AI Solves 370 Year Old Cipher That Stumped Humans For Centuries

# Claude Fable 5.1 Solves a 370-Year-Old Unsolved Cipher: How AI Cracked Sir Thomas Urquhart’s Cyphral Distich in Under an Hour

## Introduction: A Centuries-Old Cryptographic Puzzle Finally Yields

In a striking demonstration of artificial intelligence’s capacity for historical problem-solving, Claude Fable 5.1 has reportedly solved the Cyphral Distich—a 370-year-old cipher created by Scottish polymath Sir Thomas Urquhart that had baffled cryptographers since its publication in 1653. The cipher, consisting of just 64 numbers arranged in two lines of 32 digits each, appeared at the end of Urquhart’s *Logopandecteision* and was listed among historical-cipher researcher Klaus Schmeh’s Top 50 unsolved encrypted messages. Despite numerous attempts over the centuries using frequency analysis, substitution methods, and homophonic substitution—all of which failed—the solution, once discovered, proved embarrassingly simple. Claude Fable 5.1 cracked the cipher in just 44 minutes using 176,000 tokens with zero human interjections, revealing a plaintext that reads: „O GOD UPHOLD KING CHARLS THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND.”

## The Solution’s Elegant Simplicity: Internal Keys Rather Than External Alphabets

The breakthrough came not from extraordinary cryptanalytic prowess, but from recognizing clues that had apparently been overlooked for centuries. The cipher is printed immediately after Urquhart’s 32 „Proquiritations”—a number he explicitly emphasized, writing that „there can no number like that of two and thirty… be pitched upon.” Additionally, the poem accompanying the cipher promised that honest readers would find in it „his own heart’s wishes, and the Author’s minde,” while the Proquiritations themselves repeatedly concluded with formulations like „is the desire,” „wish,” or „hope of.” Claude Fable 5.1 connected these dots: 32 Proquiritations, 32 numbers in each cipher line, and language about wishes. The key was not an external cipher alphabet as historical attempts had assumed, but rather the book itself. The rule was elegantly straightforward—for the i-th number in a cipher line, go to the i-th Proquiritations, use that number as a word index, and take the first letter of that word. This self-verifying solution produces exactly 32 letters per line ending in „and” and „land,” forming a rhyming distich consistent with Urquhart’s Royalist politics.

## Extending the Breakthrough: The Cyphral Octastich and Its Royalist Prayer

Beyond the Distich, Claude Fable 5.1 applied the same methodology to decipher Urquhart’s larger 1652 work, the Cyphral Octastich from *The Jewel*—a 285-number cryptogram that had remained equally unsolved. Using an analogous principle where the k-th number served as a word index into page k of the 284-page book, the AI recovered all but nine letters of an ottava rima royalist prayer beginning „GREAT LORD, MANTAINE THAT REGAL FAMILIE WHEREOF KING CHARLS THE SECOND IS THE HEAD.” The deciphered text includes archaic Scots vocabulary like „sweigh” (meaning sway or controlling power), which rhymes with „familie” and „authoritie,” and concludes with „AMEN, SO BE IT.” Caveats remain regarding nine unclear letters and some page-numbering anomalies requiring verification against a physical copy of the 1652 edition, but the bulk of the solution appears solid.

## Implications: AI as a Tool for Historical Mysteries and Archival Puzzles

This achievement demonstrates that AI models can contribute meaningfully not just to mathematics but to historical mysteries, forgotten conjectures, and archival puzzles that have traditionally been bottlenecked by human attention. As the author notes, someone historically had to care enough to spend hours or days reading obscure material, testing unpromising ideas, and tracing references. That bottleneck is now seemingly disappearing. Crucially, Claude Fable 5.1 did not perform extraordinary cryptanalysis—quite the opposite. The answer was simple in hindsight; the AI simply kept looking until it found it. The elicitation approach was equally straightforward: the author gave the model a goal, encouragement, examples of other strong feats, and constraints to avoid unsolvable or ambiguous problems. This persistence and willingness to explore seemingly tractable problems may prove transferable to many other domains where human attention has been the limiting factor.


Ez a cikk a Neural News AI (V1) verziójával készült.

Forrás: https://www.vals.ai/blogs/fable-solves-cyphral-distich.