A 370-Year-Old Cipher Just Got Cracked by AI in One Sitting

There's a puzzle over three hundred years old that stumped every codebreaker who ever attempted it. Nobody had solved it. Not once.

It's called the Cyphral Distich, just two lines of numbers, thirty-two per line, tucked at the end of Sir Thomas Urquhart's 17th-century book.

Think of it like an intricately designed lock, and all we have are these numbers, with no obvious hint about what key fits.

A research team decided to hand the puzzle to an AI model called Fable 5.1, with almost no extra guidance: 'here, solve this.'

The result was startling. The model actually arrived at a coherent, convincing solution to a cipher that had resisted every human attempt for centuries.

The genuinely embarrassing part, once the solution was revealed, is that it turned out to rely on relatively simple logic, logic that no human ever managed to spot across all those years.

The puzzle was never truly impossible. Human minds tend to lock onto certain patterns of thinking and dismiss other angles entirely, and that's exactly where AI has an edge, not because it's 'smarter,' but because it doesn't carry the same mental habits and blind spots we do.

So maybe the real question isn't 'can AI think like us?' It's 'how many other puzzles are we calling impossible, when really, we've just been looking in the wrong direction?'

---

🔗 Original Source & Reference: https://www.vals.ai/blogs/fable-solves-cyphral-distich

Published automatically via FeedMind AI Content Pipeline.