Köşe Yazısı

I'm Done With Self-Checkouts — and You Should Be Too1

I have a confession that may cost me a few friends: I actively enjoy queuing for a human cashier.2 In an age obsessed with speed, I will happily wait three extra minutes to avoid the beeping, blinking nightmare that is the supermarket self-checkout machine.

You know the scene. "Unexpected item in the bagging area." An accusatory red light. A weary staff member sprinting over to type in a code for the fourth time.3 We were promised effortless convenience; what we actually got was unpaid labour and a robot that treats every one of us like a suspected shoplifter.

And it is not merely annoying. Retailers themselves admit that self-checkouts have fuelled a real rise in losses, and many shops are now quietly bringing human staff back.4 Meanwhile, for every machine installed, a genuine job — and a genuine conversation — silently disappears.

Because here is what we lose that no efficiency report ever measures: human contact.5 For an elderly neighbour living alone, a cashier's simple "How's your day?" might be the only conversation they have all week. Is saving ninety seconds honestly worth that?6

So next time, walk straight past the machines. Choose the human queue.7 It is a tiny act of resistance in a world that keeps trying to automate us out of each other's lives — and your local cashier will thank you for it.

Let Students Use AI — and Then Teach Them to Argue With It1

Every few weeks, another school proudly announces a total ban on artificial intelligence, as though a firewall could somehow hold back the tide.2 I understand the panic. But I want to argue for the opposite approach: we should hand these tools to students deliberately — and then teach them to distrust every word they produce.

Let me be fair to the sceptics, because they are not wrong to worry.3 A student who lets a machine write every essay learns nothing except how to type a prompt and paste the result. The muscle of writing — of wrestling a messy thought into a clear sentence — simply never develops. If we do nothing at all, we risk raising a whole generation that quietly outsources its thinking before it has ever properly learned to think for itself.

But banning the technology solves none of this; it simply pushes it into the shadows. Students already use these tools at home, at midnight, in secret, with no one at all to guide them.4 A ban does not stop the behaviour — it only guarantees that it happens without supervision, without discussion, and without anyone ever pointing out where the machine quietly got things wrong. The real question was never whether our students will use AI. It is whether a single adult will bother to teach them how to use it well.

And the numbers make the pretence absurd. Surveys already suggest a clear majority of students have used AI for schoolwork, whatever the official rules happen to say on paper.5 We can keep writing rules that everybody breaks, or we can grow up and face the world as it actually is. Pretending otherwise is not principled. It is simply more convenient for the adults in the room, who get to feel they have done something without having to learn anything new themselves.

So imagine, instead, a classroom where a teacher projects an AI's confident, fluent and completely wrong answer, and asks thirty teenagers to tear it apart.6 That is not cheating. That is the single most important lesson we could teach them: that sounding certain and being right are not remotely the same thing — a lesson, frankly, half the internet still hasn't learned.7

The tools are not going away, and no ban will make them.8 Our choice is genuinely simple. We can keep pretending, and let students learn the hard way, alone and unguided. Or we can do our actual job — and teach them to think alongside the machine without ever once surrendering to it. I know which classroom I would rather sit in.