Conference Agenda

As large language models (LLMs) become increasingly integrated into cultural heritage research, practices such as artifact reconstruction, content generation, and virtual restoration are rapidly advancing. Generative artificial intelligence is unlocking new narrative potential for jiandu (bamboo and wooden slips), offering innovative approaches to cultural expression and historical reconstruction. However, general-purpose language models often face limitations in domain knowledge and semantic reliability, making it difficult to ensure accuracy and traceability in jiandu-based generation. To address these challenges, this study proposes a virtual restoration framework for jiandu narratives based on Retrieval-Augmented Generation (RAG) and agent mechanisms. Using the Liye Qin slips as a case study, the LIYE-Agent system integrates structured knowledge, semantic retrieval, and task-oriented generation to achieve coherent scene reconstruction. Findings from discipline-informed participant evaluation demonstrate its effectiveness in enhancing generative accuracy and consistency, providing methodological and technical support for generative narrative design and the digital dissemination of cultural heritage.

Research on generative artificial intelligence (GenAI) in cultural design has primarily focused on generative control and creative efficiency, often overlooking the misalignment between AI task definition and cultural creation. This study explores how GenAI can be culturally adapted within traditional pattern-making practices. Using an ethnographically informed generative probe, field experiments were conducted in the Huayao cross-stitch community, which practices a form of Chinese intangible cultural heritage. The findings reveal how cultural practitioners negotiate creative logic and cultural judgment when collaborating with AI, identifying four key orientations: liveliness, locality, creation-in-progress, and community creativity. Building on these insights, the paper redefines AI’s role as a generative reflexive actor and introduces the perspective of Culturally Situated Task Definition (CSTD) to explain how cultural tasks are redefined through the mutual shaping of technology and practice, advancing an understanding of the mechanisms underlying AI’s cultural adaptation and meaning co-construction.

The centrality of modernist and institution-centred narratives has narrowed the scope of Brazilian graphic design throughout history, limiting its cultural and epistemological diversity. This article explores the value of researching graphic artefacts normally understood as not pertinent to design research, focusing on album covers from the Brazilian rock scene of the 1980s. Based on the formation of an archive of 155 covers, this study emphasizes the need for appropriate methods to analyse mass-culture artefacts whose wide circulation shapes collective design practices and visual imaginaries. By reflecting on the criteria of selection, interpretation, and preservation that determine what is collected, analysed, and preserved (or ignored), this paper advocates for the establishment of more plural, inclusive, and culturally situated collections. Expanding the scope of objects considered legitimate for design research has the potential to strengthen critical perspectives on the discourses that shape design history and to diversify the field’s reference framework.

Visual thinking, a foundation of design, is often framed as a reflective conversation between a designer and their medium. Recent advances in generative Artificial Intelligence and immersive technologies, such as Augmented and Virtual Reality, are profoundly reshaping this conversation. This article explores how these digital tools augment the visual design process, moving beyond traditional methods. Through a series of autoethnographic, practice-based examples, we demonstrate several key transformations, particularly: AI-driven dialogues that rapidly iterate within a visual language and translate across languages; embodied, spatial sketching in AR/VR that enables new forms of visuospatial reasoning; and novel methods for perspective-taking and collaboration in virtual spaces. By analyzing these new modes of interaction, we reveal significant opportunities, critical limitations, and the need for a critical digital-visual literacy. The article concludes by outlining future research directions and critically integrating these tools into design education and practice to augment individuals’ design thinking.

Generative AI tools such as ChatGPT and Vizcom can produce creative 2D images, however, their stylistic appearances remain difficult to control. This study proposes a design pathway based on LoRA fine-tuning technique. First, a fine-tuning model of Nike A.I.R. concept shoes and automotive surface styles were trained. Then, these two different vocabulary elements were combined and generatively tested to explore their feasibility in footwear design. The process included dataset preparation, parameter adjustment, generation testing, and VR 3D modeling. Results showed that this method can effectively generate innovative forms within a given framework, making it particularly suitable for early design exploration. In addition, by adjusting the weights of the fine-tuning model and using the VR design software Gravity Sketch, designers can effectively control the proportions of the generated image elements, enhancing the control and creativity of AI-assisted footwear design.

As professional collaboration increasingly shifts to digital and asynchronous modes, workers use collaboration technologies not only to coordinate tasks but also to manage their professional visibility and reputation. To examine how these practices could evolve as AI becomes more embedded in the collaboration environment, a speculative design approach was adopted to make near-future scenarios experiential. We prototyped speculative AI systems that support visibility through employee self-report reflection, performance scoring, AI-enhanced meeting appearance, communication delegation, and AI-use disclosure as probes for a series of immersive and interactive scenario-based interviews. The findings uncover selective disclosure of AI use to manage professional image, conditional trust in delegating self-representation to AI, attempts to manipulate AI evaluation, and stress resulting from continuous assessment. We contribute (1) empirical insights about potential AI-mediated visibility practices, and (2) design recommendations for collaboration technologies to support control, authenticity, and healthy visibility practices in AI-powered professional environments.