Meet Aeneas, Google DeepMind's AI For Decoding Ancient Texts
Google DeepMind has developed an AI model that can predict where ancient Latin texts originate, estimate their age and restore missing sections.
The model is called Aeneas and it represents a significant advancement in the field of epigraphy — the study of ancient inscriptions.
Developed by a team of researchers from universities in the UK and Greece, as well as a team from DeepMind itself, Aeneas addresses one of the most longstanding challenges in the fields of history, archaeology and anthropology.
"Aeneas can retrieve relevant parallels from across our entire data set instantly" because each text has a unique identifier in the database, says Yannis Assael, a research scientist at Google DeepMind.
Training on vast historical datasets
Aeneas was trained on inscriptions from three of the world's largest databases of Latin epigraphy.
The combined dataset contained text from 176,861 inscriptions — plus images of 5% of them — with dates ranging across a millennium from the 7th century BC to the 8th century AD.
The model comprises three neural networks, each designed for different tasks: restoring missing text, predicting geographical origin and estimating age.
Along with results, Aeneas provides a ranked list of similar inscriptions from the dataset to support its conclusions.
Outperforming human experts in accuracy tests
The team tested Aeneas by asking 23 epigraphers to restore text removed from inscriptions and to date and identify origins both independently and with the model's assistance.
Working alone, experts dated inscriptions to within around 31 years of the correct answer.
Dates predicted by Aeneas were accurate to within 13 years — significantly outperforming human specialists.
When specialists had access to the model's predictions and similar inscription lists, they achieved greater accuracy in geographical identification and text restoration than either working alone or relying solely on the model.
Real-world applications
Testing on the well-known Res gestae divi Augusti, which details Roman emperor Augustus's life, showed the model's predictions aligned with historians' assessments.
The tool correctly identified spelling variations and features historians use to determine age and origin, without being misled by dates mentioned within the text.
When examining a Latin altar inscription, Aeneas notably included another altar from the same region in its similar inscriptions list, despite not being programmed with geographical or temporal connections.
Addressing limitations of manual research
Anne Rogerson, who studies Latin texts at the University of Sydney, highlights the model's capacity to analyse vast amounts of data beyond any individual's capability.
"It's giving a hypothesis based on the evidence base that it's working from, so it's a rational guess rather than a wild stab in the dark," Anne says regarding the model's reliability when compared with more general AI tools.
The tool can help historians locate similar inscriptions — a process that typically requires weeks or months of manual research.
Thea Sommerschield, an epigrapher at the University of Nottingham, explains that finding comparable inscriptions is "incredibly time consuming" using traditional methods.
Current constraints and future development
The team acknowledges Aeneas faces limitations due to its smaller training database compared to models like ChatGPT and Microsoft's Copilot, potentially affecting performance on unusual inscriptions.
Anne suggests the model might prove less useful for unique inscriptions or those from periods with fewer available artefacts.
Despite these constraints, the research demonstrates how specialised AI applications can augment rather than replace human expertise in historical scholarship.
The development represents Google DeepMind's continued expansion into academic research applications, following previous work on ancient Greek text decipherment.
