Next generation of AI education

AI school classroom


Creating a hands-on and active learning approach to teaching 7 and 8 year-olds how to use, understand, and have a critical view of AI-powered image generation systems.


In this project, we worked closely with two year 4 teachers at Ryelands Primary and Nursery School in Lancaster. In 2022 and 2023 generative AI systems – which can create images and text automatically – took the world by storm and these technologies are here to stay. We wanted to research how to explain to primary school children how to use these AI systems, as well as help them come to their own conclusions about the ethical challenges associated with the technology.

In collaboration with the teachers, we designed a series of six lessons, each with its own lesson plan, set of worksheets and a custom-made generative AI platform that the students used directly to generate images. During the middle of the project, we decided to work with the students to create a new vision for their school using AI-generated images which we then turned into an imaginary school prospectus. At the end of the project, we invited parents to the school to view an exhibition about the project. Each family got a printed copy of the prospectus to take home.

In terms of research, the project was a great success. The hands-on approach to learning worked fantastically. We continue to work with the school and are looking to turn the project into a learning package that can be shared nationally.

Results and Outcomes

Tab Content: For Partners and Engagement

Because they are so busy, without input from this University-based project, it is unlikely that this school (and many schools like it) would be able to invest the time into thinking about how to teach their students about rapidly evolving technologies like AI. By creating resources and a platform that the school could use, we helped to demonstrate that the subject itself is nothing to be afraid of, so long as the right resources are in place. While we are not doing any kind of comparative study at this stage of the research, anyone involved in the project could immediately see it was a success. The students enjoyed every lesson, and simply by looking at the quality of the images they were generating, it was plain to see their skills improve through the weeks. An unexpected side-effect was an improved vocabulary for a group of young people who had missed out on crucial schooling during the Covid-19 pandemic, showing that working with AI can have a positive impact on other areas too.

Initially, we contacted the head teacher, and after an initial conversation, she allowed us to liaise directly with the two year-4 teachers whose classes we would be working with. They were incredibly supportive of the project and critical to helping us (after all, we are researchers, not teachers) understand how to approach working with the students.

We hope that this pilot will be the basis for something bigger. If we could roll out a learning package based on this work nationally, it could easily reach 20,000 schools, 250,000 teachers, and in just a few years, 10s of millions of young people.

Tab Content: For Academics

This project hinged on having a good working relationship with the school. We had the project concept but didn’t have a school to partner with. We utilised the network of colleagues in our department to find somebody who had previously worked with schools. In this case, it was this kind of personal introduction that led us to partner with Ryelands specifically. Building and nurturing partnerships takes a lot of time, but it can pay dividends and as time goes on allows you to speed up the process of collaborating.

Working with experts in any domain is a privilege. In our case, working with experienced teachers allowed the project to be a great success. When forging new ground, as is often the case with cutting-edge research, it’s imperative to accept uncertainty. Doing this allows great new ideas and perspectives to emerge; however, it also imports a significant amount of risk into any project. Building a trusting relationship with your expert partners will pay huge dividends. They must trust that you will do your best to deliver, and you must trust that they will understand that some imperfections are inevitable when trying out new ideas.

In our case, we combined multiple methodological and disciplinary perspectives to allow this to happen. Our entire project was organised around the fundamental principles of Design Research—this should not be surprising as it was funded by a broader initiative that seeks to gather evidence about the value of Design Research (see Those Design Research principles include embracing emergence and unpredictability, allowing plans to evolve until the last moment, and learning through the acts of designing, making, and trying things out in practice.

Back to listing