AI Deciphers Ancient Scroll Buried by Vesuvius Eruption
Artificial intelligence has managed to decipher the words from an ancient scroll that had been tightly wrapped for nearly 2,000 years beneath the volcanic ash of Vesuvius. The densely rolled papyrus was unearthed in the ancient Roman city of Herculaneum, and opening it without causing damage, let alone reading it, would have been a challenging feat, according to Science Alert.
Buried after the eruption of Mount Vesuvius in 79 AD, these scrolls remained hidden until the 18th century when workers stumbled upon the remains of a luxurious villa, possibly belonging to Lucius Calpurnius Piso Caesoninus, the father-in-law of Julius Caesar. Unfortunately, very few such ancient texts have survived to the present day. If these scrolls can be read now without unwrapping them to avoid disintegration, scientists may gain invaluable insights into life and learning in the first century.
Therefore, in March, researchers at the University of Kentucky launched the Vesuvius Challenge, "Revive the Ancient Library from the Volcanic Ash. Win $1,000,000."
Thousands of X-ray images of the charred, carbonized scrolls from Herculaneum were made available with untrained artificial intelligence software that could be used to interpret the scans.
Currently, two students are vying for the top prize – Luke Farritor, a computer science student at the University of Nebraska-Lincoln, and Youssef Nader, a doctoral candidate in biorobotics at the Free University of Berlin in Germany.
Working independently, they both identified the word "πορϕυρας" (or "porphyras," using modern Greek characters), which means "purple." This marked the first complete word deciphered from the scripts with the help of artificial intelligence.
The top prize of the Vesuvius Challenge, $700,000, will be awarded to the first team that reads a scroll by December 31, 2023.
Recently, artificial intelligence autonomously detected a supernova, a bright and highly energetic celestial event, without human intervention, by analysing over 1.4 million astronomical images from nearly 16,000 sources.
Although supernovae are conspicuous and high-energy events, they are difficult to detect. Traditional search methods rely on astronomers visually sifting through vast amounts of data from robotic telescopes, constantly scanning the night sky for new sources of light. The use of AI is expected to significantly streamline this process.