Google DeepMind's Aeneas model can restore fragmented Latin text

Google DeepMind has created an AI model called Aeneas that can restore and contextualize fragmented ancient Latin inscriptions. The model is designed to be a collaborative tool for historians, offering "interpretable suggestions" that can help them situate inscriptions within their broader historical context. Aeneas can quickly identify parallels between similar texts based on factors like wording, syntax, or geographic region, a task that would otherwise be a time-consuming process for researchers. It can also restore textual gaps of unknown length, similar to solving a crossword puzzle. The model is multimodal, meaning it can analyze both textual and visual input, making it the first of its kind to use this multi-pronged approach. DeepMind has made Aeneas open-source and free to use, and has also upgraded its Ancient Greek text model, Ithaca, to be powered by Aeneas' contextual and restorative capabilities. Overall, Aeneas represents a valuable tool for historians, helping to streamline and enhance their work in deciphering and understanding ancient inscriptions.