AI Technology Aids in Deciphering Ancient Scroll Damaged by Vesuvius
'Having certainly strained ourselves to the utmost through research and learning, we will no longer be inferior to them,' reads a scroll virtually unwrapped with the help of AI
By The Register
Researchers have used artificial intelligence and advanced imaging to read text from a fragile ancient scroll carbonised by the eruption of Mount Vesuvius nearly 2,000 years ago.
Researchers have used artificial intelligence to help read an ancient scroll that was carbonised by the eruption of Mount Vesuvius nearly 2,000 years ago.
The breakthrough is part of the Vesuvius Challenge, a global project using machine learning, computer vision and advanced scanning to read the Herculaneum scrolls without physically opening them.
The scrolls were buried and burned during the eruption in AD 79, the same disaster that destroyed Pompeii and Herculaneum.
For centuries, many of the manuscripts could not be safely unrolled because they were too fragile. Attempts to open them risked turning them to dust.
The latest work uses high-resolution scans to see inside the damaged scrolls, then applies AI techniques to detect layers of papyrus, identify traces of ink and help reconstruct readable text.
One translated passage reads: “Having certainly strained ourselves to the utmost through research and learning, we will no longer be inferior to them in any respect.”
The line is striking because it comes from a document that modern researchers themselves have only been able to recover through years of research, imaging and machine-learning work.
The Herculaneum scrolls are among the most important surviving texts from the ancient world. They were found in the Villa of the Papyri at Herculaneum, a Roman town buried by Vesuvius.
The villa is believed to have contained a major ancient library, with hundreds of carbonised papyrus scrolls discovered there.
The Vesuvius Challenge was launched in 2023 to accelerate efforts to read the texts by bringing together researchers, engineers, computer scientists and volunteers.
The project says it uses machine learning and computer vision to read carbonised scrolls that cannot be opened by hand.
Recent reporting says researchers have now recovered substantial sections of text, including work linked to ancient philosophy and ethics.
The technology is important because the ink and papyrus are both carbon-based, making them extremely difficult to distinguish with standard imaging alone.
AI can help by recognising subtle patterns in scan data that may indicate writing, even when those marks are invisible or almost invisible to the human eye.
This does not mean AI is replacing historians or classical scholars. Human experts are still needed to verify readings, translate the Greek, interpret damaged passages and place the text in historical context.
Instead, the technology is giving scholars access to material that was previously unreadable.
The breakthrough also shows one of the more positive uses of AI: not just generating new content, but helping recover lost human knowledge.
For archaeology and historical research, similar methods could eventually help with other damaged manuscripts, inscriptions and artefacts where text is hidden, faded or physically inaccessible.
The work is still developing, and not every scroll can yet be read fully. The process remains technically difficult, involving scanning, digital reconstruction, ink detection and scholarly interpretation.
But the progress made so far suggests that more lost texts from the ancient world may be recoverable.
For readers, the wider point is clear: AI is increasingly becoming a research tool, not just a consumer chatbot or business productivity system.
In this case, it is helping scholars look inside a document sealed by a volcanic eruption almost two millennia ago.