Scientists have been able to interpret an almost 2,000-year-old Roman manuscript using artificial intelligence. AI can also provide Norwegian archaeologists with a number of new tools.
The short version
- Scientists have been able to interpret an almost 2,000-year-old Roman manuscript using artificial intelligence.
The summary is created with the help of artificial intelligence (AI) and quality guaranteed by Aftenposten journalists.
In 79, Mount Vesuvius erupted. The volcanic eruption is famous for burying the Roman city of Pompeii.
city HerculaneumHerculaneumHerculaneum was an ancient city in the Campania region of Italy, between Naples and Pompeii, at the foot of the western slope of Volcano Vesuvius. Herculaneum, like Pompeii, was completely destroyed by the eruption of Vesuvius on August 24, 79 AD. Source: SNL The eruption also buried it. There, the glowing ash swallowed up a collection of writings called the Herculaneum Papyri.
The collection of writings had been buried untouched in an ancient Roman villa for nearly 1,700 years, before it was found in 1750. Until now, its contents have remained a mystery.
Texts damaged by volcanic ash. They are in such a fragile state that it is impossible to open the manuscripts of the text without them disintegrating.
How then does one read the documents?
After 275 years, researchers found the solution.
The competition was launched with $1 million in the pot
The breakthrough came from the Vesuvius Challenge, a competition launched by Brent Sills, a computer science professor at the University of Kentucky, in 2023.
Wise minds from all over the world were invited to a race whose goal was to find a way to interpret historical documents.
Prizes totaling $1 million were promised to those who managed to crack the code.
On February 5, 2024, organizers reported a breakthrough.
A team of three computer-literate students Youssef Nader, Luc Varitor and Julian Schliger won the first prize of $700,000 for their contribution.
1 of 4Image: Vesuvius Challenge
Reconstruct text using artificial intelligence
Researchers have used a CT scanner for the first time to create a high-resolution 3D image of a paper roll. Use Computer vision Computer vision Computer vision tasks involve methods for acquiring, processing, analyzing, and understanding digital images. And Machine learningMachine learningMachine learning is a discipline within artificial intelligence where statistical methods are used to allow computers to find patterns in large amounts of data. Source: SNL The three students were able to crack the code.
They have developed algorithms for automatic character recognition. Artificial intelligence used traces of ink on documents to create a virtual copy of the roll of paper.
The text turns out to be a philosophical work about pleasure, food, and music.
But the researchers' work is not finished yet. So far they have only been able to recreate a small portion of a single scroll of text. The document is part of a collection of 800 works, which will now be interpreted, the magazine reported hour.
Artificial intelligence could become an important tool for Norwegian archaeologists
Maybe there are unsolved mysteries or runic staves from the Viking Age that we can explain with the help of artificial intelligence?
– First you have to find the runes. Once this is done, artificial intelligence could be a useful tool for further analyses, says archaeologist and researcher at NTNU, Arne Andersson-Stamnes.
According to Stamnes, only imagination sets limits on how AI can be used in archaeology.
In Sweden it was presented National Archives It is exactly an AI model that can interpret handwritten texts in Swedish from the 17th, 18th and 19th centuries with an accuracy rate of up to 95 percent.
Stamnes says AI can be used to identify inscriptions, patterns and symbols. One could potentially create a database and use artificial intelligence to compare this data.
It can provide information about where things are made.
-By comparing symbols, it is possible to prove that different objects were made in the same workshop or even by the same person.
This can provide insight into past trade, communication and contact networks. For example, you can find out if a Norwegian Viking hammer ended up elsewhere in Europe.
“Web specialist. Lifelong zombie maven. Coffee ninja. Hipster-friendly analyst.”