It's influential but the most is stretching it. It showed that monads can be used to isolate I/O but have a very real cost when it comes to complexity, that la

It's influential but the most is stretching it. It showed that monads can be used to isolate I/O but have a very real cost when it comes to complexity, that laziness by default is a bad idea and that typeclasses are a good way to introduce ad hoc polymorphism. Definitely a good result for a research language but far less ground-breaking than ML was (admittedly putting the bar very high).