AI helping to solve ancient secrets
When Mount Vesuvius erupted in 79 AD, it buried Pompeii, Herculaneum, and several other nearby villages under ash and rock. In the 1700s, archaeologists recovered nearly 2,000 papyrus scrolls from a Herculaneum villa that had been engulfed by the eruption, but many were so fragile they couldn’t be unrolled, let alone read, without crumbling. The European Research Council gave the GreekSchools Project €2.5 million (about $2.7 million) to develop tools that could be used to read what was written on the charred Herculaneum scrolls, without destroying them. This led to the creation of a “bionic eye” that uses a combination of AI and several advanced scanning techniques, including optical imaging, thermal imaging, and tomography (the technique used for CT scans), to capture differences between parts of the scrolls that were blank and those that contained ink — all without having to physically unroll them. On April 23, team leader Graziano Ranocchia announced that the group had managed to extract about 1,000 words from a scroll titled “The History of the Academy” and that the words revealed Plato’s burial place: a private part of the garden near a shrine to the Muses.
The unearthing of ancient scrolls, combined with the innovative application of artificial intelligence (AI) to decipher their contents, signifies a monumental leap forward in our understanding of bygone eras. The breakthrough came after a global competition was launched to accelerate the reading of the texts. The Vesuvius Challenge offered $1 million in prizes to anyone who could solve the problem and find a way to read the remaining 270 closed scrolls, most of which are preserved in a library in Naples, which is around 8 miles west of Herculaneum. Three students managed to read 15 columns in a scroll with the help of coding machines powered by AI. In the end, the judges, who included Janko, decided that a team of three students — Luke Farritor from the U.S., Youssef Nader from Egypt, and Julian Schilliger from Switzerland — should share the $700,000 grand prize. The trio were able to read 2,000 letters from the scroll after training machine-learning algorithms on the scans. After creating a 3D scan of the text using a CT scan, the scroll was then separated into segments. A machine learning model — an application of AI — then detected the inked regions, allowing them to decipher the text.
Already 2 years ago, researchers at the University of Notre Dame were developing an artificial neural network to read complex ancient handwriting based on human perception to improve the capabilities of deep learning transcription. In research published in the Institute of Electrical and Electronics Engineers journal - Transactions on Pattern Analysis and Machine Intelligence, Scheirer outlines how his team combined traditional methods of machine learning with visual psychophysics — a method of measuring the connections between physical stimuli and mental phenomena. For example, analysing the amount of time it takes for an expert reader to recognize a specific character, gauge the quality of the handwriting or identify the use of certain abbreviations. Using deep learning to transcribe ancient texts is something of great interest to scholars in the humanities. And leveraging AI capabilities and technologies to do so, is an unique opportunity.
The paradox of the ancient scrolls being decoded by AI highlights the transformative potential of artificial intelligence in shaping our world.
