What even the fuck do you say to this? Like, if Open AI can't get basic shit like this even vaguely right, why would anyone trust it for anything, especially more consequential matters?