“Sarah Rogers asserts that critical policy decisions about AI should be made through democratic processes, not by tech executives, citing a past example where Twitter employees prohibited referring to a convicted male sex offender as male.”