Comment by kriops

Define encryption.

To help you along, you basically have two alternatives:

1. Be intentionally vague, so the definition encapsulates just about anything and can then be applied and enforced at will. This is obviously what they are going for, by the way, and *that* is the word game being played here.

2. Some set of sets of mathematical functions, contingent on some properties pertaining to computational complexity. This is what cryptography is and, as such, is the correct way to go about it. Non-exhaustively, one property we are looking for is that some data can only be considered 'encrypted' if the computational complexity of decoding it without a secret/key is strictly higher than with said secret/key.

As I also said in another comment, I can derive an encrypted message a priori. That makes it fundamentally different from any analogies tied to physics or chemistry.