Some light category theory is pretty useful (functors, monads, lenses, etc). However, I think it often goes way overboard into premature abstraction territory w

Some light category theory is pretty useful (functors, monads, lenses, etc). However, I think it often goes way overboard into premature abstraction territory with many functional programmers. The type tetris required to make some of this stuff compile, and the mental overhead, can easily exceed the abstraction gained. The category theory should also be viewed as more a close analogy to the programming problem, rather than an exact correspondence. These abstractions sometimes leak, especially in numerical code. edit: More complex and specialized abstractions are useful on fewer problems than simple, general abstractions. In addition, they have more mental overhead. So you work harder for a less useful abstraction, and you make it harder for others to use and understand your work.