Educational technology (EdTech) solutions risk being marginalised by ChatGPT and other generative AI tools, an inquiry has heard, as teachers and students use it to create lesson plans and write assignments indistinguishable from those written by humans.
“EdTech generally has overpromised and underdelivered,” Dr James Curran, CEO of technology education provider Grok Academy said as the House of Representatives Standing Committee on Employment, Education and Training kicked off its Inquiry into the Use of Gen AI in the Australian Education System.
Many current EdTech resources are “honestly not that much more than PDFs, videos, and multiple choice put in an online system,” he explained.
“Rarely do they actually do things that we know are critical in teaching” such as explicit teaching, quality feedback, adaptation, and “actually responding to mistakes and misunderstandings in interesting ways.”
With ChatGPT promising to fill that void, Curran flagged the “two-speed problem” facing an education system where students are embracing Gen AI but “the teachers that are using the technology right now don’t understand it.”
“When technology is new, that’s when it’s particularly important that you understand how it works,” he continued, warning that pedagogically sound educational methods risk being overwhelmed by generative AI.
“There are lots of teachers that have managed to ignore that technology exists, and ignore that the Internet exists,” Curran said, “but I don’t think they’re going to be able to ignore that AI is going to change so many aspects of assessment and things in their schools.”
Despite piecemeal efforts, a lack of formal generative AI guidelines had left teachers at a loss: more than half of teachers in a recent outreach session reported receiving assignments that were almost certainly produced by generative AI – but, Curran said, “their school did not have processes and policies in place to deal with it.”
Teaching the generative AI generation
Clear rules will be crucial given that students’ use of generative AI now will shape their journey through high school, university, and into the working world.
An Australian Framework for Generative AI in Schools is expected before year’s end – with a recent consultation paper outlining six core elements including teaching and learning, human and social wellbeing, transparency, fairness, accountability, and privacy and security.
An August survey, released by the Association of Heads of Independent Schools Australia (AHISA), found that 24 per cent of primary teachers and 39 per cent of secondary teachers are using generative AI for tasks including lesson plans, learning design, curriculum unit outlines, rubrics for assessing student work, and class discussion questions.
The results confirmed that generative AI “presents some complex challenges for schools and our national school system,” AHISA CEO Dr Chris Duncan said while lauding generative AI’s ability to help teachers “quickly develop differentiated learning tasks for students.”
Yet Professor Julia Powles, director of the University of Western Australia Tech and Policy Lab, warned against overstating generative AI’s benefits, arguing that the technologies “present a grave risk of exacerbating the plight of our educational system.”
“They use computational techniques but have no notion of truth,” she explained. “They don’t know what they’re generating.”
“They aren’t some kind of magic silver bullet to everybody’s different personalised needs,” she continued. “In fact, there’s a great risk that they take you down rabbit holes quite a way from where the pedagogical basis of learning would be.”
“We are potentially letting loose technologies that really distract us from what we know about education, because of a lie of all-singing, all-dancing, magical, personalised learning assistants.”
Many students agree, with a recent University of Melbourne survey of 110 students and academics finding that 85 per cent of students who hadn’t used generative AI considered the technology as “cheating” – while 75 per cent of academics said their universities are not ready for the technology.
Professor Nicholas Davis, co-director of the University of Technology Sydney (UTS) Human Technology Institute (HTI), agreed – warning that “Your average school is very poorly placed to do thoughtful procurement of these systems, so advice and standards would be critical.”
“We need to absolutely ensure that generative AI is legal and in line with the expectations of the community and the law when it comes to gathering and using data,” he said, noting that “there are currently no standards for the efficiency, effectiveness, performance, or pedagogical efficacy of EdTech and similar products in Australia.”
Generative AI controls are, he said, “a fantastic opportunity for the federal government to set those standards for what’s expected around the transparency of those systems, and proving that there is some theory behind them.”
Objectivity about generative AI’s capabilities and limitations will be critical, warned UTS Centre for Social Justice and Inclusion industry professor Leslie Loble.
“We cannot be seduced by the technology,” she told the inquiry, warning that generative AI “is really a super fast, quite talented prediction machine that gives us the appearance of having creativity – and it’s learning from what’s out there, and making itself better.”
Research to date, she said, has shown that AI-driven tools like intelligent tutoring systems “offer tremendous promise because they adapt to where a student is at,” she said.
“More structured AI can actually have quite positive impacts,” she continued, “when it’s very closely tied to what we know works in education…. The evidence shows that it works better when it’s aligned.”