Beyond the Prompt: Why I Created AI Elevated

Practical ideas, honest reflections and a classroom teacher’s perspective on what AI can, and cannot, add to education.

AI in education is often discussed at the extremes. It is presented either as the answer to almost every problem in teaching or as overhyped technology that creates more work than it saves.

My own experience sits somewhere in the middle.

I can see genuine possibilities in AI. I have also seen it produce inaccurate information, unsuitable resources and polished-looking material that simply would not work with a real class. That space between the promise and the reality is where this blog begins.

A little about me

I am Rosemary Matthews, a post-primary teacher in Northern Ireland with more than 20 years of classroom experience. I teach Biology, Chemistry, Home Economics and Nutrition & Food Science, from Key Stage 3 through to A level.

Teaching across different ages and subjects has made one thing very clear: there is no single approach to AI that will work everywhere. A tool that helps me analyse an A-level examiner report may be of little value during a practical Home Economics lesson. An activity that works with a motivated sixth-form class may be completely unsuitable for younger pupils who need more structure, explanation and teacher direction.

I am not a technology expert, software developer or computer scientist. My expertise is in teaching: understanding a curriculum, explaining difficult ideas, recognising misconceptions, creating useful resources and knowing the pupils in front of me.

I also have a creative mind, enjoy learning new things and like exploring whether there might be a better way to solve a familiar classroom problem. That combination of classroom experience, curiosity and creativity has shaped the way I use AI.

Part of my motivation is also a desire to keep developing professionally. Pupils are already encountering AI, whether or not it is formally introduced in school, and I want to understand it well enough to guide them with confidence. I do not expect to know every tool or have every answer. I do want to be able to question what AI produces, recognise its limitations and help pupils use it critically and responsibly.

For me, professional competence does not mean pretending to be an expert. It means being willing to keep learning as the world around the classroom changes.

How I use AI in my teaching

My use of AI did not begin with a grand plan. It grew gradually as I experimented with tools and platforms such as ChatGPT, Microsoft Copilot, NotebookLM and Wayground, formerly Quizizz.

I have used AI to help develop retrieval quizzes, knowledge organisers, comparison tables, flashcards, personalised scaffolds, revision resources and exam-style questions. I have explored ways of using it to analyse specifications, mark schemes and examiner reports, identify common misconceptions and gaps in learning, adapt material for different pupils and create more imaginative learning experiences, including case studies and clinical scenarios.

The name AI Elevated reflects what I have experienced when AI is used thoughtfully.

It has not transformed every lesson, and not every AI-generated idea ever reaches my pupils. At its best, however, it has helped me make difficult ideas clearer, provide more varied retrieval and revision activities, respond more precisely to misconceptions and create imaginative resources that I might not otherwise have had time to develop.

It has also encouraged me to revisit familiar topics and think more creatively about how I teach them.

For me, elevating the classroom does not mean replacing good teaching with technology. It means using technology selectively to extend and strengthen what a knowledgeable teacher can already do.

Some ideas have worked extremely well. Some have been useful only after considerable editing. Others have looked impressive on screen but added very little to learning. A few have simply not been worth the time or effort.

Those differences matter.

I do not want AI Elevated to become a collection of tools presented as though every one of them will transform teaching. I want to share what I actually tried, what happened when I used it and whether I would use it again.

Prompting is only the beginning

There is already plenty of advice online about writing better prompts. Prompting is useful, but in my experience it is only the beginning.

The more important work often happens after the response appears.

Is it accurate? Does it match the specification? Is the level of language appropriate? Does it reflect how the topic is assessed? Will it help pupils think, or has it done too much of the thinking for them? Most importantly, would it work with this particular class?

AI does not know my pupils. It does not understand the dynamics in the room, the explanation they heard yesterday or the misconception that keeps resurfacing. It can generate a starting point, but it takes teacher knowledge and judgement to turn that starting point into something genuinely useful.

Sometimes AI reduces workload considerably. At other times, the time saved generating a resource is partly lost through checking, correcting and adapting it. I will refer to this throughout the blog as repair and maintenance.

Occasionally, it would have been quicker to create the resource myself from the beginning.

That does not necessarily make the experiment a failure. AI can sometimes help me produce something more creative, better differentiated or more carefully structured than I might otherwise have had time to develop. Workload reduction matters, but speed is not the only measure of value.

