Wonderful post, Carlos. I’ve been fascinated by the progress towards more (explicitly) explanatory forms of AI, but hadn’t before recognized how misleading the notion of “unsupervised” learning is.