This one really caught my eye. The way they described “assessment” practice here (and how “good” the AI is at providing “positive feedback” first) was so resonant to work in humanities; specifically writing (my field); reminder for how key humans are in good feedback practice and how hard it is.

YES! Like, is this actually materially constructive feedback, or does it just reault in people aiming at and only at the rubric? Because thosr are different