Speakers: Matteo Valleriani (Max Planck Institute for the History of Science/ TU Berlin/ Tel Aviv University/ BIFOLD), Oliver Eberle (TU Berlin/ BIFOLD), Hassan El-Hajj (Max Planck Institute for the History of Science)
With the establishment of digital humanities, Computational History has emerged as a new discipline, providing a natural context to integrate historical inquiries with AI and its logic. Within the research project "The Sphere", AI models were initially utilized to aid in the analysis of historical sources, facilitating the generation of large datasets and transitioning history writing into a data-driven science. Subsequently, new AI architectures and models were developed to analyze these datasets, which exhibit high degrees of heterogeneity and dimensionality. These AI models can now autonomously generate analyses of general trends and genealogies of historical sources, forming the foundational structure for constructing historical arguments. The talk will show how these models, along with the application of eXplainable AI, are paving the way for the development of an AI historian.