Computer Scientist Explains One Concept in 5 Levels of Difficulty | WIRED

Sorry, the first three explanations are individually great, but each explanation includes a step that requires a leap of faith.

In the child explanation he "proves" that he knows where the puffin is in the picture of penguins by hiding A photo behind a screen with a hole in it. BUT we have to trust that the photo behind the screen was the same photo we originally saw.

In the teenager explanation he says that cracking an encrypted message would prove he could both crack all messages using that same encryption and could encrypt messages using that same encryption. With a very low or very high level of encryption this would tend to be believable but his example doesn't hold up for a very high level of encryption unless it is a public key encryption system (which isn't really a zero knowledge proof situation).

In the college student explanation he uses a 3-color map as an NP-complete proof. That's fine, but it relies on his statement "we can take that statement and convert it into a map that is only a valid 3-color map is the statement is true" THAT is a giant leap of faith.

In the grad student discussion, both already understand and accept "zero knowledge proof" based on NP-completeness and there is no explanation. The grad student offers one theoretical example, but that example is really just "guess what number I am thinking of"

In the expert discussion, again there is no explanation just a discussion of why zero-knowledge proofs would be useful and some challenges faced in making zero-knowledge proofs workable.

Bottom line: His "explanation" of a "zero-knowledge proof" is just the definition combined with "I know one when I see it"

he has failed to explain how the process works making it reliable. A programmer could write anyting to make it look like its real.