Engaging with AI painting: exploring motivations and challenges in laypeople's creative information practices

Abstract

Introduction. The advent of generative AI has democratized art creation, enabling individuals without formal training to produce visually appealing digital artworks. However, it is not yet well-understood why and how laypeople engage with AI in creative information practice.

Method. This study conducted semi-structured interviews with 17 participants, aiming to understand the motivations for laypeople's engagement with AI painting and the challenges they encounter.

Analysis. The interviews were analysed using open coding and thematic analysis, with two independent coders achieving a substantial inter-coder reliability score.

Results. Our findings reveal that laypeople's engagement with AI painting is a practice-oriented information practice, influenced by social and contextual factors. The engagement process is iterative, starting with a user prompt and AI-generated initial output, followed by continuous refinement. Laypeople engage with AI painting for emotional motivations, personal utilitarian motivations, and social interaction. Notable challenges include image quality, technological limitations, and personal constraints.

Conclusion. These findings provide empirical evidence of the potential and limitations of AI in creative information practice. This understanding is vital for informing the design of future AI tools to enhance the human-AI collaborative experience in creative endeavours.