Israelis Develop Artificial Intelligence to Translate Ancient Texts
Researchers in Israel have developed an Artificial Intelligence program that can translate an ancient Semitic language—Akkadian—into English. Akkadian was the language of Mesopotamia in 3000 BC and eventually became extinct around the time of Jesus. Ancient and modern Hebrew are also in the Semitic language family, along with Arabic and a few other languages most of which are no longer spoken.
Archaeologists have found as many as 2 million cuneiform tablets in modern times, but only an estimated 30,000 to 100,000 have been read or published. Cuneiform is one of the oldest forms of written communication. Tel Aviv University and Ariel University researchers have created an Artificial Intelligence program that takes the markings and translates them into English.
Dr. Gai Gutherz, Dr. Jonathan Berant, Dr. Shai Gordin, and Dr. Omer Levy recently published their breakthrough findings on using a “neural machine” to translate the ancient texts. They hope this will open a new window into life in the ancient Middle East, although they caution that computers can only do so much when it comes to translating from an ancient language into a modern one.
“…Most of these documents remain untranslated and inaccessible due to their sheer number and the limited quantity of experts able to read them. It can be a complex process since it commonly requires not only expert knowledge of two different languages but also different cultural milieus,” they explained.
“Digital tools that can assist with translation are becoming more ubiquitous every year, tied to advances in fields like optical character recognition (OCR) and machine translation. Ancient languages, however, still pose a towering problem in this regard. Their reading and comprehension require knowledge of a long-dead linguistic community, and moreover, the texts themselves can also be very fragmentary.”
The use of artificial intelligence is a controversial subject these days. This may be one application where it can be put to good use, but it will still require a human touch to interpret the data because machines can only do so much—especially when it comes to translating across ancient cultures and extinct languages.
(Photo: Public Domain)