
Deep Flaw_01
List price:
1.250Ξ ~ $3,855
Incl. limited, signed 3D print, 3/7.
Deep Flaw (2016) is a pioneering milestone in AI-driven sculpture — the first sculptural 3D-printed readymade of the AI age, extending Duchamp’s lineage from Bicycle Wheel (1913) and Fountain (1917) into the algorithmic era.
Where Duchamp redefined art through the simple act of selecting an ordinary object, Deep Flaw extends and relocates this gesture into computation: selection becomes distributed across evolutionary algorithms, datasets, and machine perception.
Created by Frederik De Wilde in collaboration with researchers at the IT University of Copenhagen and the University of Wyoming, the project uses evolutionary computation and deep neural classifiers to generate three-dimensional forms from minimal, low-resolution photographic input. The resulting shapes are confidently labeled by machines as familiar categories, yet appear phallic, uncanny, or distorted to human viewers — revealing the hidden architectures through which machines “see.”
Deep Flaw stands as a 21st-century readymade: a digital sculpture in which authorship is shared between artist and algorithm, and where machine-driven and artist-in-the-loop selection becomes a sculptural gesture on par with Duchamp’s revolutionary interventions.
For collectors, Deep Flaw represents an early, historically significant convergence of conceptual art, artificial intelligence, and digital fabrication — a rare work that captures the moment when AI first began to assert its own visual logic within contemporary sculpture.
Artist Statement:
Instead of choosing a manufactured object, I work with procedures that allow evolutionary algorithms and neural networks to generate shape and form. I curate the output. The work exposes the perceptual tension between human and machine vision — transforming that tension into a new kind of post-readymade sculpture.
Credits:
Center for Computer Games Research (IT University of Copenhagen) and the Department of Computer Science (University of Wyoming, Laramie) — including Joel Lehman, Jeff Clune, and Alexander Holdt.
Notes:
This innovative R&D helped lay foundational concepts for today’s AI 3D model generators.
The full title of Deep Flaw: Deep Flaw – AI]mix – “Perfume Bottle, Mushroom, Dish, Patatas, Mickey Morph.”
- MediumVideo (MP4)
- File Size335.3 KB
- Dimensions1512 x 1512
- Contract Address
- Token StandardERC-721
- BlockchainEthereum

