Great Article. Now here's the thing. With the evolution of technology and the commoditization of Machine Learning and AI, you have this thing called - AutoML (where you can virtually drag and drop components in the entire Machine Learning food chain and then train and test it). Where do we go from here? If languages like Python, R or Julia can hide this Mathematical abstraction from developers by API's why even bother about Maths? I completed a year-long course recently on A.I. and M.L. and have observed that a lot of my peers simply gloss over topics like Eigen Vectors, stating that we do not need to go in so deep. So I ask you the same question (although I love indulging these topics myself) :- it becomes tough to build a case for getting deeper in these Mathematical topics when the very people who learn them are trying to simplify them for mass consumption.