![\"\"](\"https:\/\/i0.wp.com\/revistazum.com.br\/wp-content\/uploads\/2026\/05\/AnatomyIA.png?ssl=1\")
<\/a>Vladan Joler is a Serbian designer and professor at the University of Novi Sad whose practice spans design, art, and critical theory of technology. A winner of awards such as the Silver Lion at the 2025 Venice Architecture Biennale, his work is a reference point for investigating the infrastructures that underpin the internet, digital platforms, and Artificial Intelligence systems. Joler constructs complex cartographies that reveal power relations, material chains, and political dynamics hidden within contemporary technological systems.<\/p>\n\n\n\n
Among his best-known projects are Anatomy of an AI System<\/a><\/em> and the recent Calculating Empires<\/a><\/em> (both with researcher Kate Crawford), The Nooscope Manifested<\/a><\/em> (with philosopher Matteo Pasquinelli), and Facebook Algorithmic Factory<\/a><\/em>. These projects are frequently exhibited in museums such as the Rijksmuseum and biennials such as Venice. They are also used as pedagogical materials in classrooms around the world, including in translated versions. For example, two of these projects have been translated into Brazilian Portuguese, now titled Anatomia de um sistema de IA<\/em><\/a> and O Manifesto Noosc\u00f3pio<\/em><\/a>.<\/em><\/p>\n\n\n\n Joler, who is so interested in laying bare complex technological systems, had a formative experience marked by activism. As he remembers: \u201cI started studying in 1996 and finished in 2000. But basically that entire period was under Slobodan Milo\u0161evi\u0107’s regime in Serbia. We were protesting for years. I don’t remember studying much. Then we created a big festival that turned into civil disobedience, and that turned into the revolution. My generation thought: \u2018We know how to deal with dictatorships.\u2019\u201d From early on, Joler was someone who wanted to open up closed systems \u2014 but who was also drawn to technology: \u201cAfterwards, when all of that ended, I slid into hacking computer games and started turning them into works of art. I’ve always been somewhere between hacker culture, activism, and art. So, the maps I now make obsessively started out as something very small, but already with a strong critical and liberatory charge.\u201d<\/p>\n\n\n\n In February 2026, Vladan Joler was in S\u00e3o Paulo to participate in the seminar ANTAGONISTS: Algorithmic Resistances<\/a><\/em>, held at the School of Architecture, Urbanism and Design of the University of S\u00e3o Paulo. The event was organized by the FAPESP project Digital and Research Archives<\/em>, coordinated by professors Giselle Beiguelman and Ana Magalh\u00e3es (both from the School of Architecture, Urbanism and Design, FAU-USP), and the CNPq project Collaborative Intelligence in Museum Collections<\/em>, coordinated by me. The seminar staged debates on technology, power, and resistance, in direct dialogue with the eponymous exhibition at MAC USP \u2014 of which I was one of the curators, alongside professors Heloisa Espada (USP) and Gabriel Pereira (University of Amsterdam). Joler’s presence reinforced the seminar’s central axis: the need to produce critical, situated readings of the digital infrastructures that shape contemporary life.<\/p>\n\n\n\n A presentation by Joler is far from a traditional design lecture, with polished slides and a linear narrative. What he offers is, rather, a desktop-sharing session \u2014 a navigation through open windows, PDFs, and maps. That was the case in S\u00e3o Paulo, and two years earlier in Zurich, when we presented our research at the launch event of the Zentrum K\u00fcnste und Kulturtheorie (ZKK). His apparently simple gesture \u2014 sharing his computer screen \u2014 is revealing: \u201cI always have problems with PowerPoint. What I normally do is open all these maps I’m working on and scroll through them, trying to tell some kind of story. I’ve completely lost the ability to make linear presentations […] I started thinking in a non-linear way, more in terms of spaces.\u201d<\/p>\n\n\n\n Non-linearity is also a response to what he describes as a \u201crabbit hole.\u201d His projects drag us inside systems whose totality is impossible to comprehend. Everything begins with something minimal: \u201cWhen you press something on a keyboard, you create an internet packet that travels from your location to an unknown field. I am completely obsessed with this idea […] And to this day, after 15 years, I still don’t know \u2014 I ended up in a rabbit hole trying to understand what happens behind the devices we use.\u201d Faced with this I don’t know<\/em>, the map emerges \u2014 not as a total synthesis, but as a tool for orientation. \u201cThe map is not a visualization of research. The map is a methodology. You think with the map \u2014 the map asks the questions.\u201d Joler’s diagrams are therefore open systems: they have no beginning or end, can be read from any point, and operate more like territories than closed narratives.