2020 - 2022

Machine learning systems construct meaning through pattern recognition that operates beyond human perceptual boundaries. This work creates a system where algorithms interpret mending techniques, train neural networks, and generate images that reveal the disconnect between computational and human vision. Through collaboration with Susan Yelavich, these algorithmically generated images, then individually reworked, become physical stitching that completes the circuit of translation between artificial and tactile intelligence. The resulting textile works function as echoes rather than replicas, revealing the fragility that binds code, consciousness, and thread—systems where altering a single element can cause complete structural collapse.

The process functions as material archaeology of knowledge systems. Each manipulated image reveals the machine’s incomplete but systematic understanding of repair—patterns visible to algorithms but invisible to direct observation. When transferred to textile practice, these hidden structures emerge as tangible evidence of how different forms of intelligence organize visual information. The work explores how artificial learning might create its own coherent reality, one that requires human intervention to become legible within our sensory framework.

This essay accompanied the exhibition “Echo Chambers” at 1053 Gallery, Fleischmanns, NY, USA.

---

Related Work

Learning Nature • Winter Woods • Cloud Canyon • Dandelions - AI / machine learning generated images and videos