Abstract

The Fourth Industrial Revolution, driven by digital innovations, has brought significant changes to core technologies, industrial systems, and job structures. Against this backdrop, this study explores the application of AI image generation models, specifically those based on diffusion architectures, in the development of picture book illustrations. Focusing on two widely adopted tools—DALL·E 3 and Midjourney—which represent state-of-the-art implementations of diffusion-based image synthesis, the study adopts a case study approach to evaluate their stylistic characteristics, usability, and associated challenges. The research compares the visual styles generated by both tools, outlines a five-stage workflow for AI-assisted picture book creation, and analyzes limitations in AI creativity, data bias, and ethical concerns. Findings indicate that DALL·E 3 produces whimsical and intuitive illustrations suitable for narrative clarity, while Midjourney excels in artistic expression and visual depth. By highlighting the capabilities and constraints of these AI models, this study offers practical insights into how such technologies can support creative illustration processes. It also emphasizes the need for further research using objective data and statistical methods to assess user satisfaction and broader applicability.