Remembering Together, Shaping Together: Narratives of Co-Designing Cultural Heritage Artifacts in Collaborative VR
"From remembering to shaping": Narrating Shared Experiences by Co-Designing Cultural Heritage Artifacts in Collaborative VR
HCI Today summarized the key points
- •This article reports research on using VR and AI together to create shared memories of cultural heritage within a 3D space.
- •Based on old street scenes in Shanghai, the research team tested a workflow in which two people collaboratively entered VR and iteratively edited a 3D building generated by AI.
- •Rather than explaining only with words, participants directly changed the position and size of objects to reconcile differences in opinion, making it easier to share each other’s memories.
- •Even when the AI produced results that differed from expectations or overemphasized Chinese stylistic clichés, participants repurposed and revised them—turning them into new narratives.
- •The study shows that cultural heritage is not simply something to preserve; it is a living memory that people create and adjust together.
This summary was generated by an AI editor based on HCI expert perspectives.
Why Read This from an HCI Perspective
This article is highly meaningful for HCI because it shows GenAI not as a mere generative tool, but as a ‘collaboration tool’ that helps people bring shared memories to the surface and align them. In particular, the process of directly manipulating 3D objects in VR to reconcile differing viewpoints demonstrates that the interaction method can change the experience more than the on-screen outcome itself. For practitioners, it offers design hints for collaborative UX; for researchers, it raises questions about how to analyze human–AI–human interaction.
CIT's Commentary
The most interesting point is that the system does not treat AI failures only as ‘errors,’ but turns them into material for shared interpretation. Scenes where incorrectly generated 3D models are not discarded, but re-read and repurposed as installations or sculptural forms, show that AI can function not as an entity that provides the correct answer, but as a medium that triggers negotiation and reconstruction among people. That said, if this kind of flow becomes a real product, state visibility and intervention pathways become just as important as flexibility. Users need to see more clearly what they are changing right now, what the AI is inferring, and at what moment the experience shifts from memory to being driven by the generated artifacts. In Korea’s service environment, this kind of transparency is especially important. The faster the collaboration and the higher the expectations, the greater the value of an interface that helps people immediately understand ‘how much the current system can be trusted,’ rather than relying on quiet automation.
Questions to Consider While Reading
- Q.How do the moments when a user accepts an AI suggestion as a ‘memory aid’ differ from the moments when they accept it as a ‘power to change memory’ in this collaborative workflow?
- Q.If a product includes a feature to reinterpret incorrectly generated results, what mechanisms can preserve users’ creativity while ensuring the original intent is not lost?
- Q.When more people participate than just two, what interface rules should govern adjustment and consensus in a 3D space?
This commentary was generated by an AI editor based on HCI expert perspectives.
Please refer to the original for accurate details.
Subscribe to Newsletter
Get the weekly HCI highlights delivered to your inbox every Friday.