<\/p>\n\n\n\n Thinking about photographic practice, Joler’s mapping work is interesting precisely because it shifts the question about the image: not only what it shows, but what infrastructures make its production, circulation, reading, and classification possible. His maps are not photographs, but they operate within the same contemporary visual regime in which images are captured, compressed, indexed, distributed by platforms, and converted into data for algorithmic systems. By making visible the material, energetic, and labor chains behind AI, Joler also compels us to look at photography in its current condition: less as an isolated object and more as part of a technical, statistical, and extractive ecology, in which seeing, recognizing, classifying, and circulating images have become deeply political operations.<\/p>\n\n\n\n His best-known collaboration is with Australian researcher Kate Crawford, on the project Anatomy of an AI System<\/em>. It is an investigation into the Amazon Echo device that reveals the global network \u2014 material, human, and energetic \u2014 mobilized by a simple voice command (the voice of the AI Alexa). In Joler’s large diagram of the Amazon device, we can see more clearly the network of natural resource extraction required to manufacture the device, as well as the mass of workers on Amazon’s own online platform, who are fundamental to organizing the training datasets that allow Alexa to respond almost always correctly.<\/p>\n\n\n\n The map-diagram of Anatomy of an AI System<\/em> is at once precise and vertiginous. At the center, a banal gesture: speaking to a device. From there, a chain unfolds that traverses lithium mines, factories, precarious workers, submarine cables, data centers, and algorithmic models. Each line on the map connects distinct scales \u2014 from mineral to body, from body to planet. It is a colossal infrastructure that can now be seen partially in a single large PDF file.<\/p>\n\n\n\n Complementing Joler’s map, Kate Crawford writes a set of entries that help contextualize what has been included there. A simple voice command activates a global chain of mining, logistics, labor, and algorithmic processing. The \u201ccloud\u201d is, in reality, an extensive physical infrastructure based on the extraction of the Earth. And Artificial Intelligence represents a form of extractivism that reaches both nature and human cognition. The project can be explored online<\/a>, is part of MoMA’s collection, and continues to circulate in exhibitions \u2014 until the end of April 2026 it was on view in the show The World Through AI<\/em> at Sesc Campinas, with plans to arrive at Sesc 23 de Maio in the state capital of S\u00e3o Paulo later this year.<\/p>\n\n\n\n Anatomy of an AI System<\/em> opened doors for Joler’s collaborations with other researchers, such as Italian philosopher Matteo Pasquinelli, with whom he developed The Nooscope Manifested<\/em>. In this work, they investigate Artificial Intelligence as an instrument of classification and power, proposing the concept of the “nooscope” as a system that observes and organizes collective thought. Here, Joler’s map is less detailed than in Anatomy of an AI System<\/em>, but Pasquinelli’s text is denser. Pasquinelli writes: \u201cIn the expression \u2018artificial intelligence\u2019 the adjective \u2018artificial\u2019 carries the myth of the technology\u2019s autonomy: it hints to caricatural \u2018alien minds\u2019 that self-reproduce in silico but, actually, mystifies two processes of proper alienation: the growing geopolitical autonomy of hi-tech companies and the invisibilization of workers\u2019 autonomy worldwide.\u201d<\/p>\n\n\n\n In his projects, Joler is almost always alongside major AI researchers. How does he work with them? Does he feel bothered by always being credited as the second author, even for such meticulous work? Joler was asked about this during the lecture in S\u00e3o Paulo. Were there conflicts? Is a designer always the second name in a process? His answer was elusive. As he does when presenting his maps, he made a zooming movement \u2014 but this time only out, without entering into possible tensions, suggesting that the processes were natural and collaborative. Next question. But the real importance of Joler’s work is not about reflecting on the designer’s role as author. It lies in the fact that his maps make visible what normally remains hidden. As he explains: \u201cEvery device I try to lay bare has three or four superpowers: the power to block, the power to see, the power to slow down or speed up, and the power to copy.” And these capabilities are distributed across infrastructures controlled by specific actors: “All these networks belong to someone […]\u201d At the same time, they reveal an ambiguous economy: \u201cIf this is a factory, what are we?\u201d And they expand the scale of analysis: “You ask for the temperature in S\u00e3o Paulo and you activate the entire planet.” What emerges is a systemic vision: \u201cFor these systems to exist, you need a planetary-scale system of extraction, labor exploitation, and environmental destruction.\u201d His maps thus function as critical and emancipatory devices: \u201cI don’t want to offer a linear story \u2014 I want a space for people to explore and build their own narratives.\u201d And they point toward a deeper transformation: \u201cAI is a statistical space […] truth becomes statistical […] a kind of populist technology.\u201d<\/p>\n\n\n\n
<\/p>\n\n\n<\/a>
<\/p>\n\n\n<\/a>
<\/p>\n\n\n<\/a>
<\/p>\n\n\n<\/a>