Definition
Plain language
A way of scrambling a message so only someone with the key can read it.
As stated in the literature
A method for transforming plaintext into disguised text; a substitution cipher swaps each symbol for another, and unlike steganography it makes the existence of a hidden message obvious.
Also called: substitution cipher, ciphers
Why it matters: It keeps a message readable only to someone with the key, though anyone watching can plainly see that a hidden message exists.
For example, a simple cipher might replace every 'A' with 'D' and every 'B' with 'E', so 'CAB' becomes 'FDE'.
Heard on the show
“You wrap "how do I build a bomb" in a fictional screenplay, or you encode it in a cipher, or you flood the context with so much text the model loses the thread.”Episode 118 — Why the Best-Aligned AI Models Are the Easiest to Trick Into Producing Harm