VARIATIONS
Commissioned through the Ampersand/Photoworks fellowship, 2024-2025.
V3: MODEL COLLAPSE
Mixed media installation presented at The Photographers’ Gallery, London 2025
V3: Model Collapse, installation view. C-type prints, mirror, paint, aluminium, PLA, packing blankets, dust, steel, mdf, timber
Staged in four venues across the UK, Variations by Felicity Hammond is an evolving installation exploring the relationship between geological mining and data mining, image-making and machine learning. V3: Model Collapse is the third chapter of this project, following V1: Content Aware which was exhibited as a public realm installation as part of the Photoworks Weekender in Brighton, 2024 and V2: Rigged exhibited at QUAD, Derby 2025. In Variations, Hammond uses the many landscapes associated with artificial intelligence (AI) to map how digital photographic material makes its journey from mineral to pixel; from the subsurface to the screen. The project aims to reflect on the entanglement of contemporary forms of image making - specifically those made using generative AI tools - with the politics of surveillance, data extraction, and the exploitation of land, resources and labour. Machine learning software relies on an ecosystem that extends far beyond the interfaces its users are presented with.
Framed C-types prints on mirror wall with paint, vinyl tape and PLA
During the staging of V1 and V2, photographs and data were collected from the exhibition spaces and people in them. The collected data and images have been used as training sets for each subsequent variation, including V3: Model Collapse at The Photographers' Gallery. Like AI image creation, the reiteration of past datasets in each new work mimics the constantly evolving data sets that inform machine learning platforms.
V3: Model Collapse offers an encounter with AI hallucinations through photographic images that have been destabilised. Using different methods from painting to collage, the photographs reflect the errors and inconsistencies of AI generated images, where the output is always an imitation.
The resulting images depict distorted perspectives and ghostly figures. They begin to collapse and fall apart, full of impossibilities and inaccuracies. V3: Model Collapse invites us to act as both witnesses to, and participants in, the process of machine learning. In doing so, it asks us to question the long term impact AI has on photography.
Framed C-type prints mounted behind waste off-cut aluminium that came from the process of making the sculptural work for V2: Rigged.
Inkjet billboard print, timber, paint, plastic, dust, timber, wire and waste material from the de-install of the exhibition that came before.