With so many AI tools, platforms, and big promises flying around, it’s easy to feel like everyone else is miles ahead. They aren’t. The real work begins with a few grounded moves, a little curiosity and a healthy dose of professional judgment. Here’s a practical place to start. And if the order doesn’t suit you, skip ahead to wherever you are in your own journey with AI.
Play First, Play Privately
The best way to learn AI isn’t by reading about it; it’s by trying things where the stakes are low and the outcomes don’t really matter. Generate recipes from a photo of your fridge, ask it to create a music trivia based on your oddly specific niche or have it design a walking tour of unusual local sites. Just avoid sensitive data and steer clear of finance, health, or anything that could get strange, fast. This stage is about curiosity, not correctness.
Find Your Focus
Once you’ve built some intuition, return to the one thing that already anchors learning in your school; the mission, vision values or learner profile. Choose the touchstone that comes up most often in conversations about teaching and learning. These anchors help you decide why you are using AI, not just how. If your school has guidelines, check those first. What matters most at this stage is clarity of intent.
Know the Problems
AI gets things wrong, sometimes with executive level confidence. It can reproduce biases, fumble nuance, and stumble on cultural or linguistic cues. Your local data protection requirements matter (e.g. GDPR, PDPA, etc), as does your school’s policy. If you are unsure about either, ask your tech team early. It is far easier to prevent issues than unwind them later.
Start With the Obvious Stuff
There are plenty of soft targets that make early wins easy. You don’t need a dozen platforms, just choose one or two school endorsed tools and deepen your skill with those. Some teachers see bigger gains from refining prompts than from collecting apps. Focus on tasks tied directly to your work; drafting rubrics, aligning standards, levelling texts, or prototyping curriculum documents. You are not removing challenge; you are reducing friction so you can focus on learning, improvement, and hopefully clawing back some time.
Look for High-Value Moments in Your Pedagogy
Once you feel comfortable, begin stretching your use cases. Where might AI genuinely improve learning or remove friction? Common sweet spots include multilingual scaffolds, formative feedback, inquiry planning, differentiation and resource creation. In a PYP context, you might ask the model for three provocations linked to a key concept such as perspective or responsibility, and choose the one most likely to spark learner interest. Not every moment needs AI, but a few well chosen ones can make a real difference.
Use AI as a Thought Partner
When your confidence grows, explore reasoning-based prompts that support professional judgment. Contrastive explanations, self-consistency checks, and role-play for alternate perspectives can all strengthen your work. You might ask AI to outline a group project using a structured reasoning approach, or to provide contrasting interpretations of a primary source. These moves amplify your thinking rather than replace it and they sit comfortably alongside practices like UDL and inquiry design.
Define Your No-Go Zones
Before shifting to student use, draw your line in the sand. What things stay human? summative judgments, wellbeing conversations, sensitive student information, behavioural notes, and anything requiring professional discretion. Decide what you will not use AI for and why. These boundaries may shift later as safeguards improve, but being deliberate about change matters as much as the boundaries themselves.
Shift (Thoughtfully) to Student Use
When you are ready, consider how students might use AI safely and purposefully. Communicate with your team or leadership about your intent, and be clear with families before you begin. Purpose, process, and guardrails matter far more than the tools. Start small, such as co-creating success criteria with AI or generating multiple perspectives for a class discussion. Student use will vary widely across year levels and curricula, but it should never come as a surprise.
Reflect
Choose any lens that helps you think about alignment, impact, and effort. Whether you use SAMR, TPACK, UDL, ATLs, or your own homegrown acronym, the key question is whether AI is supporting learning in a meaningful way. Build a simple reflection routine after each experiment: What worked? What didn’t? What might you try again?
Share Your Findings
Talk openly about what you are learning: the good, the bad, and the strange. De-privatising practice is one of the strongest safeguards schools have and an excellent way to scale what works. Start with a trusted colleague if that feels easier. Small internal conversations build collective wisdom and reduce risk.
Or Just Take a Shortcut!
If you prefer a structured, inquiry-led pathway with exemplars across K12, curated tools, and a community to collaborate with, the Pedagogy-First AI Accelerator begins March 9. Across four weeks, you will play, plan, build, critique, and leave with a polished AI Learning Exemplar ready for your classroom. It is a practical and humane way to move from curiosity to confident action. PS: Dr. Dana Watts will be co-facilitating it too, so that’s a win already.
And once you have a foundation, you can begin imagining what comes next: personal learning guides, adaptive challenge pathways, creative agents that deepen inquiry, and new feedback loops that support learner agency. These possibilities are real, but they deserve thoughtful, steady steps. Start small, stay grounded, and let your practice grow with clarity rather than urgency.
You can connect with John Burns on LinkedIn. Register for the Pedagogy-First AI Accelerator here.
