Visual artists often use critique to improve their work, but traditional feedback methods are limited by scheduling constraints, emotional difficulties, and one-size-fits-all approaches. While Artificial Intelligence (AI) systems like Multimodal Large Language Models could provide tailored and convenient feedback, existing conversational interfaces have not been optimized for this purpose. We conducted a co-design study to understand artists’ requirements and preferences for an AI feedback tool. Participants valued the objectivity and availability of an AI tool, but wanted non-intrusive guidance that preserved their sense of ownership. Such a tool must adapt to project life-cycles, maintain contextual memory, and support free-form input with spatially grounded multimodal output. Differences in AI persona preferences necessitate configurable, but “sticky”, controls. From these findings, we developed a prototype, which our participants assessed, leading to design insights and interface refinements. Our work helps inform the design process of future AI-based creativity support tools.

In the creative practice of text-to-image (TTI) generation, images are synthesized from textual prompts. By design, TTI models always yield an output, even if the prompt contains unknown terms. In this case, the model may generate default images: images that closely resemble each other across many unrelated prompts. Studying default images is valuable for designing better solutions for prompt engineering and TTI generation. We present the first investigation into default images on Midjourney. We describe an initial study in which we manually created input prompts triggering default images, and several ablation studies. Building on these, we conduct a computational analysis of over 750,000 images, revealing consistent default images across unrelated prompts. We also conduct an online user study investigating how default images may affect user satisfaction. Our work lays the foundation for understanding default images in TTI generation, highlighting their practical relevance as well as challenges and future research directions.

Image generation using generative artificial intelligence has become a popular activity. However, text-to-image generation—where images are produced from typed prompts—can be less engaging in public settings since the act of typing tends to limit interactive audience participation, thereby reducing its suitability for designing dynamic public installations. In this article, we explore body prompting as input modality for image generation in the context of public event settings. Body prompting extends interaction with generative AI beyond textual inputs to reconnect the creative act of image generation with the physical act of creating artworks. We implement this concept in an interactive art installation, Artworks Reimagined , designed to transform existing artworks via body prompting. We deployed the installation at an event with hundreds of visitors in a public and private setting. Our semi-structured interviews with a sample of visitors ( N = 79) show that body prompting was well-received and provides an engaging and fun experience. We present insights into participants’ experience of body prompting and AI co-creation and identify three distinct strategies of embodied interaction focused on re-creating, reimagining, or casual interaction. We provide valuable recommendations for practitioners seeking to design interactive generative AI experiences in museums, galleries, and public event spaces.

Artificial Intelligence (AI) tools have recently gained widespread interest for image creation, but tool developers have largely focused on technical capabilities or specialized domain uses, rather than visual artists as users. We collected survey data from 89 practising visual artists and conducted follow-up interviews with 30 of them, to better understand their diverse needs and values. Through reflexive thematic analysis, we explored visual artists’ attitudes towards collaboration in art creation both with human artists and with AI- and other technology-based support systems. Our results suggest that the focus of popular AI tools on high-quality, finished images does not meet the needs of visual artists. Instead, they wanted reference images, ideation support, and variant exploration. We identified similarities and differences between how visual artists view collaboration with other artists or with machine support, enabling designers of new tools to adopt a more user-centered approach.