The world of historical cryptography has just received a serious jolt. Anthropic's latest language model, Claude Fable 5.1, has managed to decipher the famous Cyphral Distich cryptogram, created by Scottish writer Thomas Urquhart over 370 years ago. This achievement, which I will detail further, could change our understanding of artificial intelligence's capabilities in the field of decryption.
The Cyphral Distich consists of two lines, each containing 32 numbers. This cryptogram was discovered at the end of Urquhart's treatise Logopandecteision, dated 1653. For centuries, the cipher remained impenetrable: in 1899, it was brought up for public discussion in Notes and Queries, and later it made its way into authoritative cryptographic literature of the 20th century. Researcher Klaus Schmeh even placed it at 28th in a ranking of the 50 most challenging undeciphered messages in history.
The Key to the Puzzle: The Book Itself
Previous decryption attempts failed because they relied on frequency analysis and substitution ciphers, assuming the key had to be sought outside the text. However, Fable 5.1 paid attention to critical details. The cipher was printed immediately after Urquhart's 32 "procuritations"—numbered wishes ending with phrases like "this is the wish" or "hope." The author specifically emphasized the number 32 in the text.
The model identified a pattern: the i-th number in a line refers to the i-th procuritation, the value indicates the ordinal position of a word within it, and the first letters of those words form the message. The decrypted text turned out to be a couplet in support of Charles II: "God support King Charles the Second and make him the supreme ruler of this land."
The result is self-verifying: each line contains exactly 32 letters, the lines rhyme, and Urquhart's political stance as a Stuart supporter fully aligns with the content.
A Second Victory and Disagreements in the Community
Using the same method, the model decrypted the more complex Cyphral Octastich cryptogram of 285 numbers, where the key was the pages of the book The Jewel (1652). However, nine letters in the fifth line remained undeciphered due to a missing leaf in the digitized copies.
Yet, not everything is so clear-cut. Specialists in historical ciphers have already expressed doubts. The Reticuli Labs report points to ten discrepancies in the algorithm: to obtain the word KING, the letter "K" is needed at position 11, but the corresponding section contains no words starting with that letter. Testing 65 counting variants yielded only eight matches out of 64—a level of pure chance.
My analysis: although Fable 5.1's achievement is impressive, the disputes surrounding the methodology show that we are only beginning to understand how AI interacts with historical mysteries. Perhaps the truth lies somewhere in between, and we will need a physical copy of the 1653 edition for a final verdict. In any case, this is a demonstration that modern LLMs are capable of unconventional thinking that goes beyond simple brute-force searching.