[e-dant]

The issue this paper is grappling with (to what extent humans have a place in the middlespace between them asking an ai to do something, and the ai doing it) is interesting (although I disagree with how the paper tries to solve it).

I’m empathetic to a non-native English speaker using ai to help communicate. I mean, seriously, I would do the same thing if the lingua franca was Japanese.

The author is here saying, well yeah, there’s this weird thing that happens when I use ai by which the ideas that come out the other end are a bit different from what the author intended.

I think the sentiment “well you shouldn’t have used ai” is incomplete.

The paper is not great, but it’s an interesting question.