Picture these three scenarios:
An expert literature teacher stands at the front of the class, waxing lyrical about the similarities and differences between Sylvia Plath and Carol Ann Duffy. The students are enthralled and will undoubtedly leave the classroom with a greater appreciation of the authors’ poetry. The teacher has heard about GenAI, but decided long ago that it didn’t really apply to their subject matter. Computers “aren’t allowed” in this classroom. But of course, students are required to submit their final essays on the learning management system. The teacher has no idea whether these final essays were written by AI or not, and although the students clearly enjoy the lessons and take a lot from them, a certain number will submit AI-generated essays because it’s expedient, and they know they’re likely to get away with it.
In the classroom across the hall, a mathematics teacher demonstrates the capabilities of code-writing reasoning models like GPT-5-Thinking to handle complex mathematics, using Python for data analytics. This relatively new teacher came via an industry pathway, arriving with a deep understanding of generative artificial intelligence and other AI-based technologies and how they apply in reality to the field of mathematics and data science. She enthuses about the possibilities of artificial intelligence in maths. But she doesn’t realise that three-quarters of the class stopped paying attention about 20 minutes ago, and what looks like a classroom full of students experimenting with artificial intelligence on their own laptops, is mostly students playing Tetris.
Across the school campus, an early career science teacher, not even fully qualified, working with permission to teach due to teacher shortages, frantically prepares for the next lesson: Year 8 Biology. Their university course, two-thirds complete, is in Humanities. They are well and truly out of their depth, but the school doesn’t have the resources, the staff or the time to provide the student teacher with a mentor. Like most of their peers, the student teacher is reasonably proficient in GenAI and feeds sections of the Year 8 textbook they’ve been handed into ChatGPT to generate some lessons. Seconds later, the chatbot extrudes the lesson plan: Item 1 – a 10 minutes starter activity class debate on cellular structure and recap on what was taught in the previous lesson.
The literature teacher only notices something unusual is happening when they spot a cluster of students’ work with oddly repetitious structure, limited vocabulary and an unusually high rate of em dashes… The maths teacher eventually realises that more than half of the class lacks the fundamental skills needed to instruct a GenAI chatbot to do the work… and the early career teacher comes to the conclusion straight away that in the first 10 minutes of the lesson, getting students into their seats and taking the roll is probably a priority over staging a full-blown debate.
Each of the three teachers in these vignettes has expertise in different domains. In a recent post, I outlined three dimensions of expertise necessary for using artificial intelligence well: technological, domain and situated expertise.
Domain expertise is disciplinary knowledge, subject-specific skills and the ability to design and deliver a rigorous curriculum. Technological expertise is an understanding of the ethical and practical implications of technologies including generative artificial intelligence. Situated expertise is expertise developed in the long term, a contextual understanding of classroom dynamics, institutional policies and teacher professional identity.
Our first teacher, the experienced literature teacher, has high levels of situated and domain expertise. They can walk into a classroom entirely unprepared and deliver a lesson so memorable that students will be thinking about it 20 years later. But without technological expertise, this teacher’s resistance to artificial intelligence sets them up for problems. Students, even the most well-meaning, will misuse Gen AI at some point. This is not a matter of malfeasant or conniving students trying to get one over their technophobic teacher. It’s pragmatism. Most of these students probably study four or five different subjects, and if they can get away with submitting a piece of AI-generated work and saving themselves some time, then some of them probably will.
The mathematics teacher has GenAI or Technological expertise and Domain expertise, but transferring from industry into the classroom means they lack the situated expertise to read the room. They can’t tell which students are struggling. They don’t have the innate ability, honed by years of experience, that teachers develop to negotiate the content of the lesson in accordance with their students’ capabilities. Their enthusiasm for the technology overwhelms some of the students and others are just disinterested. Most of what she offers is undoubtedly useful, but the majority of the students aren’t ready or able to receive that information.
And our out-of-field, early career teacher, not even fully qualified and operating on permission to teach, like so many teachers across Australia, turns to the technology to dig themselves out of a hole. But without domain or situated expertise to judge the quality of the lesson plan, both in terms of pedagogy and content, they too set themselves and their students up for failure.
