We introduce a definitional shift for current AI reasoning as mere rambling, explaining recent studies on what models are more likely doing.
