Pretty interesting! A big surprise for me was to find the explainability cluster quite far from the causality one. But I guess it stems from a cultural facet:

Pretty interesting! A big surprise for me was to find the explainability cluster quite far from the causality one. But I guess it stems from a cultural facet: causality is mainly the purview of statistics (with Pearl at the helm) with a strong medical sciences focus; when explainability is more of a reaction to algorithm used in the industry (trees, GLMs, neural networks, etc.); which you deploy for performance, and only then you care about knowing the why.