How is technology transforming historical research and interpretation? The Digital Futures of History panel, chaired by Professor Steven Connor, brought together experts from digital humanities, modern history and art history to discuss the promises and pitfalls of AI.
Professor Stuart Dunn, Professor of Spatial Humanities in the Department of Digital Humanities at King’s, compared two extremes – the logic of algorithms trained on billions of datasets and the complete absence of logic inherent in superstition.
What I want to provoke us to consider is a comparison between AI-generated content -the logic of the algorithm, driven by theoretically limitless computational power over billions of training datasets across the Web – with the complete absence of logic inherent in superstition. Is there a meaningful difference between the two?
Professor Stuart Dunn, Professor of Spatial Humanities
Professor Dunn warned that as AI-generated content becomes harder to distinguish from authentic sources, some people may treat it as ground truth, even when it fills gaps in historical knowledge.
He argued that historians and researchers need to critically examine how AI influences knowledge and culture. The challenge, he said, is ensuring that digital tools help us understand the past without replacing evidence with illusion.
In 2025, AI has made tremendous progress, and historians can do a few things with it now. But AI is not a good historian.
Dr Vincent Hiribarren, Reader in Modern History
Dr Vincent Hiribarren, Reader in Modern History, shared an experiment using AI to analyse archives on the war of Cameroon (1945-1971). ‘It took me three to four days to read the whole batch of documents as a human being. It took less than five minutes for AI,’ he noted, while cautioning against confirmation bias and speculative answers when models are pushed beyond factual limits.
Dr Hiribarren tested prompts to summarise documents, identify key actors and interpret events. While AI proved fast and was able to decipher handwritten texts, it also had significant limitations: vague answers to complex questions, confirmation bias when repeating language from sources and speculative interpretations.
‘Historians use the present to ask the questions of the past. But as the saying goes, the past is a foreign country, and we need to be careful with that’, Dr Hiribarren warned.
Audience at the Digital Futures of History event. Image credit: Richard EatonProfessor Stephen Whiteman, Professor of the Art and Architecture of China at the Courtauld, addressed the challenges of applying AI in art history. Art history, he says, is particularly interested in the types of challenges and questions that the digital raises because the digital itself is a form of media or mediation.
Professor Whiteman noted that while techniques like network analysis, spatial analysis and computer vision offer new possibilities, their success has been mixed due to the nature of questions art history asks and the challenges of large, uncurated datasets: ‘Implicit in the building of a lot of the data sets in art history are histories of institutions, histories of collecting, often built on histories of colonialism, so the data sets are problematic in precisely the ways that art history as a conventional practice seeks to unpack.’
Digital tools, Professor Stephen Whiteman argued, can complement traditional scholarship and offer opportunities for newer types of analysis, argument and presentation of history, when applied critically.
