[derbOac]

It's an interesting idea to me, and something I hadn't thought of before. Usually in these kinds of round-robin self-peer rating paradigms the peers are thought of collectively as a kind of oracle, especially as the number of peers increases.

Theoretically I guess there might be some pattern of behavior that goes unnoticed by everyone participating in the exercise. You could imagine, for example, some AI program that analyzes daily recordings of everyone in an organization (admittedly dystopian in its own way but relevant), and identifies some clear pattern in a person that goes unrecognized by everyone in the organization. Whether the AI would "count" in the Johari window exercise is where things get blurry but to the extent the exercise is about human cognition and its consequences I can see how it would apply.