Double-Loop Learning

Awesome!

Technically, we could say?

(1) Single-loop: fixes actions within fixed rules, like Reinforcement Learning.

(2) Double-loop: questions and adapts the rules, somewhat like Meta-Reinforcement Learning.

Five Whys is supposed to get you to this