digitisation of the historian’s profession, by focusing on new practices that have received little or no explicit attention from historians, such as the intensive use of cameras in the reading rooms of archive centres or the consultation of digital libraries – practices that have profoundly altered the material and sensory culture of historians. CulturHist aims to shed light on this new material and sensory culture, to identify its potential biases and methodological implications, such as the citation of decontextualised documents consulted online. The project also examines the consequences of this digitisation for the institutions that act as intermediaries, forging links between historians and their archives: museums, libraries and archive services.

A short essay for a publisher’s short books series, taking apart the word “numérique”: its history, the materiality of the immaterial, the betrayed utopia, surveillance, technocriticism, and whether another digital world is possible.

The UNSURE project will try to figure out how European citizens relate to EU democratic institutions at a time when political debate has frequently hinged on the democratic deficit of such bodies. It will focus on sources such as newsgroups, facebook groups and social media.

Reading the setbacks of European integration through the records left by internet forums — Usenet, mailing lists, IRC, blogs, comment threads, social media — to reach the views of citizens who are not specialists of Europe, in France, the United Kingdom and Germany.

I forgot most of them.

A long-term history of European central bankers as an epistemic community, from the debt settlements that followed the First World War to the monetary union — arguing that the community predates the Delors Committee. Submitted from the Universität zu Köln.