Table of Contents
Fetching ...

LACE: Exploring Turn-Taking and Parallel Interaction Modes in Human-AI Co-Creation for Iterative Image Generation

YenKai Huang, Zheng Ning, Ming Cheng

TL;DR

This paper introduces LACE, a Photoshop-integrated co-creative system enabling dual-mode interaction (parallel and turn-taking) between professional artists and generative AI. A pilot with $N=21$ participants demonstrates that the hybrid LACE workflow significantly improves ownership, usability, and art perception compared with single-mode turn-taking workflows, highlighting the benefits of flexible participation styles. By combining layer-based editing with continuous AI suggestions, LACE supports iterative, high-control workflows while preserving authorship. The work lays a foundation for framework-driven comparisons of participation styles in co-creative AI and outlines directions for integrating flexible, iterative AI collaboration into professional art practices.

Abstract

This paper introduces LACE, a co-creative system enabling professional artists to leverage generative AI through controlled prompting and iterative refinement within Photoshop. Addressing challenges in precision, iterative coherence, and workflow compatibility, LACE allows flexible control via layer-based editing and dual-mode collaboration (turn-taking and parallel). A pilot study (N=21) demonstrates significant improvements in user satisfaction, ownership, usability, and artistic perception compared to standard AI workflows. We offer comprehensive findings, system details, nuanced user feedback, and implications for integrating generative AI in professional art practices.

LACE: Exploring Turn-Taking and Parallel Interaction Modes in Human-AI Co-Creation for Iterative Image Generation

TL;DR

This paper introduces LACE, a Photoshop-integrated co-creative system enabling dual-mode interaction (parallel and turn-taking) between professional artists and generative AI. A pilot with participants demonstrates that the hybrid LACE workflow significantly improves ownership, usability, and art perception compared with single-mode turn-taking workflows, highlighting the benefits of flexible participation styles. By combining layer-based editing with continuous AI suggestions, LACE supports iterative, high-control workflows while preserving authorship. The work lays a foundation for framework-driven comparisons of participation styles in co-creative AI and outlines directions for integrating flexible, iterative AI collaboration into professional art practices.

Abstract

This paper introduces LACE, a co-creative system enabling professional artists to leverage generative AI through controlled prompting and iterative refinement within Photoshop. Addressing challenges in precision, iterative coherence, and workflow compatibility, LACE allows flexible control via layer-based editing and dual-mode collaboration (turn-taking and parallel). A pilot study (N=21) demonstrates significant improvements in user satisfaction, ownership, usability, and artistic perception compared to standard AI workflows. We offer comprehensive findings, system details, nuanced user feedback, and implications for integrating generative AI in professional art practices.

Paper Structure

This paper contains 13 sections, 8 figures, 2 tables.

Figures (8)

  • Figure 1: The LACE system supports two interaction modes: Turn-Taking and Parallel. Turn-Taking employs a classic sequential participation style, where human and AI iteratively refine the prompt and generated artifact in sequence. In Parallel mode, LACE maintains the AI loop for prompt iteration while introducing an independent Artist loop, allowing artists to refine their work in real-time without interruption.
  • Figure 2: System architecture of LACE
  • Figure 3: Interface and Workflow of LACE
  • Figure 4: The charts compare Workflow 1 and Workflow 2 against Workflow 3 (LACE) using Likert scale scores across six categories. Red asterisks (*) indicate statistically significant differences between the workflows.
  • Figure 5: Preference of Workflow by Tasks
  • ...and 3 more figures