Using Google AI for Feedback

[DoctorHarves]

Over the course of a recent Level 1 NCEA Science assessment, I found myself reflecting on how much stronger the learning process can be when digital tools, AI, and teacher judgement work together rather than separately.

As I use benchmarks for my assessments, my students were uploading their assessments to Google Classroom every lesson. That gave me an ongoing window into their progress at each benchmark stage of the task, from brainstorming ideas to choosing a question and selecting investigative techniques.

To support the process, I created a Gemini Gem marking assistant linked to the standard that provided formative feedback for each student at each benchmark. I still reviewed every response before it was returned, because the professional relationship, the context of the student, and the judgment around what feedback is most useful all still matter deeply. But what AI helped me do was respond more consistently, more efficiently, and with a clearer overview of the class as a whole.

What I valued most was that this gave me a much better sense of where students were before the final submission. Instead of only reacting at the end, I could mentor more effectively during lessons. I saw patterns emerging, allowing me to step in earlier, ask sharper questions, and support students in ways that were much more targeted.

Once the assessment was complete, I used the feedback from the class and asked Google Gemini to identify the wider trends and themes. From there, I uploaded that response into NotebookLM to create a presentation for the class.

The goal was not just to look back at what had happened, but to help students clearly see where improvement was still possible and where reassessment opportunities might sit.

That part felt especially powerful. Rather than reassessment being framed as simply “another go”, it became a more informed conversation about growth, gaps, and what quality improvement actually looks like.

This is one of the most useful applications of AI in education. Not replacing teacher expertise. Not automating care. But helping make feedback more visible, patterns more obvious, and next steps more actionable.

Google Classroom gave me the workflow. Gemini helped me work with feedback at scale. NotebookLM helped turn those insights into something meaningful that students could action.

In the middle of it all, the teacher’s role became even more important. Noticing, mentoring, questioning, and helping students move forward with confidence.

AI did not replace the feedback process. It helped me deepen it.

[jill_dubois · Accepted solution]

Loved reading this - I am doing the same with Gems and find that I am constantly refreshing my instructions as my perspective changes. I appreciate this from what you wrote..."I used the feedback from the class and asked Google Gemini to identify the wider trends and themes. From there, I uploaded that response into NotebookLM to create a presentation for the class."

You've just sparked some real joy here for me to seek some new options and continue with NLM!

Best,

Jill

[DoctorHarves]

Thanks for the feedback, Jill.

[ghoshsanchita]

Thankyou for sharing your use-case! 

[Allison75]

This is such a great way to lift the level of teaching and learning while consolidating teacher workload. Excellent post! Thank you for sharing.

Best,

Allison