The AI revolution has already begun. In education circles, AI continues to buffer out STEM as the buzzword du jour. Splashed across headlines, inboxes and policy agendas, AI and its consequences are now on the minds – and in the hearts – of those steering Australian schools into the future.
Innocuously, algorithm-matching systems have taken over many of the processes that keep schools ticking. From automated grading, to online tests that take the learner down a unique rabbit hole of inquiry, to physical chatbots popping in to direct classes, AI is
quietly infiltrating the sector even as the debate over its use still rages.
But just how will these emerging systems of intelligence change the nature of teaching and learning?
Do teachers face an uncertain future, as automated robots expand their repertoire of skills and expertise? Could we be heading towards a scenario where educators are mere slaves to clever machines? Or could AI offer great promise in our classrooms? These are the burning questions we grapple with, as we plough on into the (hyper intelligent) future.
Will AI replace teachers?
As we hustle towards automating our classrooms, there persists a common fear among educators. If technologies can perfectly play the role of learning facilitators, without the risk of having to take sick days, or demanding a pay cheque, then what is the point of hiring humans for the gig?
According to Professor Neil Selwyn, from the Faculty of Education at Monash University, this bubbling panic makes sense given the wider circumstances.
“It’s a reasonable fear to have, because companies are pumping billions of dollars into trying to do just that,” he shares.
This is not the time for complacency, the expert says.
“So even though (some) teachers and educators are quite complacent about not being replaced, you could argue that that’s false complacency ... I think educators need to speak up about what they can add value to, because systems are already being developed that will automate a whole bunch of stuff that takes place in the classroom.”
Faced with the threat of AI, Selwyn argues that over the next 20 years teachers will be placed “under increasing pressure to convincingly justify their existence”.
Nevertheless, he is convinced teaching will never be the sole domain of machines.
“What is interesting is that teaching is always assumed to be quintessentially human,” he begins.
“If you think about learning as involving a conversation between a knowledgeable expert and someone who wants to learn, and the emotional and pastoral stuff, that is always seen to be a bit beyond the pale of something that can just be automated.
“So even those that are really pro-AI in education, most people will say we still need a human to do some of the more ‘touchy feely’ stuff, and the stuff that is going to be automated is the more procedural.”
Selwyn contends that despite forecasts about the demise of many professions, and the declining need for human doctors, lawyers and accountants, education is one sector that is safe.
“Education experts and educators have tended to be quite complacent that their job is way down the list of occupations that can be automated.”
Those who are pushing the AI agenda on teachers are likely to have limited understanding about the nuances and complexities of the profession, Selwyn says.
“I’d be quite insulted if anyone told me, as a teacher, that a machine could do what I do … a lot of the pressure for these changes is coming from outside education, so it’s corporations, it’s policy makers, it’s possibly even parents and employers,” Selwyn notes.
“They are seeing what a teacher does in a very simplified, instrumental, procedural way, and so that’s the insulting bit – people aren’t recognising teachers for the professionals they are; professionals that have very detailed knowledge, and a lot of it is to do with interpersonal skills…”
The expert says the most crucial parts of teaching occur in the interactions that take place in the classroom – the moments of teacher-student connection that play out “under the radar”, unbeknown to principals or officials.
“Only teachers really know about that,” he says, “and that’s why I think we need humans, to do all the stuff that machines can’t do.”
It’s a sentiment echoed by Toby Walsh, Scientia Professor of Artificial Intelligence at UNSW Sydney.
“There’s very important parts to being a teacher; understanding that emotional and social intelligence is something that computers don’t have…” he notes.
Even if machines were able to reach a human level of intelligence in the next 50-100 years, Walsh is still convinced that the demand for human educators will prevail.
“Teaching is a very people-facing profession, and all the people-facing professions are the ones that will be the safest ones,” he says.
“Even if we could get computers to do [those jobs], I think we would just prefer people to do them.”
AI in the classroom
Gold Coast teacher Kirsten Ford is mid-way through her English lesson with a bunch of independent and inquisitive Year 7s.
Having pre-programmed an AI-powered ZenoBot avatar, the educator from Trinity Lutheran College is free to meander round the class while her capable assistant delivers the more formulaic snippets of tuition. Able to speak more than a dozen languages, respond to student questions and record their answers, her assistant has quite the skill-set.
“I think the ZenoBot can help with some of the mundane aspects of teaching,” Ford shares.
Indeed when the opportunity to trial the technology arose, Ford had no qualms about seeing how it might alter her practice.
