All Articles

Decoding Ancient Scrolls With Machine Learning

October 28, 2023
x min read

Machine learning is all the rage in Silicon Valley.

But it also might have the potential to solve a surprising problem: decoding ancient scrolls.

What’s happening:

  • A SpaceX intern has won $40,000 from Scroll Prize for identifying the first ever words on an ancient scroll from over 2000 years ago
  • The breakthrough came from leveraging a machine learning model to identify ink on a scroll that was previously believed to be completely illegible due to being carbonized by heat

The big idea:

  • Nat Friedman, prolific angel investor and former leader of Github, launched Scroll Prize as a collaborative project offering $1M to decode ancient scrolls which were once considered to be illegible through machine learning and computer vision
  • Daniel Gross, angel investor and former Y Combinator partner, is another founding sponsor of Scroll Prize

Going deeper:

  • A vast collection of scrolls that were owned by Julius Caesar’s father in law were nearly thought to have been destroyed by hot mud and volcanic debris in 79 AD
  • In 1750 AD those scrolls were discovered when an Italian farmer was digging a well and came across more than 600 individual scrolls that had been carbonized by heat and were impossible to unravel
  • In 2015 there was a technical breakthrough by a team of researchers at the University of Kentucky which used computer vision to read a carbonized scroll entirely through technology and without unrolling it

Why it matters:

  • Uncovering what is inside ancient scrolls could unlock deep insights into Greek and Roman philosophy, science, politics, mathematics and much more

The intrigue:

  • Scroll Prize has a long list of notable supporters and sponsors including John and Patrick Collison, Matt Mullenweg, Tobi Lutke, Aaron Levie and many more

Discover what’s next. The world’s biggest ideas, disruptive trends, most exciting early stage companies and groundbreaking entrepreneurs.

By clicking Subscribe you're confirming that you agree with our Privacy Policy.
Thanks for subscribing!
Keep an eye out for a welcome email shortly.
Oops! Something went wrong while submitting the form.