Artificial intelligence decodes writings dating back more than 5,000 years

Artificial intelligence decodes writings dating back more than 5,000 years

Image: Reproduction/Daily Mail.

A group of scientists from Germany developed an artificial intelligence program capable of decoding cuneiform texts, which are considered the oldest ever known to humanity.

The writings were made using wedge-shaped objects, most of which date back more than 5,000 years.

The new system uses 3D models of Sumerian tablets. It works in the same way as optical character recognition (OCR) software, which converts images of writing into machine-readable text.

The most trusted AI software

AI provides much more reliable results than previous methods, such as images.

The program allows tables to be compared with each other, which opens the way to new research directions. The researchers used 3D models of nearly 2,000 cuneiform tablets.

“Everything can be found in them: from shopping lists to legal decisions. The tablets provide a glimpse into humanity’s past,” explained Hubert Marra, assistant professor at the University of Halle-Wittenberg.

Cuneiform tablets are pieces of unfired clay onto which writing is pressed. The writing system at that time covered several languages.

Therefore, in addition to ideal lighting conditions to correctly recognize the symbols, a lot of prior knowledge of the languages ​​in use at the time is necessary.

The researchers’ work provides new access to exclusive materials and enables new lines of investigation.

Currently, there is only one prototype capable of reliably distinguishing between symbols from two languages. However, it is known that there are 12 cuneiform languages.

The Germans are using artificial intelligence to decipher ancient writings

Credits: Oeste Magazine.

You May Also Like

About the Author: Osmond Blake

"Web geek. Wannabe thinker. Reader. Freelance travel evangelist. Pop culture aficionado. Certified music scholar."

Leave a Reply

Your email address will not be published. Required fields are marked *