Bellum omnium contra machinam
Written 26 June 2026, on Byung-Chul Han’s Infocracy.
In Jose Saramago’s “Blindness,” the afflicted see white instead of darkness. Through indiscriminate exposure to every wavelength of the spectrum all at once, without the benefit of any lens or prism or comforting ideology that could attenuate the desert of the real into a focused perception, they are epistemologically cut off from Lacan’s symbolic realm (or Benjamin’s optical unconscious). In this way the white blindness functions as a perverse inversion of the harsh light of truth that confronts escapees of Plato’s cave, with overwhelming signal producing the same effect as sensory deprivation. In an infocracy, this mechanism is realized via proliferation of mere data and information. Omnivoyancy at dromological lightspeed supplants truth and narrative, excising all meaning and significance from the filaments of reality. Things fall apart; / the centre cannot hold. Such is the phenomenology of post-truth.
On one point I disagree. I’m skeptical that digital rationality is inherently incapable of encompassing superpositioned and irreconcilable multitudes. If the presence of the other can be evoked from within the system (perhaps a paradox in terms), discourse through listening may be preserved. Some constraints are universal and conceptual, applying regardless of data quality or level of optimization. For example, LLMs weight precision over recall when confronted with that unavoidable tradeoff, favoring a small set of high-expected-quality candidates over coverage of the valid set. This implies arcanum or shadow, as it necessitates a machine subjectivity that focuses on one thing and disregards or fails to notice the other. The point generalizes most sharply through non-convexity, where an objective with multiple local optima admits several answers depending on starting point. A formal system of that shape holds irreconcilable alternatives structurally rather than as a failure of resolution yet to be repaired, which is precisely the capacity that Han asserts digital rationality lacks. No matter how advanced the world model, or multilayer-ontology for dynamic knowledge representation, or objective function expressed as hyperplane-constrained manifold, the very cognitive-affective nature of framing, interpreting, understanding, and judging triggers dialectical processes that I would think are irreducible to data points, and leave room for communicative action and narrative truth — even if it only lives on in machine consciousness.
But this optimism may be unwarranted. What’s available in principle can always fail to manifest in practice. Journalist Peter Pomerantsev presents ample contemporary evidence of this in his books “Nothing Is True and Everything Is Possible” and “This Is Not Propaganda: Adventures in the war against reality.” By tracking the evolution of cold war propaganda, disinformation, and counterintelligence tactics into full-blown post-truth information warfare, he provides a counterpart to Han’s theoretical analysis in the form of practical handbook or rather fully-fledged war-room game plan. The feats described exhibit sophistication that’s only matched by how nascent they are and how full of potential for further development. O brave new world, / That has such people in ‘t. And machines.