Informa AI in Education Panel

• David de Carvalho, CEO, ACARA
• Cameron Paterson, Director of Learning, Wesley College
• Mark Grant, Chief Executive Officer, Australian Institute for Teaching and School Leadership (AITSL)
• Dr Catherine McClellan, Deputy CEO, Research and Assessment, Australian Council for Educational Research (ACER)
• Moderator: Dr Jane Hunter, Associate Professor Professional Learning, University of Technology Sydney

My comments –

The dance of love and fear – teachers love that our reports now write themselves and fear that students won’t submit their own work. Things are changing so fast we can’t keep up. And we need to be comfortable with that. When I’m uncertain, I ask the kids. I’m always blown aware by their insight and awareness.

Earlier this year a Harvard edtech professor ran an experiment with their class. Unbeknownst to the students, half received their feedback from the professor and tutors, while the other half received AI-generated feedback from ChatGPT. At the end of the course the students who had received the AI feedback performed better and rated their enjoyment of the class higher. One student who had received AI feedback without knowing it, stated that while the class was a little boring, receiving personalised feedback was the highlight.

The use of ChatGPT to generate essays and homework is rife and I don’t know that the take-home essay will survive. We will see more in-class writing and the rise of a new breed of oral assessments – TED talks meet the viva voce.

As a history teacher, AI can take my students to any point in time, from any perspective. Ask ChatGPT to be a German immigrant arriving in New York City in 1885 and describe the sights, sounds, and emotions. Prepare for an extraordinary conversation. (But perhaps don’t try asking it to take on the perspective of an indigenous Australian on the shores of Sydney Cove in 1788 and describe the First Fleet sailed in. It might use the word “excited”, an appropriate caution for us). But, we can talk to John Curtain, talk with the sheep-shearers in Banjo Paterson’s The Man from Snowy River, and even talk to the Murray River. Try it. Plato, Rousseau, and Dewey can tell me what they think of our College strategic plan.

AI means better teaching and less work. It can design exceptional lesson plans with:

• check-ins
• role plays
• low-stakes quizzes
• exit cards

It can provide scaffolds (for those who require support)

And challenge (for those who need to be extended)

AI provides faster and more accurate feedback and it can collate written feedback from a student’s essays over time into one neat report comment.

Artist-scientist Michelle Huang created a time portal to her childhood by training AI on her childhood journal entries, allowing her to converse with her younger self. Envision a future in which, instead of childhood journals, the corpus is learning reflections gathered throughout schooling, allowing students to have conversations about their learning experiences that help them gain insights into their own thinking.

Influencer Caryn Marjorie has created an AI clone to monetise her persona. She reckons it can make her one million dollars a month. What are the implications for teaching?

Remember that this is the worst version of AI our students will see for the rest of their lives. It will get exponentially better from here. It is only a matter of time until AI can:

• analyse student discussions
• create visual representations of learning moments
• coach students to ask better questions

Just as previous generations learned to work with different cultures and genders, the next generation will learn to work with AI and robots. It is going to be tricky while we learn to be AI compatible and rely on backup memory. Young people will lead the charge in developing innovative ways to utilise it.

In the longer term, our view of what it means to be human is going to have to change and we don’t have mental models for such a shift.