That looks interesting, and I'm definitely not saying there's no intersection between CT and DS. There's some cool things I've seen with CT and probability theory recently. But to me it often seems like theory in search of a problem, and a far cry from functor becoming a common word among ML practitioners.