Reflections on what the four FLANZ keynotes taught me about the scale of change, and why it matters for educators right now
Having had more time to reflect on the keynotes and workshops I attended at the recent FLANZ conference one theme ran through almost everything I heard. Each was saying the same thing: change happens at different levels, and a great deal of the frustration educators feel comes from mistaking one scale for another.
I want to use that thread as a way into something that’s weighing on a lot of us in Aotearoa at the moment: CHANGE! Curriculum, pedagogy, assessment, and the arrival of AI – and happening all at once. It’s an enormously complex picture – and something we need to grapple with amid all of the ‘noise’ we’re hearing in the lead up to this year’s election. My argument is that it becomes a little more navigable, and a lot more hopeful, once we are clear about where responsibility actually lies.
Three perspectives on change
The idea is simple. The micro level is the individual: the teacher, the learner, the classroom. It’s about habits, attitudes, relationships and the daily choices people make. The meso level is the group: the school, the department, the team, the professional association. It’s about culture, leadership, communication and how a community makes sense of what’s being asked of it. The macro level is the system: legislation, funding, national curriculum, qualifications, and the larger economic and technological forces that act on all of us.
Mark Nicholls made these three levels explicit in his keynote on Blendagogy, crediting Olaf Zawacki-Richter for bringing them into the distance education conversation. For Mark, the macro level is where agencies like TEC, MoE and NZQA set policy, funding and quality assurance. The meso level is the school or department, with its programme norms, timetabling conventions, workload expectations and supported technologies. The micro level is where teaching actually happens, in the day-to-day choices of the kaiako with their ākonga.
His observation resonates for me. The majority of kaiako design at the micro level, yet the meso and macro layers shape what’s possible there. Ignore them, and you can spend years working in good faith against the very institution you’re part of. But look at them closely, he said, and you may find avenues to extend your practice too.
Here’s an example that will be familiar to many of us right now. At the macro level, there is a mandate for explicit teaching of structured literacy, a response to system-level literacy problems that have been identified. It’s a clear national direction. Now picture a kaiako whose classroom runs on a broader, interests-based approach where learners are reading and writing about things they care about, with high engagement. The teacher’s conviction is that this is what makes children want to read and write at all. From where they sit, the mandate can feel like an instruction to abandon what works. Left on their own, they may comply reluctantly or quietly resist, and either way they carry the tension alone.
This is where the meso level earns its place. A school that takes responsibility for the response, rather than simply passing the directive down, can design a balanced approach, ensuring explicit, structured teaching is used to address the identified literacy needs, alongside the interests-based approaches that keep engagement high. The teacher’s commitment is then an asset the school builds on, and the macro mandate becomes one half of a design problem the school owns, rather than something done to its teachers.
Three levels, three different jobs. The macro level names the problem and sets the direction. The meso level decides how to honour both that direction and what it knows works for its learners. The micro level brings the craft and relationships that make it real for each child. Remove the meso level and the teacher is left to fight the system or give in to it.
What the keynotes demonstrated
Adrian showed us a system that couldn’t see what it already knew. As a 19-year-old he was already trusted to work as a sous chef and had won a national competition, yet the qualification still required him to reproduce his knowledge in its kitchen, on its menu, at its chosen time. That is a macro and meso problem: how qualifications and institutions define evidence of learning. But the second half of his story is the one I find hopeful. It began at the micro level, with a couple of colleagues trialing the use of iPods to replay demonstrations for students who couldn’t make it to class. That small crack led to questions about where learning lives, then to project-based learning, then to a degree, and eventually, after COVID, to a formal pathway recognising work-integrated learning. Change that started in one kitchen ended up reshaping a programme.
Kathryn showed us what happens when the levels fall out of step. Institutions are rushing out AI guidance, and much of it tells students to be “responsible” and “transparent” without saying what that looks like. Worse, she pointed out, we can’t ask learners to be transparent if our institutional culture makes honesty risky for them. And behind all of that sit the tools themselves, built and controlled by technology providers none of us have any say over. Yet she also gave us micro-level examples: the teacher who told her chatbot to “do better” as if the prompt weren’t her own responsibility, and the work of co-designing guidance with students so that every guideline comes with an explanation and a concrete practice. Her central point was that humans shouldn’t be a safety check sitting “in the loop”. We should be the ones leading it.
