And they were right, too. Language is far more homogenous these days than it's ever been. As for the parent commenter's point, as we encode in software what "correct" language sounds like, it will naturally exert pressure to conform to whatever notion of "correct" that you chose; conformity is literally what things like spellcheck are for. And it's not necessarily a bad thing for languages to be more uniform, especially from the perspective of non-native speakers, but it remains true that it will constrain the future mutation of the language, even if it won't totally halt it.