Piecing Together an Ancient Epic Was Slow Work. Until A.I. Got Involved.

In 1872, in a quiet second-floor room at the British Museum, George Smith, a museum employee, was studying a grime-encrusted clay tablet when he came across words that would change his life. In the ancient cuneiform script, he recognized references to a stranded ship and a bird sent in search of land. After he had the tablet cleaned, Smith was certain he’d found a prototype of the biblical flood story.

“I am the first man to read that after more than 2,000 years of oblivion,” Smith reportedly said in a frenzy of excitement.

Smith realized that the tablet, which had been excavated in what is modern-day Iraq, was a small part of a much longer work — one that some then thought could help shed light on the Book of Genesis. The discovery made Smith, who came from a working-class family and had largely taught himself cuneiform, famous. He dedicated the rest of his life to searching for missing pieces of the poem, making multiple trips to the Middle East before dying of an illness on his final trip in 1876, at age 36.

For 152 years since Smith’s discovery, successive generations of Assyriologists — experts in the study of cuneiform and the cultures that used it — have taken up his quest to piece together a complete version of the poem known now as the Epic of Gilgamesh. Fragments of the epic, which was written more than 3,000 years ago and was based upon still earlier works, have re-emerged as tablets have been unearthed in archaeological digs, found in museum store rooms or surfaced on the black market.

The researchers face a daunting task. There are as many as half a million clay tablets housed in the Mesopotamian collections of various world museums and universities, along with many more tablet fragments. But since there are so few experts in cuneiform, many of these writings are unread and many more are unpublished.

So despite a generation-spanning effort, about 30 percent of Gilgamesh remains missing and there are gaps in modern understanding both of the poem and Mesopotamian writing in general.