2 ms·
Some questions 1/how do we know we found the solution and not a decryption that gives a plausible message/solution. Aka how do we know this isn’t a false posit
by amznbyebyebye 6y ago
Some questions
1/how do we know we found the solution and not a decryption that gives a plausible message/solution. Aka how do we know this isn’t a false positive solution
2/have cryptologists studied possibility of “synthetic” cypher generation to bootstrap examples upon which an ML model could be trained to decrypt such message in the future ?
- kadoban 6y ago1/ I think the "key" doesn't have enough entropy to encode the text generated. This should be good evidence towards it not being a false solution. If your key can't realistically encode enough information to "pick" out an arbitrary false plaintext, and yet the plaintext you have seems genuine judging by content, then you probably have a real decryption. Also the way the cipher was done is plausible to the time period (that is, it's something you can do with pencil and paper), and the steps seem like something a human would do.