Zhiwan Cheung

PhD Candidate

Zhiwan Cheung is a practice-based artist and PhD candidate in Computational Media and Arts at Hong Kong University of Science and Technology (Guangzhou). His research investigates AI as a creative collaborator, with attention to how machine learning systems shape representation and memory. Working across video and multi-media installation, he develops AI-driven workflows that treat generative models as co-authors within the studio.

His current projects explore predictive drift in image-to-video systems, the politics of training data, and digital reconstruction using Gaussian splatting and neural radiance fields. Across these methods, he studies how computational systems translate lived environments into synthetic images, and how traces of personal and cultural history persist through technical infrastructures.

He received his BFA from Cornell University and his MFA from Carnegie Mellon University. He has exhibited work nationally and internationally in venues such as NURTUREart Gallery in New York, Pica Pica Gallery in Berlin, and The Andy Warhol Museum in Pittsburgh. Zhiwan also received a Fulbright grant to research the Fluxus art group in Berlin, Germany. His work connects artistic practice with critical AI research, foregrounding questions on authorship, embodiment, and the cultural implications of generative systems.

A Matter of Orientation

Zhiwan Cheung / 2023

VR Installation

"A Matter of Orientation" translates Edward Said’s critique of Orientalism into a new digital reality configured by large language models and gamification of storytelling. Players are immersed in an interactive VR temple that recontextualizes Oriental objects from San Francisco’s Asian Art Museum (AAM) into deep-fake videos, image-to-3D models, text-to-image stories and architecture, and generative script writing. In their algorithmic recreation, these historical artifacts amplify the cultural distance already traveled away from their geographic home.

Dong Xuan

Zhiwan Cheung / 2024

Single-channel Video

“Dong Xuan” reinterprets the classic German fairy tale, Hansel and Gretel, blending physical, geographical, digital, and linguistic landscapes. Named after a Berlin-based Vietnamese market, inspired by Vietnam’s largest indoor market in Hanoi, the video captures performances within a culturally hybrid space. Processed through an AI neural network, the footage undergoes a transformation where images degrade and morph into each other, symbolizing a journey through digital limbo.

18th Wenyuan Street

Zhiwan Cheung / 2024

Two Channel Video Installation

“18th Wenyuan” involved scanning my grandparents’ home in Guangzhou, China, now replaced by an elementary school, and creating a virtual environment from the scans in Unreal Engine 5, a video-game engine. From certain angles, the digital scans loose perspective accuracy, resulting in abstraction. This digital space represents the memory of a place that no longer exists.