“We were using it to basically program little lessons of more of the rote learning [tasks], so basic grammar lessons on verbs or nouns … so the kids were able to tap in and use the ZenoBot, who would deliver the instructional section of the lesson…”
Although admitting her new-age colleague still needs a great deal of fine-tuning, Ford is now excited for AI’s potential to offer “an extra hand in the classroom”.
“It’s hard work managing 25 kids!” she laughs. And yet, there is a long way to go with this.
“I’ve been giving it a lot of thought … could I help to develop [the technology] to be more personable for the students? Because that’s where I feel it is letting kids down at the moment; it’s still quite robotic, and actually, in my personal experience, my students prefer the sound of my voice, they prefer my stories in between teaching, they prefer my energy…”
Selwyn is certainly not sold on the idea of ZenoBots and the like assuming the role of teachers’ assistants. In fact, he fears we might be unwittingly traversing a very slippery slope. One that leads to more control, more regulation. The result? A profession crushed by its own intelligent technology.
“There is a fine line between having an assistant and having a supervisor,” Selwyn posits.
“Are we going to have teachers directed by the assistant? We know from every other piece of technology that is used in schools, that despite the best wishes of designers, they are often used in ways to control, monitor and standardise teachers’ work.
“Now that’s my concern: that the assistant slips into being more of a director, and then you don’t need specialised teachers, you just need someone who can follow instructions and teaching becomes less of a high-status professional job.”
It might be a stretch to suggest Ford’s handy avatar could end up stripping her of a career come Term 4, but as the technology improves, and the kinks and limitations are absolved, Selwyn’s concern burns a little brighter.
“Teachers [might] become de-skilled and you or I could teach VCE physics because all you have to do is point the students in the right direction, so it’s a really interesting area,” he concludes.
This June, however, Professor Rose Luckin told EduTECH delegates in Sydney that they need not fear being displaced by AI.
Often quoted for her insights into how these new forms of intelligence will transform the way we teach and learn, the professor of Learner Centred Design at University College London is certain that AI and teachers will always have to work in tandem – not in exclusion.
Luckin believes that ‘super intelligence’, the kind that can be used to help teachers unlock and monitor deep learning, will require a combination of human intelligence and AI. She says that machines simply do not, and will not, ‘get’ the social realm of existence.
China: a case study in AI
While the West sits around pondering the implications of AI technology, picking apart test scores and fretting over international league tables, China is running a very different race. An AI race, to be precise. And it hinges on training its students to become the best computer scientists in the world.
At least this is according to Alex Beard, a former UK teacher who now works at Teach for All, a global network that aims to “tackle the complex challenges facing children in disadvantaged communities”.
Beard travelled to the communist country to see just what was happening in schools.
“China is playing the ultimate long game with its AI strategy,” he tells Australian Teacher Magazine.
“In the West, we’re focused on creating the right economic incentives, or finding ways to import talent from abroad to fuel our AI industries. Winning the long game depends on developing that talent internally. China not only has the raw materials – hundreds of millions of kids – but it has the knowledge of how to run a high quality education system, the culture of learning to sustain it, and a system of government – for all its many flaws – that allows true long-term planning.”
Recently, the South China Morning Post reported that almost a quarter of the country’s schools are trialling ‘thinking’ technology to mark students’ essays. The AI-driven grading machine is designed to assess everything from linguistic style and structure, to the sophistication of logic presented. Minus the human error, of course.
Yet the pilot hasn’t been all smooth sailing. The publication notes that parents from the majority of schools taking part were not informed, and that test results were strictly classified. In some classes, even the students had no idea their essays had been assessed by a machine.
Taking things up a notch, in May Sina News, an online news portal, published a series of photos showing three cameras perched above the blackboard at a middle school in Hangzhou. They had nothing to do with security. Rather, using facial recognition technology, the cameras monitor students’ engagement and attention levels as they sit at their desk.
“Since the school has introduced these cameras, it is like there are a pair of mystery eyes constantly watching me, and I don’t dare let my mind wander,” one student is quoted as saying.
Beard says this kind of measure unearths some gnawing ethical questions.
“Facial recognition technology can undoubtedly increase the efficacy of online learning experience,” he begins.
“Imagine an AI that is continually watching you, reading your expressions and tracking your behaviours to understand when you’re most highly motivated to solve maths problems. Machines already have the capacity to know us better than we know ourselves. Yet is that a world we want to live in?”
As Walsh puts it, “the industrial revolution provided us with machinery that liberated our muscles and in some sense the AI revolution will provide computers that will liberate our minds.”
The question is, will Aussie educators be part of the conversation?