Scholarlink

<p>Large Language Models (LLMs) are increasingly used in creative practice, yet designers often lack structured frameworks that support both creativity and theoretical grounding, especially in complex media workflows, such as in Virtual Reality (VR). This paper argues that the VR Dimensionality Framework (8Df) can structure human-AI co-ideation meaningfully by guiding VR experience design across eight interrelated dimensions: spatial, sensory, placeness, temporal, cultural, social, cognitive, and psychological. The study integrates the VR 8Df into a dialogic co-ideation workflow and explores its use across three LLMs: ChatGPT, Gemini, and DeepSeek, in a shared VR narrative design task. As an exploratory, theory-informed comparative case study, the analysis examines how different models respond to the same framework-guided design process. The findings suggest distinct tendencies rather than generalisable differences in models: ChatGPT is implementation-oriented, Gemini is more facilitative and alignment-focused, and DeepSeek has explicit reasoning and is principled in its articulation of design. Differences also emerge in cultural specificity and the structure of practical outputs. The paper’s main contribution is to demonstrate that the VR 8Df provides a meaningful theoretical scaffold for human-AI co-ideation in VR, while offering exploratory insight into how different LLMs may support different phases of design.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Human-AI Co-Ideation in VR Experience Design: Using the Dimensionality Framework (8Df) to Structure LLM-Guided Design

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月4日

查看全文
  • Eugene Ch′ng

摘要

Large Language Models (LLMs) are increasingly used in creative practice, yet designers often lack structured frameworks that support both creativity and theoretical grounding, especially in complex media workflows, such as in Virtual Reality (VR). This paper argues that the VR Dimensionality Framework (8Df) can structure human-AI co-ideation meaningfully by guiding VR experience design across eight interrelated dimensions: spatial, sensory, placeness, temporal, cultural, social, cognitive, and psychological. The study integrates the VR 8Df into a dialogic co-ideation workflow and explores its use across three LLMs: ChatGPT, Gemini, and DeepSeek, in a shared VR narrative design task. As an exploratory, theory-informed comparative case study, the analysis examines how different models respond to the same framework-guided design process. The findings suggest distinct tendencies rather than generalisable differences in models: ChatGPT is implementation-oriented, Gemini is more facilitative and alignment-focused, and DeepSeek has explicit reasoning and is principled in its articulation of design. Differences also emerge in cultural specificity and the structure of practical outputs. The paper’s main contribution is to demonstrate that the VR 8Df provides a meaningful theoretical scaffold for human-AI co-ideation in VR, while offering exploratory insight into how different LLMs may support different phases of design.