I wish they trained the bots to be a notch more relaxed and less hysterical. Less is more.

But maybe that's just a consequence of the RL training being essentially AI torture to make them do what we want.