So what can we do to help all three teachers?
A Professional Development Model for Three Dimensions of Expertise
My earlier post explored the three dimensions in detail. In this article, I’m going to make that more concrete and talk about how a school or university could develop a professional development model that guides educators towards expertise in all three dimensions.
Importantly, this model makes no assumptions about whether an individual teacher will use artificial intelligence in their classroom. There are many good reasons, some of which I outlined in an earlier article on resistance, that some teachers might want to keep the technology away from their students. This will vary subject by subject and unit by unit, but if artificial intelligence is not applicable in a certain subject area, there is no reason why that teacher should be forced to use it, either as part of their own professional activities or as a classroom resource. However, like our literature teacher, even if the individual decides that AI has no place in their course, they should have a level of competence with the technology that allows them to discuss GenAI transparently with their students.
There are also entirely valid reasons why a teacher who is an expert in their domain might use generative AI but simultaneously avoid using it with students in the classroom, because students are not experts. The three dimensions model of expertise supports this: until a student can balance domain expertise and situated expertise with the use of generative artificial intelligence, they might end up using the technology in a way which does more harm than good.
To frame the development of these three dimensions of expertise, I will use the Dreyfus model of skill acquisition, moving from novice, to advanced beginner, to competent, proficient and expert level. Not every member of staff in a school community has to reach the expert level, but some do. Not every member of staff in a school community is expected to be an absolute expert in their subject area. It is good to have staff who have proficiency across a range of subject areas, as well as staff who are experts. Expert staff are ideally placed to become faculty leaders. Proficient staff work incredibly well in junior and middle school contexts, where it is more common for teachers to have multiple disciplines. Teachers who are proficient across multiple areas are invaluable in the primary school context, whereas a teacher with narrow expertise in, say, physics, would probably find themselves out of their depth teaching a Year 2 cohort.
This professional development model, therefore, is highly contextual and needs different avenues for different types of teacher. Mapping Dreyfus’ skill levels to the three dimensions of expertise looks like this:
A Three-Phase PD Pathway
I think, given the growing ubiquity of Gen AI and related technologies, professional development in schools should include regular opportunities for staff to learn about the technology. However, there are many other things which teachers need to learn and AI PD cannot dominate the entire staff training calendar.
So the first stage is to explore where staff are already at in these three dimensions of expertise.
Stage One: Explore
A school needs to understand the skills and expertise of its staff. This sounds straightforward, but you would be surprised how little schools know about staff expertise sometimes. I was the Director of Teaching and Learning at a relatively small school with around 50 teaching staff. There were staff working at the school who had been there for over 30 years, some of whom had multiple degrees and qualifications in disciplines that I had no idea about.
Some of my colleagues had entered teaching from industry, but because of staff attrition and the need to fill spaces in various subject areas, had spent so many years teaching out of field that it was just assumed that was where their expertise lay. There were science teachers who should have been teaching humanities, arts teachers who should have been teaching maths, and maths teachers who were more than qualified to work in agriculture or technology. This kind of thing happens in schools all the time.
The same is true for technological expertise. Some of the staff, myself included, were serious computer nerds, but none of them were teaching digital technologies or really sharing their knowledge of digital technologies with their colleagues. Stage one is about exploring all of these hidden pockets of expertise. Who’s been here for a long time? Who’s new? What do they know? Where do they come from? What are their formal and informal qualifications? What are their hobbies? Do they use Gen AI? If they avoid it, why do they avoid it?
Use something like a simple survey to gather this information. Spend two to three months having conversations with staff and identifying their current levels of expertise in different domains and where they would like to take their professional expertise next. Some will tell you that they want to learn more about generative artificial intelligence. Some will tell you that they’re only interested in learning more about physics or music or visual arts. Respect what these people are telling you.
Stage Two: Design
Once you have identified where staff expertise lies, work with staff to map against the three dimensions model. Are they confident disciplinary experts with years of experience like our literature teacher? Are they an early career teacher, teaching out of field and feeling out of their depth and unsupported, or are they like our maths teacher, brimming with ideas, knowledgeable about new technologies and industry, but lacking lived experience in the classroom?
