Exploring Opportunities to Support Novice Visual Artists' Inspiration and Ideation with Generative AI

TL;DR

The paper investigates how generative AI can support novice visual artists in the early ideation and reference-gathering stages. Through a two-phase study—phase 1 with interviews to identify needs and phase 2 with co-design workshops—the authors map novice challenges to six open-source capabilities and develop six prototypes to probe in practice. Findings reveal concrete capabilities desired for early-stage support (reference management, 3D-staging, structured style refinement, explainable learning, and community feedback) and highlight tensions around autonomy, unexpected outputs, and the ethics of AI-assisted creation, including attribution and compensation. The work advances artist-centered AI design and offers policy-oriented guidance to ensure provenance, consent, and fair compensation, aiming to blend AI assistance with preservation of personal voice and craft in novice art practice.

Abstract

Recent generative AI advances present new possibilities for supporting visual art creation, but how such promise might assist novice artists during early-stage processes requires investigation. How novices adopt or resist these tools can shift the relationship between the art community and generative systems. We interviewed 13 artists to uncover needs in key dimensions during early stages of creation: (1) quicker and better access to references, (2) visualizations of reference combinations, (3) external artistic feedback, and (4) personalized support to learn new techniques and styles. Mapping such needs to state-of-the-art open-sourced advances, we developed a set of six interactive prototypes to expose emerging capabilities to novice artists. Afterward, we conducted co-design workshops with 13 novice visual artists through which artists articulated requirements and tensions for artist-centered AI development. Our work reveals opportunities to design novice-targeted tools that foreground artists' needs, offering alternative visions for generative AI to serve visual creativity.

Figures (3)

• Figure 1: Overview of Phase 1 & Phase 2 Methods.
• Figure 2: Mapping of higher-level and granular challenges, relevant existing tools, and broader generative tool categories We find that users broadly face 4 higher-level challenges: (1) quicker and better access to references, (2) visualizing combinations of references, (3) perspective and feedback from another artist, and (4) support to learn art techniques and styles, which expand into ten more granular challenges. Based on these needs, we developed 6 prototypes demonstrating the potential for AI tools to augment creative workflows.
• Figure 3: Development of Co-design Probes. Six interactive prototypes were developed to address needs identified through user interviews and to explore AI-assisted design workflows, belonging to three broad categories. (1) Structure-to-style refinement tools help artists progressively convert initial "structured" forms (e.g., a basic shape, outline, style, or geometric layout) into more aesthetically complex or stylized versions, such as visual editing, style transfer, and medium transfer. (2) Reference, capture, and 3D-staging tools help search, store, and combine references to source material (e.g., style combination) in both 2D and multi-view 3D environments (e.g., interactive 3D creation from a single image). (3) Transparency, learning, and community-based tools help make the most salient parts of the input more transparent, such as step-by-step tutorials of visual content to help artists learn new techniques and styles.