AI helps find the roads to Rome: puzzling out inscriptions with AI. From the article:

Artificial intelligence has already been used to fill in gaps in ancient Roman scrolls, but a new system goes much further. It can fill in missing words from ancient Roman inscriptions carved on monuments and everyday objects, as well as dating and placing them geographically.

AI often introduces errors in its analysis of even simple modern texts, so there are concerns that relying too much on this technology might distort rather than enhance our understanding of history. But historian Prof Dame Mary Beard of Cambridge University has described the technology as potentially “transformative” to our study of past events.

She said that the system, called Aeneas, after a Greek and Roman mythological figure, could accelerate the rate at which historians piece together the past from ancient texts. “Breakthroughs in this very difficult field have tended to rely on the memory, the subjective judgement and the hunch/guesswork of individual scholars, supported by traditional, encyclopaedic databases. Aeneas opens up entirely new horizons.”…

Dr [Thea] Sommerschield developed Aeneas along with her co-research leader Dr Yannis Assael, an AI specialist at Google DeepMind. It automates the process of contextualising based on parallels, in the blink of an eye. Aeneas draws on a vast database of of 176,000 Roman inscriptions including images and uses a carefully designed AI system to pull up a range of relevant historical parallels, to support the work of historians, according to Dr Assael….

In tests of the system with 23 historians the team found that an historian working with Aeneas came up with more accurate results than either Aeneas on its own or an historian on their own. “The feedback was that Aeneas was not only allowing the historians to accelerate their work but it also revealed parallels that they had previously not identified,” according to Dr Sommerschield.

For other posts on AI and history, see here.

For other posts by Pallab Ghosh, see here.