The more useful question is whether the final result improves teaching or learning enough to justify the time spent on it.

The idea comes first

I rarely begin by asking, “What can this AI tool do?” I usually begin with a teaching problem.

Perhaps pupils repeatedly confuse two similar processes. Perhaps a class needs more retrieval practice, but I do not want to keep giving them the same style of quiz. Perhaps an A-level topic needs a clearer visual explanation, or an examiner report contains useful insights that are difficult for pupils to access. Sometimes I simply wonder whether an established activity could be made more engaging, more focused or easier to adapt.

The idea comes first. The tool comes second.

Many ideas also grow through conversations with colleagues. Teachers who are already experimenting with AI can spark new possibilities, but colleagues who are cautious about it are just as important.

A sceptical question such as “Will pupils actually learn from that?”, “How long did it take you to fix?” or “Would the original method have worked better?” can expose weaknesses that enthusiasm alone might miss.

I value those conversations, including conversations with people who disagree with me. Thoughtful dissent makes ideas stronger. It keeps the focus on learning rather than novelty and reminds us that choosing not to use AI can sometimes be as professionally sound as choosing to use it.

Knowing when to stop

Exploring new technology can easily lead down a rabbit hole. One more prompt, one more revision or one more new platform can turn a supposedly time-saving task into an evening’s work.

I am learning to stop when the output needs more repair than it is worth, when a familiar method would work better, when the technology creates unnecessary barriers or when the activity risks reducing the thinking pupils need to do for themselves.

I also stop when I find myself continuing simply because the technology should be able to do something, rather than because the result is genuinely useful.

Giving up on a particular tool or idea is not a failure. It is part of evaluating it properly. Teachers make these judgements every day about resources, explanations and activities. AI should be subject to the same professional scrutiny.

Why I am starting this blog

There is no shortage of information about AI in education, yet I have often struggled to find one place where teachers can see practical classroom uses explained in enough detail to try for themselves.

Much of the discussion focuses on lists of tools, collections of prompts or broad predictions about the future. Far less attention is given to what happens after the prompt: the checking, adapting, testing, repairing and sometimes abandoning.

AI Elevated is my attempt to share that part of the story.

Some colleagues who have seen the work I have been developing encouraged me to share my experience more widely. I was initially hesitant. I am more comfortable quietly exploring, creating and refining than putting myself or my work on display.

However, I also think there is value in classroom teachers sharing what they are learning while they are still learning it, honestly and practically, without pretending to have every answer.

Through this blog, I plan to share:

·       practical and creative ways I have used AI in teaching

·       honest reviews of tools and courses I have explored

·       examples of what worked, what needed significant adaptation and what did not work at all

·       reflections on workload, teacher judgement, pupil learning and classroom reality

·       ideas that teachers can adapt across different subjects, age groups and school contexts

·       the thinking behind the ideas, not simply the finished prompt or resource

These are not prescriptions. What works for me may not work in another classroom, and an idea developed for Biology or Nutrition & Food Science will rarely transfer unchanged to a different subject.

What I hope will be transferable is the thinking behind it: identify the problem, choose an appropriate tool, evaluate the result and adapt it using professional judgement.

What I hope AI Elevated becomes

Starting AI Elevated is also a way of pushing myself beyond my comfort zone. Every now and again, I think it is important to do something that feels slightly daunting, particularly when it offers an opportunity to learn, connect with others and contribute something useful.

I am not writing this blog because I believe every teacher should use AI, or because I think technology can replace the knowledge, relationships and judgement at the heart of teaching.

I am writing it because teachers deserve realistic examples, honest evaluation and space for thoughtful discussion.

If something I share gives another teacher a useful starting point, sparks an idea for a completely different subject, saves them from an unproductive rabbit hole or helps them make a more informed decision about AI, then the blog will have achieved something worthwhile.

AI Elevated will be a record of what I am trying, what I am learning and what I am still questioning. I hope it becomes a useful place for curious teachers, cautious teachers and sceptical teachers alike.

Because the most valuable conversations about AI in education will not come from technology alone. They will come from teachers sharing, questioning, testing and learning from one another.

If that sounds useful, subscribe below to follow the journey, see what is inside my AI toolkit and hear about the real-world ups and downs of using AI in the classroom.