According to AI Author and Founder of The Next Word, Matthew Esterman, in any given school staffroom, there is a spectrum of teachers who relate differently to AI.
On the one end, there are the “do not talk to me about tech” types, he said, whose skin bristles each time it comes up in a morning tea chat.
On the other end, are the “boots and all” types, who go gangbusters over LLM (large language model) releases and cannot wait for the tech to permeate more areas of their working life.
And while each staffroom differs in its make-up, Matthew – who co-authored The Next Word: AI & Learners – says all share one common trait:
A nagging desire to know more about AI’s role in teaching – whether driven by dread or curiosity.
“We have had four years of ChatGPT now and people are starting to realise it cannot be ignored, it’s not just a fad,” he said, ahead of The Sydney Morning Herald School Summit.
“It’s a bit like continuing to teach during a hurricane. Yes, it’s possible, but eventually the roof is going to blow off.”
Implications for assessment
One of the biggest sticking points with AI’s role in teaching, Matthew says, is how it might impact student assessment. Generally, the more AI-exposed a task is, the less reliable traditional examinations become.
“AI questions every single part of what assessment is. When we give people the digital equivalent of a PhD-level expert that acts as their study or work buddy, that fundamentally changes the way we should examine their ability.
“Moreover, why should students do a three-hour sit-down exam when it’s unlikely they’ll ever have to do that form of assessment as work, or in life going forwards? It’s no different to testing their mental arithmetic when calculators exist – totally pointless as a leading metric for future employability or success post-school.”
As to which form of assessment is most appropriate for AI-exposed disciplines, Matthew says it’s an ongoing debate. With AI now part of a student’s study group, the reliability of every assessment type has been called into question.
“We’re still having conversations about [it]. We always have, and I guess we always will. But we do know that traditional assessments alone aren’t going to cut it. And that, of course, is a reality that we can’t ignore – no matter your AI personality type.”
Differentiation
AI’s role in learning differentiation is also a worthy discussion – and one which could help win over the most trepid technology adopters, Matthew said.
At present, it is largely up to teachers or teaching aides (TAs) to tweak lesson content to match students’ circumstances. But with TAs often stretched too thin, some schools are turning to AI to automate part of the task.
“Say you’re a primary teacher, you’ve got thirty kids in the room, each of whom has a different learning need, which might change from day to day. You can now say, hang on, Student X is in the classroom today. I know her background and what she normally needs. But I also know she’s just had a traumatic experience in her family. So maybe we don’t have to talk about that topic today. Maybe we’ll get to that later.
“As a teacher, that’s so much on your brain, but now you do have this assistant who can help you think instantly about how you might tweak the lesson that day for Student X, so that she’s learning more effectively in her circumstances. That was just impossible unless you had several human assistants in the room with you. But now we have that capability in the palm of our hands.”
When this is done responsibly, by a teacher who knows how to talk to the technology, Matthew says it can be transformative.
“It can provide a rich, connected, and highly personalised experience with students,” he said.
Workload management
Matthew says it is natural for people to feel alarmed upon hearing that a large chunk of their professional work could be performed by a robot. But on closer inspection, many people realise that they are already automating parts of their role.
“Sometimes I ask people who seem cautious about the AI revolution in teaching, ‘do you use ChatGPT?’ And they say, ‘yes of course I do because it helps me with emails, for example.’ And I say, ‘okay, so you’re already automating part of your job. You’re choosing the tasks it does, so you feel in control and it’s less scary that way’.”
Child safety
Like it or not, Matthew says AI is now “part of the plumbing” and that its governance is now a child safety issue, as much as anything else.
With statistics showing that one in five high schoolers either has or knows someone with an AI boyfriend or girlfriend, he says even AI-avoidant schools will need an AI strategy.
“That’s not just a technology issue. That’s a welfare issue. And it cannot be solved by just giving or restricting access to a certain platform. There is a lot more work to be done here.”
Continuing the conversation
Sharing more on AI’s role in teaching and schools, Matthew Esterman will lead a pre-conference masterclass for leaders who want to get “hands on” and build something that works in their context.
The masterclass will cover:
– AI First Aid – the essentials of what AI is and can do
– Assessment Architecture – how to rethink and redesign assessment with AI in the loop
– The Differentiation Engine – how we leverage AI to reach more students with effective teaching & learning.
Learn more and register your tickets here.