The Herculaneum papyri (Naples, Italy), buried by the eruption of Mount Vesuvius 2,000 years ago, have remained shrouded in mystery for over 250 years. However, an unprecedented breakthrough has allowed the deciphering of the first words of these ancient texts thanks to the combination of artificial intelligence and the expertise of researchers. Although there is still a long way to go, this discovery marks a milestone in the study of ancient literature.
The findings date back to the 1750s, when hundreds of papyrus scrolls were discovered in the remains of a luxurious Roman villa. The library of the Villa of the Papyri has the potential to offer invaluable insight into the thinking of ancient Rome, but until now, these documents were doomed to remain as fragile charred blocks.
The library in question, located at the National Library of Naples, has been the subject of study by scholars and scientists for centuries. But the challenge lies in the fact that the scrolls are rolled up and disintegrate when attempting to physically unroll them, making analysis difficult.
When AI Meets Ancient History
In 2023, the “Vesuvius Challenge” was launched by tech executives to encourage the use of machine learning and computer vision techniques to decipher these ancient texts. Recently, a significant breakthrough was announced: the generation of the first image of the interior of one of the three scrolls located at the Bodleian Library at the University of Oxford.
Brent Seales, a computer scientist at the University of Kentucky and co-founder of the challenge, expressed his excitement after the successful scan, stating that the scroll contains more recoverable text than any previously scanned from Herculaneum.
How Does This Technology Work?
The scanning was conducted by Diamond Light Source, a laboratory that uses a synchrotron to generate high-intensity X-rays. Scientists employed artificial intelligence to merge the images, locate ink, and enhance text readability, thus creating a 3D representation of the scroll and allowing for virtual unrolling.
Despite the advancements, AI has limitations. So far, only a few words have been distinguished, with “disgust” being one of the few. However, the academic community is invited to participate in this collaborative effort to complete the text.
Peter Toth, curator of the Bodleian, commented on the need to improve the images and expressed his optimism about the technology’s potential to facilitate the process. With thousands of scrolls still to be deciphered, the future of the study of ancient literature looks brighter than ever.
0 Comments