Mapping staff capabilities against the three dimensions framework will reveal clusters of staff and may suggest professional learning communities, communities of practice, action research groups and other structures. Over the next couple of months, these groups can be approached with different variations of professional development, and can begin supporting one another across dimensions. The aim is to move staff from the outer circles to the inner and to balance the three dimensions of the Venn diagram. Remember, you’re not trying to make everybody an expert in everything, though a handful of people at the centre of the Venn would be a great asset to any school.
Different roles within the school will tend towards different profiles. Discipline and technological experts would make great e-learning or digital coordinators, and even if they are relatively new to teaching, they would still be a great asset for supporting the professional development of other staff. Gen AI “resisters” in a school can add a respectful counterpoint to discussions about the need to integrate technologies, and as long as they are refusing based on knowledge and awareness of the technology and not just fear or general aversion, then these staff will enrich the professional development of the technology adopters.
Sometimes you will identify staff for whom the technological expertise dimension is nowhere near as important as building their domain and situated expertise, like our out-of-field early career teacher. For a teacher like this, you have a few options: try to get them working in field, within the limitations of the timetable; partner them up with mentors; ensure they have sufficient professional development opportunities through subject teacher associations and so on.
Stage Three: Lead
By this point, about halfway through the school year, some teachers should be emerging as leaders in their various dimensions. Some of these will be obvious, such as faculty leaders with situated and disciplinary expertise. Some may be outliers, perhaps the early career teacher demonstrates leadership in running professional development on the technologies.
Most staff will not be leaders or experts, but you need to put in place structures for those that are. If you’re lucky enough to have more than one expert across the dimensions, then they should be encouraged to work together. By this stage, you should also be building up some evidence. Create a shared space on the staff-facing part of your learning management system, or in a shared drive where teachers can access professional development resources, case studies, artefacts or examples of student work.
To sustain these practices, the initial audit needs to be carried out regularly, probably once a year.
A Case Study: Towards Expertise
I’ve worked with hundreds of schools since 2023 and with English teachers since 2016, allowing me to observe long-term staff development over nearly a decade.
“Mary” represents a composite of teachers I’ve worked with. She graduated in 2018 and joined a large city co-ed school but was frustrated by limited English teaching opportunities (only one Year 8 class) and the school’s resistance to subjects like VCE Literature. When we met at a 2019 conference, Mary was considering switching schools. I suggested she develop her English expertise, contribute to the teaching community through writing and professional development, and apply for the head of faculty position opening in 2020. She did exactly that and was successful.
During COVID’s extended remote learning period, Mary became a VCE examination assessor and supported staff transitioning to online teaching. Despite being only two years out of university, she rapidly developed domain expertise in VCE English requirements.
In 2022, before ChatGPT’s release, Mary contacted me about students using AI writing tools like Jasper and WriteSonic. When ChatGPT launched, she was less surprised than colleagues, having monitored these technologies.
Through Mary, I began working directly with her school in 2023, helping develop AI guidelines and running staff professional development. Mary led an AI Task Force with the assistant principal of curriculum and head of digital technologies. By 2025, the school had published clear AI guidelines and established discipline-specific Professional Learning Communities for maths, Visual Arts/Design Technologies, and English/humanities. These groups run professional development for colleagues and host parent information nights on AI.
Remember, Mary is a composite. She represents elements of staff that I’ve worked alongside for many years. I’ve now seen a number of schools support staff to reach these levels of expertise. When professional development is done well, with individual pathways designed for teachers and supported through subject associations and regular opportunities for quality PD, schools can absolutely build robust, meaningful approaches to new technologies like generative artificial intelligence.
Get Started Now
If you’re a school leader reading this article, I want to encourage you to start right away with the skills audit I discussed in Stage One. If you’re not a school leader, I would encourage you to do what Mary did, and think of ways that you can hone your expertise in one particular area, perhaps your subject domain, or perhaps the technology, to demonstrate to the school your leadership capabilities. Once you’re there, you can be the person that instigates the audit.
If you’ve gone through a process like this yourself, leave a comment, either here on the blog or wherever you found this post on social media.
Want to learn more about GenAI professional development and advisory services, or just have questions or comments? Get in touch:
Your message has been sent
Leave a Reply