When Polished Work Feels Empty: Judgement Begins Where Scripts End

You read the submission once, then again. On the surface it is perfectly serviceable. The sentences behave themselves. The structure is tidy. The claims arrive in the correct order and leave no obvious mess behind. It is the sort of piece that should let you keep moving. And yet something in it does not carry any weight. The writing looks complete, but the thinking feels oddly unfurnished. When you ask the student a simple follow-up question, they can circle the answer without ever really entering it.

Most teachers know this moment. Many are less willing to trust themselves in it than they should be. They have been trained, often quite aggressively, to prefer what can be codified, standardised, and justified in advance. So when a piece of work feels slightly off, there is a temptation to treat that perception as professionally embarrassing, as though judgement only becomes legitimate once it can be reduced to a rubric descriptor. That temptation should be resisted. The sense that something is not sound is not irrational leakage from an otherwise proper process. It is part of the process.

This matters because teaching has never been the execution of a script. It has always involved reading situations that are not fully closed, interpreting signs that are real but not always easily measurable, and acting without the luxury of perfect proof. Students do not arrive as controlled variables. They arrive with prior knowledge, anxieties, sudden insights, and forms of competence that rarely line up neatly with the boxes we build for them. The teacher’s work has always included discerning what sort of learning, or imitation of learning, is actually in front of them.

That is why the polished-but-hollow submission rankles. It reveals something that procedure alone cannot settle. So what do you do? Do you make a grand accusation that you cannot prove, do you let it go because you are not prepared to die on that hill, or do you simply require everyone in the class to do an oral defence of their work? The answer to bad judgement is not the abolition of judgement. It is better judgement. Sometimes that means asking the student to defend the work orally. Sometimes it means stepping sideways rather than confronting head-on: asking what changed in their thinking between draft and submission, what they found difficult, what they rejected, what the opposite argument would be, or why they made one move rather than another. The point is not to catch them out. It is to move from product back to process.

The process of learning was always important, even before artificial intelligence entered the arena. It’s just that the education system has idolised assessing the outcome for so long. Now, the onslaught of AI reminds us that the central problem is not technological novelty but educational perception. The teacher who senses that a response is formally competent yet intellectually vacant is already doing the deepest part of the work. They are judging the quality of the process that created the outcome. 

There is a reason this part of teaching remains stubbornly resistant to managerial fantasy. Judgement does not behave like a checklist. It cannot be fully pre-scripted because the situations it addresses are open, human, and variable. The demand for certainty does not remove that reality. It only hides it beneath cleaner paperwork.

The sharper question, then, is not whether teachers should judge. They already do, and must. The question is whether schools still have the nerve to admit that professional discernment begins precisely where the AI script runs out.