Through Dawn explores Walter Benjamin’s theory on the disappearance of the "Aura" in artworks of the technological era. By placing artificial intelligence at the core of the creative process, we push the boundaries of traditional art, aiming for a reinterpretation and re-creation of the aura in today’s digital age. Through Dawn is a central piece in this series, using advanced AI model training and generation technologies to blend art with artificial intelligence. The AI model, "Jiayu∗," is trained on the artist’s previous digital works, transforming these images into sequences to teach the AI the artist’s unique visual language and shape its worldview. This model is available on a web platform for public exploration and interaction.

Elements from the artist’s previous works—particle effects, fluid dynamics, 3D scene construction, synthesized visuals, and camera movements—are abstracted into keywords to train the AI. Through this process, the AI learns to perceive a natural world made of abstract particles: mountains, streams, forests, grasses, mushrooms, flowers, birds, bamboo groves, and other symbolic elements that have been focal points of the artist’s work. Once trained, the AI generates new visual worlds based on audience-provided prompts. By integrating real-time climate data via open APIs, the work continuously merges dynamic information—such as wind speed and direction—with geographic inputs from viewers, generating vibrant, ever-evolving visual pieces. Each audience interaction imbues the work with a unique personality, seemingly reviving the "Aura"—that elusive, singular artistic charm—in this participatory creation process.

Through Dawn reflects on how art's uniqueness and existential significance can be preserved in an era where digital replication and dissemination are increasingly pervasive. It is not only a reinterpretation of the "Aura" in the age of technological art but a profound exploration of contemporary artistic practice and theory.

​