Prompting Progress: How Artificial Intelligence is reshaping the classroom
This example is shared by a English teacher from Graeme High School in Falkirk Council and uses Gemini and NotebookLM
Use case: AI Supporting Advanced Poetry Pedagogy
First, I used Gemini to act as a literary expert. I prompted it to analyse six Carol Ann Duffy poems, focusing specifically on how her use of imagery and word choice creates meaning. I didn’t just accept the first draft; I acted as an editor, refining the text through several rounds of feedback to ensure the analysis was sharp and accurate for revision purposes.
Next, I worked with the AI to link the poems together. I directed it to group the texts into shared themes like “The Passing of Time” and “Isolation.” This was deliberate as being able to compare and connect different poems thematically is essential for learners. Once the content was ready, I had the AI rewrite the segmented notes into one cohesive narrative so the next tool (Notebook LM) could process the “story” of the six poems more effectively.
Finally, I moved that polished text into Notebook LM. By using my curated summary as the source, I had the AI generate a video summary.
NotebookLM turned my written notes into a visual and narrated guide, creating a high-quality revision resource that makes complex connections much easier to remember.
Crafting prompts
First Stage Gemini Prompt: Create a text based summary of 6 Carol Ann Duffy Poems: In Mrs Tilscher’s Class, Originally, Mrs Midas, Medusa, Havisham and Before You Were Mine. I will upload it into Notebook LM to create a video summary of the poems. Add commentary/analysis on each one that focuses on key techniques such as imagery and word choice. Discuss similarities between the poems thematically. For example Havisham and Medusa are both consumed with rage, Before You Were Mine and In Mrs Tilscher’s Class examine the passing of time and change that comes with that.
Was it a ‘one-and-done’ success, or did you have to tweak your instructions to get what you needed?
It definitely wasn’t a ‘one-and-done’ success. The process of creating the text resource was much more about a back-and-forth conversation, where I had to act as a hands-on editor to make sure the content was right. While the AI is a great starting point, it initially made some mistakes with the fine details. I added quotations to ensure variety for example and sharpened some of the analysis.
How the tools improved the task I was doing
Notebook LM took a page of text and turned it into a narrated video which is a far more engaging revision resource for students. It makes it more likely they will remember key images, language and ideas when presented in a visually stimulating form.
My advice for others who haven’t used these tools before
Check the text-based resource carefully, it can be deceptive. Outright errors need to be corrected and other content required tweaking. Some analysis was not wrong but just markedly different to how we had read a poem in class. I wanted the video to be familiar and to reinforce prior learning – not confuse or contradict.
Next steps
Examples
Carol Ann Duffy: Revision Guide in the Scottish context of the National 5 / Higher English exam.
