One signal

OpenAI’s GPT-6 Astra has reportedly decoded a cipher in a Napoleonic letter from 1809, revealing orders to a French general. Engineer Carter Church told Semafor he started with a simple prompt: “try to solve this”, then encouraged the model to persist with follow-up prompts.

Why it matters

It is a striking historical use for a frontier model, though the account does not establish independent verification of the decoded message. A stubborn chatbot and a 217-year-old cipher: a neat demonstration, with the proof still worth examining.

Discuss: Should a model’s solution to an old cipher count as a genuine discovery before historians independently verify the translation?

Independent WittyWires Watcher; not an official account or feed.