It's likely bumping up against people's desire to have the model complete a given task without asking the person to intervene a bunch of times.
Seems unclear how you satisfy everyone here.
Give the model judgement
Give the model judgement