Why You Should Treat AI Like a Smart Student
Let’s be honest – as teachers, we have favourite classes.
One of my favourites was a senior business class. It was full of intelligent, creative, and engaged students. The conversation flowed, the chemistry was great, and we got a lot of learning done.
There was one student in particular who, while extremely capable, was sneaky. Let’s call him Billy. On his tests and assignments, Billy would always answer with full confidence, even if he had no idea of the answer.
When marking his work, I would often be impressed by the amount of referencing he was doing. His references looked solid. Until I started to verify them. He’d made half of them up…
This made Billy tricky. He was intelligent, engaged, supportive. He did some of the best work I’ve ever seen in the senior business class. When he understood something, he really understood it and could explain it to others.
However, when he didn’t understand it, he acted like he did. When he had no references, he made them up.
He had flashes of brilliance mingled with complete fabrications.
Here’s the key point: you need to treat AI like you would treat Billy.
Don’t always expect it to understand you at first. You will need to provide guidance along the way.
Even when it does understand the task, you still won’t get what you want the first crack. Hemmingway said all good writing is editing. All good AI use is iteration.
Double-check all the output if you’re asking for facts, data, and analysis. In my experience, AI is often directionally correct but can fudge key details. This is not what you want if a key detail ends up being part of a lesson plan or reading.
But, don’t write off Billy because of these things!
As a teacher, my life would certainly be easier if I had a couple of Billys to run errands for me.
Would I get Billy to write a parent letter, put together a few ideas for a new unit, or turn my dot points into full sentences for student feedback? Absolutely. Would I treat any of his work as a final draft? Absolutely not.
Conclusion
Why am I asking you to think of AI like Billy? Because most people default to thinking of it as a super-intelligent guess-what’s-in-my-head-machine. They expect perfect results the first time and every time. They expect 100% accuracy and a three-second turnaround. AI may get there one day, but it’s not there yet.
If you can change the way you think about AI from a super-intelligent guess-what’s-in-my-head-machine to a diligent but overconfident senior student, you’ll have the right basic assumptions to get the best out of AI.
Until next week,
Happy Teaching!
Paul Matthews
