Researchers from LMU Munich and the University of Baghdad used AI to reconstruct a Babylonian poem that had been lost for more than 2000 years.
The text of Babylon, who praises Babylon and the God Marduk – was written 3000 years ago and last studied in 100 BC.
According to the team behind his rediscovery, it is merged from 30 clay rags that have been excavated over the years, using artificial intelligence to join the dots.
“We have used a specialized AI program to analyze and match text fragments based on combinations of cuneiform signs,” told Professor Enrique Jiménez, professor of old Eastern languages to LMU, told Decrypt.
Jiménez and his colleagues use approaches based on natural language processing to indicate that fragments belong to a single text, as detailed in a methodological article last year.
Working from the electronic Babylonian library platform, which contains 1,402 manuscripts, the researchers use N-Gram matching as their primary reconstruction method, although other methods include vocabulary overlaps and searching for the longest common strings (of text).
According to Jiménez, the rediscovered poem was important enough to be taught as part of Babylon’s curriculum.
Writing in the diary IraqHe and co-researcher Anmar A. Fadhil also suggest that the author was probably a member of the priest class of Babylon, given that the poem includes a section that describes priests as the ‘free citizens’ of Babylon.
In addition to celebrating the natural resources and beauty of Babylon, the hymn also contains passages that praise the acceptance by the city of foreigners and support for the poor.
It reads: “The foreigners under which they do not humiliate. The modests they protect, the weak ones they support. Under their care, the poor and needy can thrive. For the orphans they offer help and favor.”
The reconstruction of old texts with the help of AI is increasingly occurring with scholars; In 2023, a 21-year-old student won the headlines for the development of a machine learning algorithm to decipher old Greek letters in a sealed scroll from Herculaneum.
Jiménez told Decrypt That AI becomes “indispensable” for researchers, “in particular for reconstructing damaged or fragmented texts.” He added that: “Although languages such as Akkadian and Sumeric are still under -represented in large language models, we actively work on improving computational tools for old nearby studies.”