The Ercolano papyrus burned 2000 years ago is deciphered

2024-02-15 17:52:28Lifestyle SHKRUAR NGA ZULFIKAR ABBANY
The Ercolano papyrus burned 2000 years ago is deciphered
The Ercolano papyrus considered a challenge to Vesuvius

Scientists have used artificial intelligence to decipher the text on a papyrus scroll burned 2,000 years ago. Now artificial intelligence (AI) goes back in time looking to the future.

More than a thousand scrolls were buried and covered by volcanic debris when Mount Vesuvius erupted about 2,000 years ago. They were in a library in a Roman villa in the ancient city of Ercolano and were discovered in the 1800s by a local farmer.

Many people have since tried to read the texts on the ancient papyri, but most attempts have failed, as the documents were destroyed. For centuries they were left coiled underground, charred and brittle.

The winners - Youssef Nader, Luke Farritor and Julian Schilliger - beat the challenge by managing to read four paragafas without ever opening the papyrus scrolls.

Winning the Vesuvius challenge

They used machine learning, a form of artificial intelligence, to read the ancient Greek text. Nadery, Farritor and Schillinger contributed independently to the Vesuvius scroll community and now share the €650,000 grand prize.

The goal was to decipher four paragraphs of text, each of at least 140 characters, with at least 85% of the characters "recoverable," or readable.

View of papyrus damaged by the volcano on Vesuvius 2000 years ago stored in the National Library of NaplesView of papyrus damaged by the volcano on Vesuvius 2000 years ago stored in the National Library of Naples

The Ercolano papyrus burned 2000 years ago is deciphered
The damaged Papyrus of Ercolano housed in the National Library of Naples

Their work has uncovered what are believed to be unknown texts by Philodemus, the villa's so-called philosopher-in-residence.

In the text, Philodemus writes about living a good life through the pleasures of beauty, music and food. The researchers say this and future discoveries in the texts will give them a "unique window into the classical world".

How scientists used AI to read a Herculaneum scroll

The blinds are digitally "unwrapped" using computed tomography (CT), or X-ray pictures, and machine learning technology.

First, in late 2023, prize organizers photographed the scrolls at the Diamond Light Source particle accelerator near Oxford in Great Britain. This produced high-resolution CT scans of the parchments.

The scans were then converted into a 3D voxel volume. Voxels are 3D pixels, similar to the building blocks used in the video game Minecraft.

The second step is known as segmentation. They traced the crumpled layers of the rolled papyrus in the 3D scan. This allowed them to decompress, or flatten, the images.

And the third step was the discovery of ink on papyrus. They used machine learning to identify areas of ink on flattened papyrus segments.

The Ercolano papyrus burned 2000 years ago is deciphered
Roll of carbonized papyrus

But here comes the clever part. The machine learning model was not trained to recognize ancient Greek letters, optical character recognition (OCR), or any other language patterns. Instead, he simply detected ink blots on the CT scan and then combined and reconstructed each individual position - and where the letters appeared, there was writing.

Decades of work bear fruit

One of the organizers of the award, Brent Seales of the University of Kentucky, had been working on deciphering the Ercolano scrolls for decades. Seales was the first to use CT scans, but found that the ink was difficult to detect because it had a similar density to the papyrus it was written on.

But developments gained speed when Seales, Silicon Valley entrepreneur Nat Friedman and engineer Daniel Gross launched the competition in March 2023.

A breakthrough came within a few months. A former physicist, Casey Handmer, noticed a cracked texture in the text and called it cracking.

Co-winner Farritor, a university student and part of a SpaceX intership, used Handmer's observation to train a machine learning model and deciphered the first whole ancient Greek word: ΠΟΡΦΥΡΑϹ, meaning purple .

In October, Nader, an Egyptian doctoral student in Berlin, was able to read several columns of text.

Schillinger, a Swiss robotics student who previously won three prizes for segmentation, enabled the 3D mapping of the papyrus scrolls.

But there is still much to discover in 2024. Vesuvius' next challenge is to read an entire work or scroll by the end of the year. /DW



Video