Image:  St. Gallen, Stiftsbibliothek, Cod. Sang. 904 https://www.e-codices.unifr.ch/en/csg/0904/189. The word ‘lathaeirt’ in centre of the top margin probably means ‘hungover’

Irish people have a great appetite for history, according to Insight funded investigator Pádraic Moran. A Celtic Studies specialist at the University of Galway, he now working with AI developers in what he describes as a research ‘pivot’.

‘I have always created resources for people interested in the Middle Ages; web tools for example. I work with manuscripts – handwritten books from the preprinting era – and especially on notes written in margins.’

Notes written on old scholarly texts can tell us a lot about the people and the period. Moran describes a Latin text copied by an Irish scholar who complains in the margin about the cold, the quality of the ink and his hangover.

‘I want to drive more public engagement with this period because I know that people are very interested in Ireland’s mediaeval history. However, much of the material is only accessible to specialists.’

Two years ago, Pádraic started working with the Insight Centre to see if, by combining his work with AI, he could ‘squeeze more value out of my research.’

This cross-disciplinary collaboration has led to a pilot project with Insight, starting in 2026, to create a chatbot interface for early Irish history resources.

‘Imagine a classroom in County Roscommon where the teacher is exploring local history with pupils. Right now, it’s quite hard to find out what happened in a local area in the Middle Ages. The information is there but hard to access. Our data source for the project is a set of texts called the Mediaeval Annals containing 10 to 20,000 recorded events. They are hard to search and make sense of. Our chatbot would use AI to structure the data.’

The end result, Moran hopes, would be a resource for non-specialists looking to find out what was going on at a particular time and place in Ireland.

‘That Roscommon teacher could focus on the local monastery in the seventh century, for example, and find results for classroom use.’

AI is opening up promising new areas of research in the emerging field of digital humanities. However, the field relies heavily on the human expertise of scholars like Moran, who, in the end, must sit down and read those scribbled notes, make sense of them, and know what a hangover is.