HYPERMINER_EXTRACTED EARTH: Colonialism, the Anthropocene, and the Extractive Gaze
The extraction of Earth’s most valuable materials has long been entangled with colonial expansion, capitalist accumulation, and the epistemic violence of Western scientific and economic systems. From the colonial-era plundering of gold, silver, and rare spices to contemporary mining operations driven by machine learning and hyperspectral imaging, the logic of extraction remains fundamentally tied to systems of control and asymmetrical power relations. HYPERMINER_EXTRACTEDA EARTH situates itself within this discourse, critically interrogating the hyper-acceleration of resource extraction in the age of artificial intelligence, remote sensing, and data-driven decision-making. Within the framework of the Anthropocene,
HYPERMINER_EXTRACTED EARTH points to the centrality of resource extraction in shaping both planetary and political futures. The Anthropocene, often framed as a neutral geological epoch, is inseparable from colonial histories: the industrialization that triggered climate change was built upon extractive economies that disproportionately exploited colonized lands and labor. Uranium mining offers a particularly potent example of this entanglement. As a material that underpinned the nuclear age, uranium was extracted from colonized and Indigenous lands—from the Belgian Congo’s Shinkolobwe mine, which fueled the Manhattan Project, to the Navajo Nation’s uranium mines, which left behind environmental devastation and radiation poisoning. This history reveals how the Anthropocene is not simply an era of human impact on Earth but a consequence of specific political, economic, and colonial forces that continue to dictate whose lands are sacrificed and whose bodies bear the cost of resource extraction. From a post-colonial perspective, the expansion of AI-driven mining suggests not a rupture from colonial extractivism but its intensification. The transition from human geologists to algorithmic detection does not undo the colonial logic of resource allocation—it merely accelerates it, rendering landscapes legible as sites of capital potential rather than as ecosystems or lived spaces.
HYPERMINER_EXTRACTED EARTH confronts viewers with the paradox of hyper-extraction: the transformation of natural resources into objects of aesthetic and economic value at the cost of environmental and cultural destruction. This artificial geological formation does not belong to deep time but to an accelerated temporality of human intervention, where minerals are not given time to form but are violently forced into new configurations. The work thus operates as both a critique and a provocation—revealing the tensions between technological progress, colonial legacies, and the future of planetary survival in the algorithmic age.
WHAT DO YOU SEE
In the artwork we see how deposits are allocated, extracted (speculative) and used to create a new geode composed of the most valuable natural resources found on Earth. TRIPTYCH Screen one: a cubesat in orbit is scanning Earth for rare earth deposits. Screen two: valuable natural deposits are allocated by the means of machine learning, hyperspectral imaging and detection. Valuable natural elements like rare earth metals are mined. Screen three: rare earth elements are used to create a geode (extracted earth).
PROJECT CONTEXTUALISATION
Hyperspecral data from CUBESATs equipped with machine vision in outer space are used to allocate deposits and natural resources on Earth with unprecedented accuracy and vision. As major discoveries of near-surface mineral deposits are declining globally, new methods are needed to detect economical deposits at great depths. However, this is challenging due to the relatively small size of ore deposits, the limited number of existing geological data at depth, and limitations of the geophysical methods used for their detection. Machine learning can aid in developing better models for the prediction of rock type and economical mineral deposit locations for extraction purposes without engaging in time and resource-intensive approaches.
GENERAL CONTEXTUALISATION
The need for monitoring environmental dynamics is now even more urgent given the acknowledged impact of climate change, sustainable food sources, and intensified need for green energy. Different visualisation tools exist for geospatial data, however, these are not suited to capture the complexity of hyperspectral imagery. This makes our project relevant, trans-sectoral and trans-disciplinary.
EXTRACTIVISM
Mass-scale development of digitalisation lead to a situation in which digital devices not only are a constant element of our existence and landscape but transformed us into functional hybrids of bodies welded with small boxes full of microchips. For smartphone production, in addition to aluminium, palladium, copper, or gold, there is a need for minerals of little telling names like cobalt, yttrium, lanthanum, or neodymium, whose extraction requires advanced technologies, which, consequently, carry many long-term ecological, social, and political ramifications.
Modern forms of extractivism involve not only the extraction of the planet, its resources, and the biosphere. They reach far out, beyond the Earth, embrace the cosmos and aim at other planets of the Solar System. For the first time in human history, extractivism reaches into a human as well, to their deepest desires, behaviours, and experiences. Extracting data from human life, and gathering traces of online activity – so-called behavioural surplus – has its analogies with colonial strategies. Data acquired simultaneously to the acceptance of “cookies” or terms of use, is, as a resource, processed, refined, and, eventually, divided into fractions – human desire transforms into attainable profit. This data, regarding current and future behaviours of particular individuals, is appraised and sold on the global market of Internet advertising. Serving surveillance capitalism, extractivism monetises deep knowledge of each entity and element of the ecosystem. Our every step on the Web is carefully monitored and painstakingly analysed. Unconsciously, we produce a brand-new type of knowledge, capitalised by great business, and also used for testing devices of collective control. – Ref. https://2022.biennalewarszawa.pl/en/ekstraktywizm/
TECHNICAL INFO
The audiovisual artwork is made in Touchdesigner. The installation can run realtime or presented as a video (2K/4K/8K). Realtime data from Google Earth is combined with hyperspectral images from satellites with embedded machine learning.
CREDITS
The project was realised during the “Geographies of an AI” residency organised by the Onassis Stegi Foundation, 2020/21. Warm thanks to Daphe Dragona and Michelle Kasprzak for their valuable input. We also thank SCANWORLD for helping us out with hyperspectral data and images. We also would like to thank GOOGLE EARTH.
PRODUCTION
Studio De Wilde © all rights reserved – 2021 INQUIRIES studio.frederik.de.wilde@gmail.com