We build
AI for History
Supported by Schmidt Sciences
We're looking for collaborators — historians and AI researchers.
philip.torr@eng.ox.ac.ukIt is 2050. Artificial intelligence can now explain its reasoning, simulate complex societies, and help humanity learn from ten thousand years of recorded history. Historians and anthropologists work with AI collaborators that reconstruct lost evidence, test competing explanations, and reveal new insights into how cultures evolve and societies endure. This project lays the foundation for that future.
Humans have long been fascinated by the study of history and the lessons it might teach us. Historians and philosophers have repeatedly emphasized the importance of understanding the past as a key to self-knowledge and societal insight. While recent years have seen remarkable advances in AI scientists and AI scientific co-pilots, the humanities have, by comparison, been largely neglected. This raises a central question: if AI is to benefit all of society, how can we extend its capabilities into the interpretive and causal reasoning domains of the humanities? In particular: could we design an AI historian, a co-worker that assists historians in their reasoning, perhaps even approaching human-level interpretive skill? The aim of this project is to work closely with historians to develop such an AI collaborator, modelling their workflows and reasoning practices. The goal is to make this AI historian co-worker fully open source and publicly available, serving as a new kind of scientific instrument for historical scholarship worldwide.
This project is made possible by AI2050, a Schmidt Sciences initiative launched by Eric Schmidt to support researchers working on the hard problems that must be solved for AI to be hugely beneficial to society by the year 2050.
AI for History Team at the Torr Vision Group
-
Philip Torr FRS FREngProfessor of EngineeringUniversity of OxfordSchmidt Sciences AI2050 Senior Fellow
-
Niclas GriesshaberCo-PI NFDI4Memory Incubator FundsDPhil Student in Economic and Social HistoryUniversity of Oxford
-
Lorenz HufeELLIS PhDFraunhofer HHIUniversity of Oxford
-
Sebastian Oliver EckDPhil Student in Music HistoryUniversity of Oxford
-
Gavin GreifCo-PI NFDI4Memory Incubator FundsDPhil Student in Economic and Social HistoryUniversity of Oxford
-
Zhenfei (Jeremy) YinPostdocUniversity of Oxford
-
Zachary ShinnickVisiting StudentAustralian Institute for Machine Learning (AIML), The University of Adelaide
-
Pieter FrancoisProfessor of Cultural EvolutionDirector of the Computational Humanities LabUniversity of OxfordThe Alan Turing Institute
External Collaborators
-
Yao LuAssistant Professor in NLPUniversity College London
-
Xiaoxi LuoMaster's StudentUniversity of Waterloo and Vector Institute
The AI Historian project is funded by AI2050, a Schmidt Sciences initiative launched by Eric Schmidt to support researchers working on the hard problems that must be solved for AI to be hugely beneficial to society by the year 2050. The support is awarded through Philip Torr, a Schmidt Sciences AI2050 Senior Fellow.
Join and follow us: