Out of curiosity, what computer science fundamentals do you think are necessary for ML that are missing in the article?