Google DeepMind has unveiled an artificial intelligence tool that could revolutionize the study of ancient Roman texts. Named Aeneas, after the legendary Trojan hero, the AI program helps historians fill in missing words, date inscriptions, and identify their origin, tackling one of the biggest challenges in classical studies.
Developed in collaboration with historians like Dr. Thea Sommerschield from the University of Nottingham, Aeneas is trained on a massive dataset of 200,000 Latin inscriptions — around 16 million characters — ranging from the 7th century BC to the 8th century AD.
Unlike simple keyword search tools, Aeneas detects deep contextual similarities between texts, providing nuanced suggestions and helping place fragmented inscriptions in their correct historical and geographical settings. It can estimate the province and date of origin of a text (accurate to within 13 years) and suggest likely missing words in damaged or incomplete inscriptions.
In trials, historians found Aeneas helpful in 90% of cases, including identifying relationships between inscriptions in Germany that had never been formally linked. Notably, the tool gave historically plausible date estimates for the Res Gestae Divi Augusti, matching scholarly debates.
The breakthrough has been hailed as “transformative” by top scholars like Professor Mary Beard of Cambridge and Professor Jonathan Prag of Oxford, who emphasized its potential to democratize access to epigraphic research.
Aeneas is now publicly available for academic use, marking a major leap forward in how we decode and preserve the voices of the ancient world.
