By Jodan Perry
Universities are grappling with effective approaches to guide students about GenAI use. My answer has been practiced by my people for tens of thousands of years.
I am a Worimi man, and I grew up learning through story on our Country. Bush walks with my grandmother at Karuah; my father’s voice at night; my aunties and cousins spinning yarns around campfire and torchlight. Stories containing life lessons, heroic ancestors, plenty about spirits, but overall, making sense of the world.
We didn’t write and read these stories. We told them.
GenAI produces a lot of stories too. It never runs out. Ask about Aboriginal culture and history, and you always get the confident, fluent answer. But there’s no connection to Country, respect for protocol, and no responsibility that comes with knowledge passed down through relationship. As Dr Terri Janke says, ‘AI has no Dreaming’. What it produces is disembodied, not grounded.
When I started noticing the common ‘AI slop’ in submissions, I picked up the ‘tools’ myself. What I realised was outside of the ‘profoundly delved deep’ writing everyone notices, was how this technology flattens both cultural knowledge and the intellectual labor of Indigenous peoples. It’s not good enough to attribute fake articles to authors, to confidently present their concepts wildly out of context, or to remix sources and spit out garbage, and hand it in.
I wanted to create a meaningful experience for first-year students
Indigenous peoples have fought for generations for spaces in institutions that have excluded us. Our pedagogies and knowledges are being embedded more meaningfully in education. If students (or teachers) use GenAI carelessly, like a pulse blender of words to quickly get out a bland-tasting smoothie, they aren’t learning about our people. They’re repeating an extractive process that’s occurred since 1788.
Blanket restrictions are about as useful as ‘AI detectors’. This was not the way. I wanted to create a meaningful learning experience for first-year students. They were already coming to university, with its troves of digital systems, documents and compulsory click-through modules, and inconsistent GenAI guidance across their courses.
I knew I couldn’t try and cover everything. They are already overwhelmed. So I shared my story of how I conceptualised GenAI, and my role between the new technology and where I come from. It’s called The Two Trees. One is rooted in the land, one is floating. I teach them what these are, why our knowledge matters, and provide a framework (S.T.O.R.Y) for how they can walk the bridge between the two trees respectfully.
This is the ‘trunk’ of my learning module ‘AI: Our Way For Our Course’. It’s a one-hour, video-led, interactive experience. I embed it before the Learning Path in many first year Indigenous Studies courses at the Wollotuka Institute, I’m on camera, walking alongside students as a fellow learner, not a machine learning scientist. I use GenAI throughout and model how I want them to use it in the course.
The Two Trees builds across three videos before the full framework is revealed at the end. Along the way, students:
Two trees
– Audit AI-generated text about the Voice Referendum and find fourteen problems that mirror exactly what I’ve seen in submissions. For example, GenAI hallucinating papers from the mid-2000s claiming they discuss the result.
– Navigate realistic scenarios from their own coursework and choose how to act, for example, what will they choose to do when a group work member uses GenAI to research, and produces parts of their assessment that contain factual errors?
– Reflect on relationships with knowledge: ‘Think of something you know deeply. It could be a skill, a story, a practice. What taught it to you? And what did they ask of you when they shared it? What were you expected to do with it?’
The module is living, first running in July last year, and already redesigned once. GenAI is outrunning academic publishing cycles and university policy reviews. Conceptualising through story holds meaning in ways that picking apart prompts and outputs (from endlessly updated models) cannot. It’s easy to recall and then apply.
After reaching over 500 students so far, feedback has revolved around two themes: ‘so different’ and ‘very human’. The research on why is underway. But this wasn’t a surprise. A recurring reflection since entering the academy in 2023 has been: ask a question, receive a link to a document. Information based-textual transfer of knowledge. The opposite of Indigenous oral storytelling.I could get existential here
Linguists are suggesting GenAI may be pushing education back toward its oldest form ‘to understand language – and perhaps even thinking itself – we need to start with the spoken word’. I could get existential here, riffing on consciousness or what it means to be human in the so-called ‘intelligence age’. But what resonates for me is simpler: the connection between people when we speak, learning from each other, sharing stories. Yarning.
Universities are being called upon to value ‘narrative and relational approaches to AI ethics alongside analytical frameworks’ while students are describing current institutional guidance as ‘vague or lagging’.
Indigenous Storytelling has always done this work
This responds to both. Indigenous Storying has always done this work. It’s what we know. It’s what we have always done. I just applied it to this context. As my father writes,
‘Many stories are not set in the Dreaming, but have developed over time and accompanied us as we travel forward into a changing future … and provide answers to many of life’s questions.
Stories become a collage within your memory, which impacts on how you view the world. Perhaps one small story may not change the path that you travel, but it has that possibility as many of those stories relate to life journeys’.
Old ways. New world. Same practice.
This article was originally published on EduResearch Matters. Read the original article.
The world is changing fast. We help you keep up.
We’ll send you 1 post, 3x per week.
