AI revives ancient texts and unlocks lost knowledge
Summary and headline written by AI from the source article. How we work
Researchers in several countries are using artificial intelligence to read and preserve ancient manuscripts, revealing previously unknown details about history and culture.
A team led by Youssef Nader and Julian Schilliger employed AI to decipher charred manuscripts from Herculaneum, a city destroyed by Mount Vesuvius nearly two thousand years ago. A student at the University of Nebraska, Luke Farritor, used AI techniques to read messages from the fragile scrolls, identifying over ten letters and winning a $40,000 prize.
This work offers new insights into ancient Greek philosophy and Roman life. Scientists are also applying AI to the Dead Sea Scrolls, discovered in caves near Qumran, to refine their dating. Analysis combining radiocarbon dating and AI revealed some scrolls may be a century older than previously thought, dating back around 2300 years, the time when the authors likely lived. The AI corrected for contamination from modern materials, specifically linseed oil used in earlier preservation efforts.
Engineers in Tunisia developed “Zanki,” an application that uses AI to convert old Arabic manuscripts into searchable digital text. The application allows access to millions of Arabic manuscripts previously limited to specialists, and a digital project at the Baronia Library has used Zanki to access over 1200 historical manuscripts containing 150,000 pages. In Iraq, the House of Manuscripts in Baghdad is digitizing its collection of 47,000 ancient texts, including unique copies of works by Ibn Sina and early Qurans.
Scholars are also using AI to preserve Syriac manuscripts held by the Hermezdian Antionite Chaldean Monastery, a collection of over 1100 manuscripts.
