Document Type

Article

Publication Title

International Journal of Human Research and Social Science Studies

Abstract

Generative AI systems increasingly promise rapid 3D asset production for game development, yet their practical viability depends on whether generated models can move beyond visual preview into editable, optimized, engine-ready workflows. This article presents a practice-led comparative case study of a stylized fantasy tavern environment produced through two workflows: a human-authored Blender pipeline and an AI-assisted pipeline using Meshy 6 and Hunyuan 3D. Using a fixed asset list, shared visual theme, documented prompts, production-time tracking, visual comparison, topology inspection, UV-map analysis, and post-generation labor accounting, the study evaluates whether text-to-3D tools function as production substitutes, ideation accelerators, or conditional asset sources. Results indicate that AI-assisted generation substantially reduced first-pass production time: the Hunyuan-assisted reconstruction required 238 minutes compared with 716.06 minutes for the human-authored scene. However, the apparent time savings were accompanied by substantial technical debt, including dense triangulated geometry, fragmented UV maps, inconsistent prompt adherence, material-editing constraints, clipping during placement, and loss of texture integrity during attempted decimation. The human-authored workflow required more labor at the modeling stage but produced assets with greater intentionality, editability, scale control, and technical legibility. The findings support a human-in-the-loop model in which generative AI contributes most effectively to ideation, rough prop exploration, and early prototyping, while artists and technical artists remain necessary for optimization, art direction, retopology, UV reconstruction, material refinement, and engine validation.

Research Highlights

  • The Problem: Generative artificial intelligence systems promise rapid 3D asset production for game development, but their practical viability depends on whether generated models can transition from visual previews into editable, optimized, and engine-ready workflows.

  • The Method: Researchers conducted a practice-led comparative case study evaluating a stylized fantasy tavern environment produced through a human-authored Blender pipeline and an artificial intelligence-assisted pipeline using Meshy 6 and Hunyuan 3D.

  • Quantitative Finding: The human-authored Blender scene required 716.06 minutes to produce, while the Hunyuan-assisted reconstruction required 238 minutes; the human-authored scene contained 123 objects, 141,008 faces, and 293,828 triangles; the artificial intelligence-assisted scene contained 103 objects and 32,781,505 triangles.

  • Qualitative Finding: Artificial intelligence-assisted generation reduced first-pass production time but introduced technical debt through dense triangulated geometry, fragmented UV maps, and material-editing constraints; human-authored workflows required more initial labor but produced assets with greater intentionality, editability, and scale control; generative artificial intelligence functions best for ideation and prototyping, whereas human artists remain necessary for retopology, optimization, and engine validation.

DOI

https://doi.org/10.55677/ijhrsss/15-2026-Vol03I06

Publication Date

6-2026

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Included in

Game Design Commons

Share

COinS