Imagine Yourself: Explorations in Fostering Personal Expression with Generative AI

Abstract

Generative Artificial Intelligence (AI) technology has been promoted with many exciting promises to enhance human creativity. However, it has also been shown to amplify human bias and perpetuate harmful stereotypes. In the new age being ushered in by this technology, this thesis explores how educators and designers can use this technology to support young people in exploring and expressing aspects of their unique identities. In particular, I use a design based research methodology to iteratively create Imagine Yourself, a new digital experience adapting off-the-shelf text-to-image generation technology to support young people creating personal representations and stories. Imagine Yourself combines OpenAI’s Dall-E 3 image generation technology with Scratch, a rich environment for young people to imagine and create interactive multimedia stories, animations, and more. Guided by a core value of designing for belonging, this project explores how experiences with generative AI can be designed to foster young people’s creative process in creating personally meaningful stories reflecting their own unique identities, experiences, and cultures. I discuss the iterative design process of creating Imagine Yourself in tandem with creative workshops, aiming to support more diverse representation within the image generation output and invite a tinkerable and iterative process of creating. I discuss observations and feedback from creative workshops with young people and adults, creating with Imagine Yourself. Finally, I conclude with reflections on the design process as well as a discussion of challenges, limitations, opportunities, and open questions for future work incorporating generative AI into young people’s creative learning experiences.