When AI becomes invisible in your pedagogy, that’s when the magic happens.
The real AI level-up for educators doesn’t come from crafting the perfect prompt or even running complex single tasks through an LLM. The breakthrough happens when AI enhancements become so woven into your regular pedagogy that they flow seamlessly through your teaching practice.
This week, I experienced exactly this kind of integration whilst teaching Stuart Hall’s representation theory to my A-level Media Studies students. What unfolded wasn’t revolutionary – it was evolution. Standard pedagogy, enhanced at every stage by thoughtfully chosen AI tools.
The Lesson in Practice
We started with foundational discussion work, breaking down the word “representation” itself. What does it mean to represent? How does “re-presentation” differ from simple presentation? Then came the practical grounding – because I lean towards constructivist approaches – getting students to create six photographs of our school: three showing it in a positive light, three highlighting less flattering aspects.
This wasn’t just busy work. Finding shots of the library looking fantastic (our building is genuinely photogenic) versus capturing that trolley of broken furniture under the stairs taught them something fundamental about how representation is always a choice.
For the theoretical foundation, I used reading material I’d previously run through Diffit to create differentiated worksheets with knowledge check questions. The language level came down just enough to make dense academic theory accessible to my sixth form students without dumbing it down.
After watching a Media Insider video on representation theory and taking Cornell notes, we analysed the music video we’re studying – Sam Fender’s “Seventeen Going Under” – identifying representations within the text.
Where Technology Enhanced Learning
For homework, students received a video summary I’d generated through NotebookLM. I’d fed the same reading material into the platform and asked for a video summary. The result was genuinely impressive – well-structured visuals that perfectly complemented the academic content. This wasn’t replacing the primary learning but reinforcing it through a different modality: six to eight minutes of revision that built on what they’d already read.
The following lesson began with what I called an “exit ticket at the start” – a Brisk Boost chatbot programmed to check students had hit our five learning objectives. Now, I need to be very clear about the safeguarding context here. This ran within our Censo monitoring environment with live oversight for inappropriate interactions or disclosures. My post-16 students had permission to use this specific chatbot within school systems, with full data protection protocols in place. Brisk doesn’t require student logins – just a name – and each session remains isolated with no residual data.
What fascinated me was watching the adaptive questioning. One student struggled with the initial multiple choice question. The chatbot guided him through the answer, reduced the choice set, celebrated when he got it right, then adapted the next question accordingly. When he struggled again, it coached him through until he could tackle the long-form answers and demonstrate he’d met those learning objectives.
We continued with practical camera work – experimenting with orbital shots like those used in the Fender video, playing with dolly shots and push shots rather than zooms. Theory and practice interleaved throughout.
The sequence concluded with a multiple choice quiz generated by Brisk from our original representation theory video, which I then uploaded into Wayground for tracking student progress. I’ll reuse this quiz, adding questions about reception theory as we progress through the unit.
Why This Approach Works
Look at what actually happened here. Every pedagogical principle I value was present: interleaved learning, constructivist approaches, flipped learning, retrieval practice. Nothing was out of the ordinary – everything was enhanced.
I remained in control throughout. The content stayed on curriculum. Instead of defaulting to “read this, hands up for questions, write a quiz by hand,” technology supported learning at every stage whilst following safeguarding guidance.
Creating this sequence from scratch, unsupported by technology, would have required hours of planning. With AI enhancement, the new elements took minutes to implement, partly because I’d done similar work before, but mainly because the tools handled the heavy lifting.
GenAI isn’t changing or supplanting fundamental pedagogy. It’s enhancing what works at every stage. Yes, there were multiple short instances of AI use, but each served specific pedagogical purposes through carefully chosen platforms.
Although my instinct is often to reach for raw LLMs like Claude, I deliberately used education-specific platforms designed for classroom contexts. This choice reflects growing comfort with knowing when to use which tools – and when not to use them at all.
Working in a computer suite made digital implementation straightforward, but the principle scales. Teachers might choose YouTube videos or paper-based questions or whiteboards. Having that choice makes teaching more rigorous and adaptive. I can see what students grasp and struggle with, then adjust the next unit accordingly.
The Scaling Challenge
This is where we need educators to be operating – at that comfort level of choosing whether and how to integrate AI tools based on pedagogical purpose rather than novelty or fear.
The problem? I don’t think we’re anywhere near that point collectively. I’ve reached this integration through three years of practice, learning from professional networks, and continuous experimentation. I remain in the minority.
With less than 50% of schools providing AI training, and most of that being one-off sessions rather than sustained development, we face a significant gap between where individual pioneers are operating and where the profession needs to be.
The government has made the strategic direction clear. The question becomes: when will schools bridge this gap? What do we need to make this level of integration standard practice rather than exceptional?
The Path Forward
The answer isn’t more complex technology or revolutionary teaching methods. It’s supporting educators to reach that comfort level where AI enhancement becomes invisible – woven so seamlessly into good pedagogy that it simply amplifies what already works.
We need sustained professional development, peer learning networks, and time to experiment safely. Most importantly, we need recognition that this isn’t about transformation – it’s about enhancement of teaching practice that has always been effective.
The real AI level-up happens when the technology gets out of the way and simply makes great teaching even better.
What’s your experience with AI integration? What are your favourite combo moves?
#AIinEducation #EdTech #TeachingInnovation #MediaStudies #PedagogyFirst #AIintegration #EducationalTechnology #TeacherDevelopment #FutureOfLearning #ClassroomTech