ChatGPT-6 Astra cracks 85-year-old 1941 Enigma-coded message in two days

ChatGPT-6 Astra cracks 85-year-old 1941 Enigma-coded message in two days

Note that there was a spelling mistake at the front of the original message, which wouldn’t help decoding efforts.

We’ve looked at Enigma machines before, as efforts to crack wartime ciphers from them, as well as from Lorenz machines, would lead to some of the first general-purpose digital computer designs.

The Enigma featured a 26-key keyboard, with 26 light-up letters on the ‘lampboard’ positioned above it. Behind the lampboard, three rotors were installed from a set of five. The signal from a keyboard press traveled through an input wheel, then the three manually turnable rotors with 26 pins each, then a reflector to return the signal, which lit up the lampboard. Making it even more tricky to crack, the rotors would advance every keypress. Thus, if you keep pressing the same letter, a different letter on the lampboard would be shown each time. Yet another layer of obfuscation was added via the plugboard at the front of the machine, where users swapped more letters around. Operators used up to 10 plugboard rerouting wires.

Before using the Enigma machine, four settings must be dialed in by the operator: rotor order, ring setting, rotor starting positions, and plugboard connection(s). These settings were distributed separately, and they changed following a calendar. Even if the enemy had an Enigma machine, they’d also need to know the settings on the day to decode the message(s).

We reported on GPT6-Astra's deciphering skills just last week. The LLM was behind the cracking of a 108-year-old unsolved WWI German code that had been encrypted using the ADFGVX method.

Follow Tom's Hardware on Google News , or add us as a preferred source , to get our latest news, analysis, & reviews in your feeds.

Mark Tyson Social Links Navigation News Editor Mark Tyson is a news editor at Tom's Hardware. He enjoys covering the full breadth of PC tech; from business and semiconductor design to products approaching the edge of reason.

evilpaul I'm not a cryptographer, but could they have tried asking the Germans? I bet somebody would know. Reply

jeremyj_83 I remember seeing a video maybe 10 years ago by a computer scientist and breaking Enigma. They coded a program (I think in C++ or Java) that did a brute force method to breaking the encryption. Using nothing more than a midrange laptop with an i5-4200u the program was able to break a 1940 era Enigma (3 roters out of 5 choices) in several minutes. I guess the amazing thing was ChatGPT wrote a program? Thing is a lot of people use AI to do the initial code creation now anyways. IMO this message has probably been cracked multiple times but people didn't care. Reply

derekullo jeremyj_83 said: I remember seeing a video maybe 10 years ago by a computer scientist and breaking Enigma. They coded a program (I think in C++ or Java) that did a brute force method to breaking the encryption. Using nothing more than a midrange laptop with an i5-4200u the program was able to break a 1940 era Enigma (3 roters out of 5 choices) in several minutes. I guess the amazing thing was ChatGPT wrote a program? Thing is a lot of people use AI to do the initial code creation now anyways. IMO this message has probably been cracked multiple times but people didn't care. You don't really need to crack it … I'm sure some one kept the code book logs of what the enigma settings were for each day. Here's April 1940's codes https://www.101computing.net/wp/wp-content/uploads/enigma-code-book.pngThe Enigma machine was used between 1939 – 1945 during WW2. As long as you know the day a message was sent/encrypted a 2190 line excel file could contain all of the settings for the rotor order, ring setting, rotor starting positions, and plug-board connections to decrypt any message you want. Even without knowing the day a message was sent modern computers could just try every day's settings in a few milliseconds and decrypt it. Reply

Air2004 Admin said: A German Army Enigma transmission from 85 years ago, known as the MVUEH message, has been cracked by GTP-Astra in just two days. ChatGPT-6 Astra cracks 1941 Enigma-coded message in two days : Read more Now, that there's funny 😂😂🤣🤣🤣 (Thank you for that) Reply

Joey2121 Now could we focus on cancer research using AI please? Reply

Key considerations

  • Investor positioning can change fast
  • Volatility remains possible near catalysts
  • Macro rates and liquidity can dominate flows

Reference reading

More on this site

Informational only. No financial advice. Do your own research.

Leave a Comment