Interpretability and Unification

In a recent reply to our article, “What is Interpretability?,” Prasetya argues against position that artificial neural networks are explainable. It claimed indefeasibility thesis—that adding complexity an explanation of phenomenon does not make the any less explainable—is false. More precisely, unificationist explanations defeasible increasing complexity, and thus, we may be able provide such highly complex AI models. The highlights important lacuna in original paper, omission account explanation, affords us opportunity respond. Here, argue explainable way should satisfy unificationists interpretability methods present ways which ML theories can achieve unification.