Scientists Built An AI That Reads 3,500-Year-Old Clay Tablets: Here Is What It Found

A new AI system is helping researchers piece together broken clay tablets from ancient civilisations, and it works faster than any human expert ever could.

Ancient scribes pressed symbols into wet clay 3,500 years ago.

Source : ABPLIVE AI

Show Quick Read

Key points generated by AI, verified by newsroom

• AI tool Palaeographicum reads and compares ancient cuneiform handwriting.
• System analyzes millions of characters, identifying scribal writing style nuances.
• AI aids tablet reconstruction and potentially dates undated fragments.

Researchers have developed an AI-powered tool that can read and compare handwriting on cuneiform clay tablets dating back over 3,500 years. The system, called Palaeographicum, was built by researchers from the University of Würzburg and the Academy of Sciences and Literature Mainz. It is being seen as a significant step forward in the study of Ancient Near Eastern civilisations, as it takes on one of archaeology's most time-consuming challenges: piecing together broken clay tablets and matching them to their original scribes. 

The tool processes millions of cuneiform characters in minutes, work that previously took trained experts several days to complete.

How Does the AI Tool Read 3,500-Year-Old Cuneiform Tablets?

Cuneiform is among the oldest known writing systems in the world. Ancient scribes used a stylus to press wedge-shaped symbols into wet clay, creating records that have survived thousands of years, often in fragments. 

Palaeographicum works by scanning and comparing handwriting patterns across a database that currently holds more than 70,000 photographs containing over five million cuneiform characters.

What sets the tool apart is its ability to detect subtle differences in individual writing styles. Even in an ancient script pressed into clay, scribes had personal habits; some made deeper marks, others spaced symbols differently. The AI picks up on these patterns much like a forensic handwriting analyst would.

According to Professor Daniel Schwemer, head of Ancient Near Eastern Studies at the University of Würzburg, the AI is "radically changing" research work and saving thousands of hours for scholars.

What Could This Mean For The Study Of Ancient Civilisations?

Beyond tablet reconstruction, the tool may also help date undated clay fragments by tracking how handwriting styles changed over centuries. 

Researchers are working toward future versions that could automatically identify individual scribes and follow their work across multiple documents.

The long-term goal is to build a detailed picture of writing culture within the ancient Hittite civilisation, which thrived in present-day Turkey roughly 3,500 years ago.