Contextual AI Use in the Classroom

The National Council for Curriculum and Assessment has recently published enactment guidance for each of the five new primary specifications as part of their Toolkit offerings, as well as updated Assessment guidance in light of the redeveloped primary curriculum. It is important to note that specific guidance and probing questions are provided around the use of AI for teachers in preparing for teaching, learning and assessment across these five specifications.

The guidance asks us to consider the following:

  • Does the AI tool adhere to relevant school policies and guidelines?
  • Will it meaningfully support teachers’ preparation?
  • Can the outputs be guided to ensure they support, rather than replace, teacher professional judgement?
  • Are the outputs contextually appropriate and relevant for your school setting?
  • Before use, has the output been checked against other sources?

I am going to focus on the contextual relevance of AI outputs. Often, the response to a prompt seeking information around lesson ideas can provide us with generic ideas – while these can be engaging learning experience, they can be of little relevance to the lived experiences of the children in our school. For instance, I was provided with some great ideas around using the beach for an environmental awareness lesson, with our school being located at least one hour away from the nearest beach.

Gemini Notebooks

As a Google school, it can make sense to use Gemini as our ‘go to’ AI tool. Integration with ‘Notebook LM” has meant users can creates their own ‘Notebooks’ within the web browser. these notebooks can be used to ‘train’ Gemini around the context of your role.

In one Notebook, a school leader may upload Circular updates from the DEY, Information Notes, School Policies, etc. and ask the AI tool to always draw upon these documents when a prompt is included here.

In another Notebook, a classroom teacher might upload curriculum documents, assessment guidance, preparation for teaching and learning documentation, as well as other relevant Toolkit supports. You may also include information around your class in terms of age, interest, ability levels, etc. Similarly, specific websites can be provided to pull information from. All prompts within this Notebook will pull on this to ensure outputs are given the opportunity to be more contextually relevant to your particular setting.

Ultimately, teacher judgement still needs to be at the fore, but it is possible to make our AI use more efficient and more contextually relevant, with Gemini Notebooks being one tool that makes this possible.

Shopping Basket
Scroll to Top