Creative fidelity: AI in design education

AI tools are developing rapidly, and schools are trying to keep up with the developments and their implications for how we learn, how we create and how we keep young people safe. AI is very much a double-edged sword. You can’t really talk about AI in education, and specifically in design education, without getting into some quite philosophical questions. What does it mean to learn? What does it mean to be creative or to have original ideas? And how, and what, do we assess with this new technology?

As many of you know, I’ve been working in AI and education for a few years now, exploring what it means to use ever-emerging tools safely and effectively, with pedagogy and the learner at the centre. AI tools can save teachers time and enable us to dual-code information with the needs of the learner in mind. Creating audio versions of written text, making visual infographics, and adapting the reading age or language of a text are just a few of the many ways schools and colleges are using AI right now in design and technology and in art and design. Teachers and students are also starting to experiment with how AI can enhance communication, confidence and experimentation. This is happening not only through text-based large language models, but also through the built-in AI functionality of CAD tools, visual design tools and even 3D tools.

I was recently lucky enough to judge the Create, Code, Change competition, where young learners from around the world used vibe coding techniques and tools to solve real-world problems. The work the students produced was incredible. Their creativity, collaboration and deep, empathetic understanding of specific problems were genuinely impressive, especially considering that most of the learners involved were under 14.

The digital divide in our schools is still a huge chasm. Now, with AI literacy and the need to explicitly teach critical thinking about AI outputs, there’s a real risk of that divide becoming even bigger. I was one of the authors of the DfE’s Safe and Effective Use of AI support materials for schools and colleges. These materials are freely available and go into much more detail about what schools need to consider before using AI. Important areas such as GDPR, intellectual property, safeguarding, ethics and bias are all covered. We also include sections on sustainability and climate change when considering uses of AI.

My own passion for design education has led me to experiment with the capabilities of AI design tools, and to wrestle with questions such as: Where does my creativity end and my AI tool’s creativity begin? How do I retain authorship of my work? And how can I push my own creative practice further?

Through this work, I introduced a concept called creative fidelity, and I recently published a paper on the concept and the research behind it in the Chartered College’s Impact journal. In short, creative fidelity is about being mindful of when we’re using AI tools to do the creating for us, and when we’re using them to support communication, experimentation or delivery while still retaining ownership of the core idea. The research really boiled down to the need for explicit terminology, as someone said earlier, to describe these different relationships with AI. Hence the term creative fidelity.

For example, if we’re using AI to render a sketch accurately in a variety of colours or textures, it’s essential that the tool stays loyal to the outline or sketch provided. If a student designer has designed, say, a three-legged chair, we would not want to use certain large language models to render the chair in a given material, because they would “fix” the problem. Having been trained on countless chairs, the model could predict that the chair should have a fourth leg and correct the issue. Other tools would retain the designer’s intent and stay loyal to the outline of the three-legged chair.

It also covers the essential need to think about tool selection in a similar way to how we select tools in a workshop. AI tools do slightly different things. We need to choose the right tool for the job, and we need to prompt it accordingly. We all know that when we choose a hammer for a job, we’re not going to ask it to drill a hole, because that needs a different tool. I think this is a really useful metaphor for AI tools, because some require more training than others, some carry greater risks than others, and some may even need forms of protection. For example, you might need to change your permissions, just as in a workshop you might need PPE. In a workshop, we would never hand someone a table saw without training them first, and we would expect them to use the appropriate personal protective equipment. We should think about AI tools in a very similar way. 

AI is all around us. A lot of you will have driven a car with AI in it this morning, and your iPhone has AI in it. It’s not something we can opt in or out of. It’s everywhere, and it carries huge promise and potential for pupils and students. I believe we all have a responsibility to teach it, and when we teach it, to teach it safely, effectively and morally.

I’d love to chat to any of you about it later on. Thank you very much for listening.