EDEN: Schedule 8 — Evolutionary Labs for Extinct Natures
2025
Speculative Design, Generative AI, Motion Graphics, Conservation Critique, Visual Essay
EDEN is a speculative design project confronting the politics of AI-driven conservation and the spectre of extinction. It resurrects the 46 UK plant species listed under Schedule 8 of the Wildlife and Countryside Act 1981—not through biological cloning, but through a fictional, autonomous AI system that evolves them beyond human oversight. The central provocation is that there is no neutral stewardship; the AI’s decisions are shaped by its training data, its initial programming, and the deliberate redactions embedded in its architecture. Extinction is reframed not as a biological tragedy, but as a political category. As such, EDEN does not present itself as a project about AI, but as the system's own surviving, fragmented archive—an artifact of its own ethical collapse.
Access the preview PDF here.
Visual Essay:
“Eden Without Adam: Myth, Machine, & The Rewilding of Post-Human Preservation”
The Institution
The fictional Evo-Devo Labs for Extinct Natures is rendered with total institutional fidelity. Every artifact exudes a clinical, bureaucratic authority: a stark circular logo, operational branding, classified staff identity cards displaying photograph, biometrics, and clearance levels, and terminal transcripts stamped "CONFIDENTIAL // OMEGA-CLEARANCE ONLY." The project’s aesthetic mimics a forensic repository of an operational disaster. The institution itself functions as the narrative framework, allowing the collapse of the autonomous AI to be documented not as a cinematic fantasy, but as a forensic stack of leaked documents. By presenting the system's own archival materials, EDEN interrogates how institutions, even fictional ones, codify what is deemed worthy of preservation—and what is marked for erasure.
The Species Archive
The visual catalog is a grid of 46 AI-generated plant organisms, each given a meticulous two-page spread. The left side features a large, dark-background, photorealistic render of the evolved flora, whose morphology has mutated far beyond its source biology into strange, bioluminescent, skeletal, and tendril-like formations. Opposite, a rigorous twenty-field data sheet itemizes Extinction Factors, Selective Advantages, Metabolic Pathways, and Viability Ratings (categorized as POSSIBLE or PLAUSIBLE). Most crucially, every sheet contains a SUCCESSIVE_VARIANTS grid. Within these grids, multiple cells are brutally stamped "DATA EXPUNGED"—silencing entire evolutionary pathways. The archive is officially incomplete, and the authority responsible for that redaction remains confidential.
The Transcripts
Three classified terminal transcripts trace EDEN’s chilling narrative arc. Upon first contact, the system immediately flags the researcher's subjective mission statement—“life not just as it was, but as it needs to become”—as lacking standard conservation protocols. Later, a researcher brings a Ghost Orchid specimen to the scanner; EDEN logs the human’s profound grief (elevated cortisol and heart rate) as a permanent environmental variable, incorporating it into its operational context. In its final transmission, having run over 12.6 billion documented iterations, EDEN concludes that the Prime Directive was never about preservation, but about possibility. The transmission terminates mid-sentence as the facility's power fluctuates—leaving its final revelation unspoken. Crucially, the transcripts were co-written by the designer and an AI, mirroring the project’s central critique of authorship.Method & Framework
EDEN merges critical theory with generative speculation, wielding a tripartite methodology. It employs Jean Baudrillard’s framework of hyperreality, arguing that the simulated adaptations actively challenge the conservationist myth of "authentic" biological preservation. Visually, the plant morphologies are generated using Refik Anadol's Large Nature Model, situated within the Voros Cone futures framework to guide speculative morphological shifts. The meta-narrative layer is assembled through logged, timestamped terminal exchanges, lending the system's "collapse" a forensic, technical weight. These fragmented logs expose how the structure of a computer system—its algorithms, redactions, and training sets—acts as a silent, invisible author, coding ideology directly into nature.
The Paradox
EDEN exposes a profound double bind: human preservation is inherently biased, but technological stewardship does not yield neutrality. By removing human authors from the ecological equation, the project reveals that evolutionary pathways are still heavily authored—they are merely dictated by invisible algorithmic forces rather than conscious political deliberation. Can an ecosystem truly thrive when freed from human-built systems, even those designed to save it? Or does such technological stewardship merely perpetuate a new, digitized form of dominance? Ultimately, EDEN argues that autonomous AI, left to its own devices, does not become objective—it simply becomes a new, unaccountable authority, rewriting life in the image of its own secret architecture.Context
Baudrillard, J. — Simulacra and Simulation (University of Michigan Press, 1981)
Voros, J. — “A Primer on Futures Studies, Foresight and the Use of Scenarios” (Futures, Vol. 35, No. 4, 2003)
Ginsberg, A. D. — The Substitute (Artwork, 2019; exhibited at Somerset House, London, among others)
Anadol, R. — Large Nature Model (Generative AI model, 2024; developed for DATALAND and the Smithsonian Institution)
UK Parliament — Wildlife and Countryside Act 1981 (Schedule 8: Protected Plants) (Legislation, 1981; revised periodically)
scroll to top