AI Already Controls What We See Online, Should It Also Control What We Say?

Silicon Valley is becoming increasingly enamored of the idea of using AI to proactively delete “offensive” speech before it is ever seen, but should AI control what we say in addition to what we see?