Researchers have successfully utilized artificial intelligence to decode a carbonized scroll from the ancient Roman city of Herculaneum, revealing previously invisible philosophical text. The breakthrough allows scholars to read papyri that were buried and scorched by the eruption of Mount Vesuvius.
The project provided the framework for this technological advancement. By applying machine learning models to high-resolution CT scans of the papyri, researchers identified ink traces within the layers of the scrolls. The decoded text offers new insights into ancient history.
How Artificial Intelligence Unlocks Carbonized Papyri
The primary hurdle in studying the Herculaneum scrolls has been their physical state. The intense heat of the Vesuvius eruption carbonized the scrolls, making them resemble dense, brittle lumps of charcoal. Any attempt to open them manually typically resulted in the destruction of the parchment.
To overcome this, researchers utilized volumetric CT scanning. This process creates a 3D internal map of the scroll. The AI, specifically trained on these scans, recognizes the subtle difference in density between the carbon-based ink and the carbonized papyrus. The machine learning algorithms segment the layers of the scroll virtually, allowing researchers to “unroll” and read the contents digitally without ever touching the original artifact.
The Content of the Deciphered Text
The scroll currently being analyzed contains text. The passages provide a rare glimpse into the academic life of the Roman elite. Prior to this AI-driven effort, these specific sections were entirely inaccessible to historians.
This discovery represents a monumental shift in classical philology. By automating the detection of ink, the process significantly reduces the time required to transcribe these ancient texts compared to manual restoration methods.
Technological Impact on Archaeological Research
The success of this project demonstrates the increasing utility of machine learning in the humanities. Beyond Herculaneum, the techniques developed are expected to be applied to other damaged historical documents worldwide. The project has served as a proof of concept, showing that interdisciplinary collaboration between computer scientists and classicists can resolve long-standing archaeological barriers.
The team continues to release updated datasets and software tools to the public, encouraging global participation in the ongoing effort to read the remaining hundreds of scrolls from Herculaneum. Future phases of the project aim to decode larger portions of the collection, with the ultimate goal of reconstructing entire works that were previously thought lost to history.
As researchers move forward, the next checkpoint involves the refinement of the AI models to improve the clarity of the text recognition in more severely damaged sections of the scrolls. Scholars expect to publish further findings as more segments are processed. We welcome your thoughts on how AI might continue to reshape our understanding of the ancient world in the comments section below.
Related reading