AI Cracks Napoleonic Code
An unsolved Napoleonic military mystery from 1809 was just cracked in six hours. Here is how a cybersecurity engineer used OpenAI's GPT-6 Astra to solve it.
What the video says
A new AI cracked a 217-year-old military mystery. The breakthrough was achieved by OpenAI's GPT-6 Astra.
Sentinel-1 engineer Carter Church used the model to decipher an encrypted message sent to a French general in 1809. The document, containing 24 rows of symbols, had sat unsolved for decades on Cryptiana's historical cipher list.
The newly revealed text completed a sentence about a gathering of rabble that broke off in Napoleon's memoirs. Satoshi Tomokio officially marked it solved after the code proved mechanically testable against the original ciphertext.
Astonishingly, Astra completed the entire multimodal workflow in just 6 hours from a single image and prompt.