Why AI should strengthen, not weaken, the student-teacher relationship

on

For years, I thought my job was to stand at the door. I didn’t put it that way, of course. But when I began teaching in India, and later in Fijia and Australia, I took it for granted that students came to me because I held something they didn’t. I chose the readings, explained the theories and marked the work. Like a doorman, I decided who got through.

The idea of a doorman comes from Rory Sutherland, the advertising executive who coined the “Doorman Fallacy” in his 2019 book Alchemy. In business, it describes the mistake of reducing a role to its most visible, basic task, then automating or cutting it to save money while ignoring the hidden human value it provides. A hotel sees a man who opens doors and replaces him with a sensor. What it loses is the person who greeted guests, hailed taxis, noticed who looked lost and made the place feel safe and looked after. The door was never the job.

I think higher education is now at risk of making the same mistake, and in two directions at once. Institutions may decide that a teacher is mainly someone who delivers information, and that a machine can now do that more cheaply. Worse, teachers may have come to see themselves that way too.

AI has made the question hard to avoid. A first-year student stuck on a media or management theory can ask a chatbot to explain it plainly, give three Australian examples and try again when the first answer doesn’t land. No waiting for next week’s tutorial. The teacher is no longer the only person who can explain things, and some of my colleagues sitting in empty classrooms definitely find that frightening.

Their worries are not silly. We all have marked assignments that read beautifully and say almost nothing. The grammar is flawless, the structure tidy, and the evidence thin or missing. AI can produce convincing language without producing convincing thinking, and any marker will tell you that gap is real.

But guarding the door harder is the wrong response, and I have seen where it leads. Teachers become investigators, students become suspects, and a classroom built on curiosity turns into one built on suspicion. Detection software cannot settle the matter either. A probability score is not evidence, and no student should have their degree turn on one.

I would rather ask a different question. Not “did a machine write this?” but “can you explain and defend it?”

Just ask a business student to present basic SWOT or PESTLE analysis and then tell you why one factor matters more than another, and you learn quickly whether they understand it. Ask them what the analysis leaves out and you learn more. Oral defences, staged drafts, in-class problem-solving and short reflections all work this way. They are harder to fake, and they treat students as people with minds rather than files to be scanned.

This is also where teachers show the value that doesn’t appear on any job description. As someone who crossed from India to Australia, I remember what it is to learn an unwritten academic culture: how directly you can disagree, what independent study really means, when asking for help counts as strength. No chatbot taught me that. People did, mostly through small acts of patience.

I see the same thing in my classrooms now. The student who goes quiet at the back may not need a better explanation. They may need someone to notice, and to say their question was a good one. A confident student may need the opposite, a teacher willing to push back. AI can explain a concept in ten different ways. It cannot yet sense that something is wrong with the person asking, or care what happens to them next.

None of this lets anyone off the hook. Educators must become learners again, willing to experiment with these tools and understand their limits before judging them. Students, for their part, must accept that access is not the same as understanding. Using AI doesn’t excuse them from checking what it says, owning their conclusions and being honest about the help they used.

What holds this together is trust, not policing.

Sutherland’s hotel learned its lesson only after the doorman was gone and the lobby felt emptier. Universities have the chance to learn it earlier. The point of a university is not who gets through the door but who walks out the other side: people who can think for themselves, challenge what they are told and use what they know well.

If we cling to the door, AI will make us irrelevant. If we put our energy into what happens beyond it, the conversation, the challenge and the mentoring, AI may remind us why we became teachers in the first place.

Support our Journalism

No-nonsense journalism. No paywalls. Whether you’re in Australia, the UK, Canada, the USA, or India, you can support The Australia Today by taking a paid subscription via Patreon or donating via PayPal — and help keep honest, fearless journalism alive.

Add a little bit of body text 8 1 1