Decoding the Cyphral Distich Record
Vals AI announced that its Claude Fable 5.1 model successfully decrypted a 373-year-old Cyphral Distich record created by Scottish poet Thomas Urquhart. The decryption process took 44 minutes and was conducted by Gebi Jaff from Vals AI. The model used 176,000 tokens without any prompts, employing an internal key that linked the numbers to 32 chapters of the book. The decrypted text includes a call to support King Charles II and reads: 'O God uphold King Charls the Second and make him the supreme ruler of this land.'
The Cyphral Distich record was published in the book 'Logopandecteision', released in London in 1653, where two rows of numbers were printed with 32 numbers per row, totaling 64 numbers. In 1899, the puzzle was published as an open problem in the journal 'Notes and Queries'. Klaus Schmeh listed Cyphral Distich as the 28th best unsolved encrypted message. According to Vals AI, the model was able to match the numbers with words, with the first letters of those words forming the decrypted message.
Impact on the Study of Historical Ciphers
Currently, there is no independent confirmation of Vals AI's success. Experts in historical ciphers have yet to verify the accuracy of the decryption. A report from Reticuli Labs has appeared on the GitHub platform, attempting to challenge Vals AI's results. It is known that earlier, Claude Fable 5.1 helped mathematician Levent Alpoge to disprove the Jacobi hypothesis, which remained unsolved for 87 years. Levent Alpoge, who works at Anthropic, unofficially reported his finding on the social network X.
'There is no number like the number thirty-two… on which one can stop.' Thomas Urquhart
Vals AI's success in decrypting the Cyphral Distich record could significantly impact the study of historical ciphers and encryption methods. However, the lack of independent confirmation of the results highlights the need for further verification and analysis by experts in the field. This may also mark a new stage in the development of artificial intelligence technologies capable of solving complex scientific and historical problems.