Lumi is a story of snow and luminance, of reflection and planetary light. The video essay by Abelardo Gil-Fournier and Jussi Parikka arranges historical photographic datasets through the perspective of an imaginary synthetic intelligence trained for the repainting of landscapes as part of an operation of climate restoration. Lumi can be seen as a science and climate fiction about a collective intelligence of Artificial Agents that time travel from past photographic images to future programmed landscapes. This intelligence has grown out of those moments of light computing its path, together with the landscape computing its growth and preservation. Lumi is a dataset of snow, ice, and light. It is a film about the reversal of time and the synthesis of landscapes: once made, then remade.

The interaction between light and landscape manifests through various technological mediations: from surveyors depending on the reflection and refraction of light to trace the edges of landscapes, to satellite images capturing atmospheric radiation. These observations generate different forms of representation – annotated photographs, algorithmic classifications, and composite images – each revealing how landscapes both draw themselves through natural phenomena and are drawn through human and technological intervention.

Coming together in conversation, Marco Ferrari, Abelardo Gil-Fournier, and Jussi Parikka discuss these synthetic landscapes, examining how moving images serve as a methodology for investigating distance, proximity, and elemental media. Their conversation reveals the evolution from analog to digital perception of landscapes while exploring the tensions between ahistorical landscape representation and human presence in images. The discussion ultimately frames landscape recognition as a complex interplay between questions of scale and measurement, of natural phenomena, and technological mediation.

This screening and conversation is supported by the project “Weather Reports: Wind as Media, Model, Experience” (2022–24), funded by the Arts and Humanities Research Council (AHRC: AH/W010526/1) and the German Research Foundation (DFG: 468455590).