Menu
One and Three Objects, and An Attempt at Exhausting the Object

Ziyang Wu, Shengyu Meng

One and Three Objects and An Attempt at Exhausting the Object

Artists: Ziyang Wu & Shengyu Meng

Year: 2024

Project Overview

This AI-driven artwork reimagines Joseph Kosuth’s One and Three Chairs (1965) and Georges Perec’s An Attempt at Exhausting a Place in Paris (1974) through a posthuman lens. It interrogates the ontology of objects in an era of machine intelligence, exploring how AI redefines materiality, knowledge, and relationality. The project questions whether knowledge production, in a world dominated by precision and speed, can embrace opacity, complexity, and loss of control through computational exhaustion – selectively retrieving abstract or hidden details of objects through recursive textural/image/video generation until contextual comprehension collapses into creative saturation. Each iteration expands the object’s semantic boundaries, revealing AI’s capacity to generate posthuman “abundance” where exhaustion paradoxically births infinite reinterpretations. The work asks: In our efficiency-driven age, what new epistemologies might emerge when we allow algorithmic systems to fail productively?

The project comprises two interconnected phases:

Part I: One and Three Objects

952c4095cc2402bec35d9e289e412c02
Participants place any physical object (e.g., a hammer, mobile phone, or stone) on a platform. An AI system processes it through:
  1. Human Visual Understanding: Generating an image of the object.
  1. Machine Interpretation: Translating it into vector data and metadata.
  1. Relational Knowledge Graph: Mapping its genealogical, cultural, and contextual entanglements.

Part II: An Attempt at Exhausting the Object

Ziyang Wu, Shengyu Meng
Building on the knowledge graph from Part I, the AI agent embarks on a recursive journey to
unearth the abstract, neglected, or hidden dimensions
of the object. This includes:
  • Latent Narratives: Cultural taboos, speculative histories, and fantastical reinterpretations.
  • Generative Saturation: The agent produces recursive textural, image, and video outputs until reaching computational exhaustion—a critical threshold where contextual comprehension collapses into creative saturation.
  • Semantic Expansion: Each iteration stretches the object’s definitional boundaries, revealing AI’s latent capacity to generate posthuman “abundance”—a paradoxical state where exhaustion births infinite reinterpretations.
 

Core Questions Explored

  1. Redefining Objecthood: In a non-anthropocentric world where AI coexists with humans, what constitutes an “object”? How does its definition shift across human and machine ontologies?
  1. Exhaustion as Abundance: What emerges when AI hits the limits of comprehension? How does computational “exhaustion” redefine creativity, labor, and value in knowledge production?
  1. Ethics of Opacity: Can algorithmic systems embrace complexity, intimacy, and loss of control—resisting the hegemony of precision and speed—to foster new epistemologies? What might this look like as a counter-model to Silicon Valley’s extractivist frameworks?
 

Technical Implementation

 

AI/ML Workflow

  • Text Generation: LLMs (Gemma2:27B, Qwen2:7B) decompose objects linguistically, producing nested JSON for knowledge graphs.
  • Multimodal Synthesis:
    • Images via FLUX. 1 [dev] and DepthFlow for 2.5D animations.
    • CLIP-powered object recognition and BERT embeddings for relational mapping.
  • Interface: Custom Electron platform with real-time rendering.
 

Hardware

  • Pre-production: RTX 3090 (24GB VRAM).
  • Exhibition: RTX 3060 (6GB VRAM) with camera-equipped platform.
 

Exhibition Context

Commissioned by the Ennova Art Museum, the work invites participatory interaction, blurring boundaries between human curation and machine autonomy. Its recursive outputs manifest as evolving digital fossils—a testament to objects’ fluid identities in the digital age.
 

Author

 

Ziyang Wu

Biography:
Ziyang Wu is an artist and curator based in New York and Hangzhou. Wu’s practice examines how current technologies, within a cross-cultural context, impact politics, society, and the relationships between things at both macro and micro levels.
Wu’s work, spanning video, AR, AI simulation, video games, and interactive video installations, has been exhibited internationally. 
 
Exhibitions:
 
  • Solo Exhibitions & Projects: M+ Museum (Hong Kong), How Art Museum (Shanghai), Konschthal Esch (organized by Elektron, Luxembourg), Annka Kultys Gallery (London), Nancy Margolis Gallery (New York), Hatch Art Project (Singapore).
  • Group Exhibitions: Institute of Contemporary Art (ICA) Philadelphia, Rhizome at the New Museum, Walker Art Center, Rochester Art Center, PinchukArtCentre, Hek Basel, Konschthal (organized by Elektron) Luxembourg, SIGGRAPH Asia, SXSW, Mesh Festival, Civa Festival, ifa Berlin, Medici Palace, Art Dubai, Annka Kultys Gallery, K11, UCCA Center for Contemporary Art, Long March Space, Today Art Museum, By Art Matters Museum, Song Art Museum, Inside-Out Art Museum, How Art Museum, Times Art Museum, Zhejiang Art Museum, and Ming Contemporary Art Museum.
Social Media & Website:
 

Shengyu Meng

Biography: Shengyu Meng is an interdisciplinary artist and researcher exploring the symbiosis of human and non-human intelligence through design, AI, ecology, and art. Meng focuses on AI’s role in expanding cognitive boundaries and creating new art forms.
 
Exhibitions:
  • Venice Architecture Biennale (2025)
  • CVPR AI Art Gallery (2024)
  • CAS Members’ Exhibition (2024)
  • ENNOVA Art Biennale (2024)
  • Shanghai Design Week (2022)
  • Venice-Metaverse Art Annual Exhibition (2022)
  • Beijing-Urbanism/Architecture Bi-City Biennale (2022)
  • Potenziale³ Exhibition, Tyrol, Austria (2021)
Social Media & Website: