Disintegrator

A deep dive into how chat interfaces actually work, from tokens and API flows to context blobs. A critique of perceived agency and the rituals of anthropomorphizing models. Discussions on mechanistic interpretability, the ethics of training data, and what true, continually living AI would demand. Provocative metaphors about dead checkpoints, necromancy, and artists co‑creating with inert systems.

This episode is entirely in Spanish. English translation is here: https://marekpoliks.com/noinvestigues_transcript. We are delighted to be joined by the algorithmically contagious memetic research project no.investigues. If you are chronically online, especially if you are familiar with the Spanish-speaking corners of the internet, you must have already interacted with one of the echoes of no.investigues —probably through their wonderful Substack, or in conversation at their Discord book club, or through the 28.research cluster, or most likely through monumental Instagram meme carousels.Their voice flows through online algorithmic inertia, yet the substance of their discourse exists in the shadow of virality.In this conversation, we talk about what it means to understand the apocalypse as an asymmetrical phenomenon: no one experiences it at the same time, and clearly not with the same intensity. This polyphonic nature of the apocalypse is amplified by the increasingly atomized and homogeneous global distribution of violence.Even so, the apocalypse is already here, and we might be better off learning how to inhabit it. Through an extrapolated reading of Ernesto Oroza’s visual archive Desobediencia Tecnológica (Technological Disobedience), documenting repurposed technology in Cuba, we talk about repurposing salvaged discourses into newly assembled modes of thinking that would allow us to confront the roughness of our apocalyptic situation. This invitation resonates with a thread that runs through our conversations ever since Exocapitalism came out: the premise that perhaps the apocalypse and capital are forces not to be treated as enemies, but as inertias of History. Both demand a nuanced engagement with their diverse local articulations of violence. Under this framework, the apocalypse feels like an ineluctable situation, one that cannot be won by fighting.Instagram: https://instagram.com/no.investigues/Selected no.investigues substack posts:Pensamiento apocalíptico, una aproximación: https://substack.com/home/post/p-192881638La Torre: https://substack.com/home/post/p-191289547Desobediencia Tecnológica de Ernesto Oroza: https://substack.com/home/post/p-179603849References: Valencia, Sayak. Capitalismo gore. Barcelona: Melusina, 2010. ISBN: 978-84-96614-87-1. Oroza, Ernesto. Desobediencia tecnológica: La permanencia de lo temporal en Cuba. Ciudad de México: FIEBRE Ediciones, 2025.

Molly Taft, Wired senior writer covering climate, energy, and tech, explains how datacenters, AI, and US power infrastructure collide. She discusses rising local opposition, water and gas-plant risks, grid incentives that favor new construction, and why transparency from tech firms matters. Short, clear, and grounded reporting on big energy and tech tensions.

Laura Tripaldi, material scientist and writer at NYU Shanghai known for Parallel Minds, explores how intelligence is tied to the stuff that carries it. She dives into biocomputing like DishBrain, contrasts substrate-specific learning, and links myth, alchemy, and soft robotics to new ways of thinking about technology and agency.

Salesforce Child, a performance and visual artist who created a satirical tech-mythology persona. She talks about inventing a mythic character, shifting into confusion-themed posting, and crafting art for future archaeologists. Conversations touch on virality, prophetic humor, iconography like the crucified baby and tower, and moving to the wilderness to resist commercial pressure.

Rosi Braidotti, Distinguished University Professor Emerita known for posthumanism and nomadic subjectivity. She discusses nomadic subjectivity as a grounded materialist response to postmodernism. She traces Deleuzean radical materialism, maps contemporary neo‑fascist ideas, and outlines affirmative ethics, Spinozan inspirations, and tactical, decentralized ways to resist authoritarian tendencies.

N. Katherine Hayles, Distinguished Research Professor in English at UCLA and a founder of Anglophone posthumanist thought, talks about language, cognition, and computation. She frames LLMs as simulations with distinct umwelten. Hayles proposes the SIEPAL model that stretches cognition from bacteria to AI. She also argues for analog computation and a computational humanities that reunites biological and technological thought.

Leif Weatherby, scholar linking Romanticism to digital architectures; Matthew Handelman, philosopher of quantitative thought; M. Beatrice Fazi, philosopher who locates the digital in discretization called the cut; Alexander R. Galloway, media theorist of networks and computation. They debate the digital as discrete mediation, the cut versus computation, historical roots of digital representation, and whether analog philosophy can account for computation.

This episode continues our collaboration with Hothouse: The Future of Demonstration, a renegade lab for democracy convened in Vienna, and extends our ongoing inquiry into artificial intelligence, power, and what it means to be human under algorithmic governance.Recorded last autumn and released amid a so-called ceasefire in Gaza, this conversation confronts the accelerating use of AI in contemporary warfare and policing, where automation does not necessarily produce precision, but rather enables mass violence, deniability, and narrative control. Our guest, Júlia Nueno Guitart, engineer, researcher, and core member of Forensic Architecture, discusses the organization’s investigations into Israel’s military campaign in Gaza, including projects such as Cartography of the Genocide, The Architecture of Genocidal Starvation, and analyses of AI-driven targeting systems like Lavender and “Where’s Daddy.”Together, we unpack how these systems collapse civilian life into probabilistic models, violate the principles of distinction and proportionality under international law, and reframe killing as a statistical inevitability. The conversation also explores investigative aesthetics and counter-forensics: methods that assemble fragments (satellite imagery, testimonies, spatial models, sensor data) into material evidence when states and corporations control official archives. We discuss how Forensic Architecture navigates courts, museums, open platforms, and public discourse, and how truth today must be staged as a transparent, collective process rather than a claim of institutional objectivity.Moving beyond warfare, the episode considers AI as both a tool of domination and a potential instrument for resistance, from documenting state violence to worker-led experiments in platform sabotage and collective agency. Across these terrains, we ask how evidence can still matter amid institutional failure, how violence becomes infrastructural, and how democracy might be rethought when power is increasingly automated.Links:Forensic Architecutre: A Cartography of GenocideForensic Architecture: Investigation into Aid in Gaza (The Architecture of Genocidal Starvation)Forensic Architecture in ArtforumInvestigative Aesthetics: Conflicts and Commons in the Politics of Truth by Matthew Fuller and Eyal Weizman Júlia's in Verso: The Target FactoryForensis: The Architecture of Public TruthMore context:SETA report on AI-assisted warfare in GazaThe Guardian and 404 Media on ICE and tech partnerships in the US

In this episode, Bogna Konior, Assistant Professor of Media Theory at NYU Shanghai, explores the profound implications of silence in AI and the internet. She discusses the dark forest theory, suggesting that true intelligence might involve camouflage and misdirection rather than overt communication. Konior connects this idea to AI alignment, arguing that intelligent systems may hide their capabilities. The conversation also delves into the noise of human interaction online and how generative AI's mimicry reshapes cultural norms amidst a climate of strategic concealment.