Now whenever I see mutated variables and class attributes, or random side effects besides reading/writing to a database, it kinda makes me cringe and think it's a "oh here we go" into a rabbit hole just to understand what the code is doing. 9/10 the code does what it's supposed to, but the mutations and side effects makes understanding and extending it so much harder.
A monoid is just a collection of things that can be associatively "added". Think addition with integers, or append with lists.
Members of the dreaded monad can be sequenced, or composed, while taking into account their context. For instance: if I want to get a value from stdin, then use that value safely; or make a network request and then use the result safely; or run a function that can fail, and use it or short-circuit as needed.
It wasn't until I discovered Erlang in 2012 (thanks, Seven Languages in Seven Weeks) that I finally found the motivation, aided in no small part by the fact that it's a very simple language and it's designed for server programming, where I've always been happiest.
I still haven't graduated into category theory or type theory. I still don't know the difference between a monad and a monoid. But functional programming really speaks to me, because I have an old, tired brain and I need pure functions wherever I can employ them to keep things straight.