[Paper Review] What if we have Meta GPT? From Content Singularity to Human-Metaverse Interaction in AIGC Era

Lik‐Hang Lee, Pengyuan Zhou+2 more|arXiv (Cornell University)|Apr 15, 2023

Virtual Reality Applications and Impacts4 citations

TL;DR

This paper proposes an AI-Generated Metaverse (AIGM) framework to overcome content and usability bottlenecks in the metaverse by leveraging AI-generated content (AIGC) as a catalyst for content singularity and human-metaverse interaction. The framework enables human-AI collaboration, AI generation, AI-assisted editing, and multi-user co-creation, embedding AIGC deeply into virtual-physical blended worlds to democratize content creation and enhance user experience.

ABSTRACT

The global metaverse development is facing a "cooldown moment", while the academia and industry attention moves drastically from the Metaverse to AI Generated Content (AIGC) in 2023. Nonetheless, the current discussion rarely considers the connection between AIGCs and the Metaverse. We can imagine the Metaverse, i.e., immersive cyberspace, is the black void of space, and AIGCs can simultaneously offer content and facilitate diverse user needs. As such, this article argues that AIGCs can be a vital technological enabler for the Metaverse. The article first provides a retrospect of the major pitfall of the metaverse applications in 2022. Second, we discuss from a user-centric perspective how the metaverse development will accelerate with AIGCs. Next, the article conjectures future scenarios concatenating the Metaverse and AIGCs. Accordingly, we advocate for an AI-Generated Metaverse (AIGM) framework for energizing the creation of metaverse content in the AIGC era.

Motivation & Objective

• To address the stagnation of metaverse development due to content scarcity and poor usability in 2022-era experimental metaverses.
• To explore how AIGC can act as a technological enabler for accelerating metaverse adoption from a user-centric perspective.
• To propose a conceptual framework—AIGM—that integrates AIGC into the metaverse lifecycle, from ideation to content sharing.
• To examine the challenges of subjective presence and co-perception in shared virtual environments when using AIGC.
• To advocate for a human-centered, democratized content creation model in the metaverse, where users—not just developers—can co-create and shape virtual worlds.

Proposed method

• Introduces the AIGM framework as a pipeline integrating human-AI collaboration, AI generation, AI-assistance, and human editing for metaverse content.
• Employs generative AI models to produce initial content based on natural language prompts from users.
• Supports iterative refinement through AI-assisted editing and multi-user collaboration on content creation.
• Embeds AIGC with interaction logic (e.g., tap-to-activate) and evaluates user performance and cognitive load via AI-driven assessments.
• Integrates AIGC into both standalone virtual worlds and physical-adjacent environments (e.g., AR backdrops, dynamic object removal).
• Proposes that AIGC should be embedded in the metaverse to enable real-time, context-aware content adaptation based on user context and subjective presence.

Experimental results

Research questions

• RQ1 How can AIGC overcome the content scarcity and usability barriers that hindered metaverse adoption in 2022?
• RQ2 What role can AI-generated content play in enabling content singularity and democratizing creation in the metaverse?
• RQ3 How can AIGC support subjective presence and co-perception in shared, multi-user virtual environments?
• RQ4 What are the key design principles for integrating AIGC into a human-centered, immersive metaverse experience?
• RQ5 How can the AIGM framework facilitate seamless human-AI collaboration across content ideation, generation, editing, and sharing?

Key findings

• The 2022 metaverse experiments failed due to a lack of engaging content and poor usability, especially in 3D interaction techniques, despite significant investment.
• Users struggle with 3D UIs due to the absence of clear interaction cues, leading to reliance on trial-and-error and high cognitive load.
• Current AIGC tools are insufficient for generating contextually appropriate content for subjective user states (e.g., 'calm' in a bedroom) without precise prompting.
• AIGC can serve as a savior for the metaverse by enabling non-technical users to create 3D content, similar to how Web 2.0 empowered 2D content creation.
• The AIGM framework enables a full content lifecycle—from inspiration to sharing—through integrated human-AI collaboration and multi-user co-creation.
• Embedding AIGC in both virtual and physical-adjacent spaces allows for dynamic, emotionally resonant, and personalized virtual environments.