Robots could one day replace teachers - but should they?

The idea that robots could one day displace teachers is becoming increasingly plausible – but a new Ethical Augmentation Framework, focussed on learning objectives, is helping to keep the technology’s trajectory in check.

Developed by Dr Nicole Brownlie at the University of Southern Queensland, the framework serves as a decision-making tool to help schools determine when AI should be used in classrooms.

Nicole said it is “not anti-AI”, but that it seeks to identify and preserve what ought to remain human domains.

“Just because AI can do something in school settings, it doesn’t always mean it should,” she said. “And to determine when it should, we need to ask a series of questions.

“Does it keep our expertise central and are we still the owners of the work?

“Does it allow for a deepening of relational pedagogy? Does it increase the excellence of what we do and are we making a better thing at the end?

“And then, how do we account for how AI is being used and uphold responsibility for that?”

Red versus green flags

With these criteria in mind, Nicole said there is a strong case to use AI in various educational contexts. The first is in daily tasks that don’t require much oversight, and which consume significant time.

“For example, I personally spend a long time crafting one paragraph because I agonise over every sentence. Whereas if I have a conversation with OpenAI, I can say, I need a paragraph on this. I need to mention this and this, in this tone - and it can create a paragraph for me.

“That then leaves me more time with my students, to spend more time in marking and feedback, or to have a one-on-one conversation with them to ensure that they understand the content.”

Conversely, using AI to create new knowledge or offset key educator responsibilities may be inappropriate, Nicole warned.

“A red flag would be if AI does my work for me – for example, using AI to mark assignments without any human input. Students pay for educator expertise and offloading marking to AI undermines the trust students have in them to do that.”

Context is key

However, Nicole stresses that context is key and that using AI could be appropriate in some settings, but not others.

“It depends what you are trying to achieve. If we are using it in teaching, do we want students to be able to write a paragraph? Good, then they need to do that part of it.

“Or are we assessing their oral presentation skills? If so, it's probably okay if AI creates the PowerPoint based on the script we've written, because we're not assessing their ability to create a pretty PowerPoint. We're assessing the information they relay to us.”

In this way, Nicole said AI can help isolate skills, in the same way that certain physical exercises isolate muscle groups.

“It’s all about which learning objective you’re trying to target – and the teacher’s discretion comes into play here,” she said.

Intention matters

More broadly, Nicole warns that the schools’ intention for using AI also matters – and advises against using AI policies purely for punitive purposes.

“A lot of schools are using AI to try and catch the kids that are doing the wrong thing. And I think that's also a red flag,” she said.

Instead, Nicole said AI should be celebrated as a feature of the curriculum.

“It’s important for schools to teach students about working with AI in a way that doesn’t take away their knowledge - because they're going to have to use it after school anyway.

“Students might use AI to understand a topic, just like they would use a Google search or book. And they need to be able to determine if there's bias and if the output is correct. And from that, they need to take the time to be able to write it out themselves.

“If we can teach our students these skills, that's far more effective than trying to catch the bad kids, because the vast majority of kids try and do the right thing.”

Grey areas

While the logic seems clear, Nicole admits there are many grey areas and that not everyone agrees on which contexts warrant AI.

Additionally, public attitudes around AI are evolving – meaning a use case that was considered unacceptable yesterday might be more widely embraced tomorrow.

“There is this broad acceptance now that calculators can do mental arithmetic for us – and the same will be true for AI applications, as they become more ubiquitous and mainstream.”

Despite this state of flux, Nicole said the sentiment behind the framework remains constant.

“The essence is the same,” she said. “It’s all about why we are using it, what we want it to do, and how it will help individuals be better people."

Human connection

Learning objectives aside, Nicole said there will always be a need for human connection in schools and that AI should not interfere with that.

“Teaching is essentially human job,” she argued. “We make connections with our team, we make connections with our students, we make connections with our colleagues - and it's important we're able to keep that.

“Yes, in the future, a robot could theoretically stand in front of a class and deliver a lesson, but could it give us that same experience of education where we look up to someone – where we trust, confide in, or lean on someone to help us understand the world a bit more?”

Join the debate

Sharing more of her views and insights into her research, Dr Nicole Brownlie will deliver a keynote at the upcoming AI in Education Brisbane Conference.

This year’s event will be held 8 October at the Stamford Plaza Brisbane.

Learn more and register your tickets here.