AI Experiment + Cross-Cultural Collaboration + Saint Martins Insight
Before I formally began using AI to explore Chinese aesthetics, I was already interested in the intersection of creative design and machine intelligence. Prior to joining Central Saint Martins, I had explored a range of AI tools for visual generation and stylistic simulation. Coincidentally, one of my friends had begun developing an AI-based virtual try-on platform called LavieAI. Intrigued by its potential, I joined as a collaborator to explore how we could use AI to reimagine and disseminate the essence of traditional Chinese fashion structures and color systems.
[ here of AI-generated multicultural models wearing Chinese fashion elements—highlighting silhouettes, color systems, and structures.]
At CSM, we are constantly encouraged to think beyond our disciplines. “Can technology become part of fashion? Can culture become a driver of visual expression?” These guiding questions led me to initiate a small-scale visual experiment: I selected AI-generated models of different ethnicities and body types and dressed them in traditional Chinese garments (e.g., stand-up collars, frog buttons, crimson red, moon white). The results, generated using LavieAI, were surprisingly rich in aesthetic depth and cultural sensitivity—even in its early stage of development.
Academic studies echo this possibility. Guo and Xiao (2022) argue that Chinese aesthetics is not merely a visual style but a philosophical and minimalist system of thought that must be deeply understood to be translated correctly. This aligns with my goal of using AI not just as a tool but as a cultural translator—especially in helping non-East Asian bodies visually express “Chinese temperament” and elegance.
▶︎ Guo & Xiao, 2022 – Full article
Ye Wang (2023) further points out that traditional Chinese garments are not only defined by silhouette or pattern but also by spiritual values embedded within them. The challenge, she notes, lies in reinterpreting these elements with a modern visual language. Through AI, I attempted to test whether such a reinterpretation was possible—whether machines could serve as co-creators in carrying forward these intangible aesthetic ideas.
▶︎ Wang, 2023 – Full article
Inspired by Triyanto et al. (2017), I also reflected on the concept of “aesthetic adaptation”—a cultural strategy whereby local identities are preserved by updating their visual language through contemporary tools and mediums. AI, in this context, is not just a neutral generator but a means of keeping traditional beauty alive and evolving in a globalized environment.
▶︎ Triyanto et al., 2017 – Full paper
At the same time, I came across the collaboration between Loewe and Jingdezhen, in which traditional Chinese porcelain colors were translated into a contemporary design collection—receiving widespread acclaim. This made me realize: Chinese aesthetics does not lack depth; what it lacks is a “translation mechanism” that allows it to be truly understood by the world.
▶︎ Loewe × Jingdezhen: Chinese Monochrome Collection
In line with these reflections, Kexin Li (2024) highlights both the strengths and risks of using AI for cultural communication. While AIGC (AI-generated content) has made cultural production faster and more accessible, there is a danger of Western-biased datasets misrepresenting or flattening non-Western cultural expressions. She argues for localized training datasets and interdisciplinary collaboration—precisely what I envision for LavieAI as it grows into a truly global, culturally responsive design engine.(Li, 2024).
Leave a Reply