Selena was candid about the macro level. Competency-based qualifications have limits that educators have criticised for decades, but turning that ship around is close to impossible, so we work with it. What we can do is design learning that protects the productive struggle that builds capability, and use AI as a coach rather than an answer supplier. She also gave us a meso-level warning sign: in her small survey of vocational educators, around one in five were either refusing AI outright or very reluctant to use it. If those colleagues aren’t supported to build their own literacy, they can’t model good practice for their learners.
Mark offered perhaps the most practical advice of the week. If you’re an individual kaiako, start knocking on the door of your head of department, because you need meso-level support for what you’re doing. Otherwise you risk building a Franken-course: rich and flexible to you, but disjointed and confusing for the learner who meets a different model in every classroom.
Where this leaves us in Aotearoa
Right now we are facing change on all three levels at once, and each of the big shifts looks different when you map it.
The national curriculum is being substantially overhauled, moving from a competency-based approach to a knowledge-rich one, with much more prescription about content and less emphasis on developing capability. NCEA is changing too, with more traditional end-of-year assessments that sit in tension with what schools have been used to. And with AI, schools have largely been left to develop their own adoption policies, with very limited guidance and support from the government and the Ministry. Many other countries are putting national guardrails in place. New Zealand has taken the position that the benefits outweigh the dangers and is encouraging greater uptake. There is work going on behind the scenes, but much of it isn’t explicit or well understood by schools and their communities.
The table below illustrates how these issues are addressed at the three levels of the system:
| Issue | Macro (national) | Meso (the school) | Micro (the classroom) |
| Curriculum | Shifts from competency to knowledge-rich, with more prescribed content | Interprets and resources the change, and designs how to balance capability and knowledge. | Teaches the content well while keeping learners’ capability and agency alive |
| NCEA and assessment | Introduces more traditional end-of-year assessment | Manages the tension between new requirements and existing practice, and explains it to whānau | Prepares students and designs tasks where the learner’s own judgement is visible |
| AI | Encourages uptake with limited guidance; some work not yet visible to schools | Develops its own adoption policy and explains it to the community | Models good practice and teaches students to use AI critically |
Notice the difference in pattern. With the curriculum and NCEA, decisions arrive from the macro level and the meso and micro levels are asked to implement them. With AI, the macro level has stepped back, so the responsibility doesn’t disappear. It moves down. Boards and school leaders at the meso level are being asked to make decisions about safeguards that other countries make nationally, as that is a responsibility that shouldn’t be left to individual teachers to carry alone.
The reason this matters is that locus of responsibility is easy to get wrong, and getting it wrong is costly. When a teacher feels personally to blame for something that was decided at the macro level, or when a school waits for national direction on something it could be deciding for itself, energy goes to the wrong place. Naming which level a problem belongs to is not a way of passing the buck. It’s how we work out who needs to act, and what each of us can reasonably be expected to carry in a complementary manner.
It also lets us say something honest about the macro level: much of it is not in our hands. Mark was blunt that nobody should be waiting for a funding announcement to make the education they believe in possible. Selena was blunt about the limitations of a competency-based system. We can and should advocate at that level, but we shouldn’t mistake advocacy for the only thing we can do.
What you can act on
Here is where I find real hope. Each of these keynotes pointed to things within reach of an individual educator, whatever is happening above them.
- Look for the human, not the AI. Rather than trying to police where AI was used, ask where the learner’s judgment, effort and authorship are visible in the task, and design for that. Be precise with students about what “appropriate use” means for this task, rather than leaving them to guess.
- Protect the struggle. Design tasks so that AI supports the effort rather than replacing it. Use it as a coach that asks questions rather than a machine that supplies answers, and treat mistakes as part of learning.
- Bring students into the conversation. Co-design your classroom expectations for AI with them, and turn vague guidelines into worked examples. They are asking for exactly this.
- Start small, and let it travel. Adrian’s iPods were a minor experiment that solved a problem for a handful of students. It wasn’t a strategy. Small, well-grounded changes can be noticed and adopted, and sometimes become the pathway.
- Find your meso allies. Talk to your head of department, your principal, your colleagues. A change that is aligned with the group around you is far more likely to last, and far less likely to leave you exhausted.
None of this makes the system issues go away. But it does mean that nobody has to wait for the system to be fixed before doing something that matters for the learners in front of them.
A question to take away
If you map the biggest change you’re facing right now across the three levels, which part of it is genuinely yours to act on? And who, at the other levels, needs to hear from you?
