Multi agent system helps people describe what they want in custom art
MAIA: Multi-Agent Intent Articulation for Requirement Discovery in Art Commissions
Human-Computer InteractionArtificial Intelligence
Summary
People who commission custom artwork often know what feelings they want but struggle to describe them clearly to artists. The authors created MAIA, a system where multiple agents ask thoughtful questions to help users turn vague ideas into clear, visual descriptions. This process lets users confirm each suggestion, improving how well intentions are articulated before work begins. Tests showed MAIA helps people provide better instructions that artists find easier to understand and use.
What this means in practice
- •For commissioned artists: Get clearer client briefs by using a question-driven system that turns vague ideas into verified visual instructions.
- •For user experience designers: Incorporate multi-agent inquiry systems that help users express abstract goals into concrete, verifiable requirements.
Authors
Yu-Chao Wang, Yanhong Lu, Yingjie Victor Chen, Tim McGraw
Abstract
In bespoke art commissions, laypeople know what they feel but lack the words to specify it: one participant wanted a laid-off truck driver depicted as "a ghost in his own machine" but left the medium, scale, and palette unsaid. We frame this as an articulation bottleneck at an under-served upstream stage: requirement discovery, which precedes any artist or image generator and forces the commissioner to constitute intent in the first place. We present MAIA (Multi-Agent Intent Articulation), a multi-agent system that scaffolds this stage through Socratic inquiry under a "Verification over Invention" rule, turning vague affect into a text-only brief of visual terms the user verifies. In a within-subjects study (N = 16), the full configuration produced a large, significant gain in Cognitive Support over a minimal baseline (r = 0.96, p_FDR = 0.015; LMM p_FDR < 0.001). Thematic analysis traces the same mechanism, and a validator gate structurally blocks unratified content. A complementary blind review by three professional concept artists on a sampled set of briefs corroborates this improvement from the artist's side: AI rewriting improved visual completeness and executability in all eight sampled tasks (task-level Wilcoxon p = 0.008; FDR q = 0.010), with directionally larger gains under MAIA than under the baseline (underpowered, d = 1.4-2.6).