E-Flux

artificial intelligence ethics

[ e-flux ]

Abnormal Encephalization in the Age of Machine Learning

Machine intelligence is framed as a contemporary animism and a pathological extension of Theory of Mind: the Turing Test and Singularity discourse anthropomorphize computation while masking the political economy that actually structures “thinking machines.” Against pancomputational mystifications, intelligence is recast as sociomorphic and transindividual (via Vygotsky and Simondon), with machine learning functioning as capital’s encephalizing apparatus that extracts, normalizes, and amplifies social relations—making the central task not to build a One Mind but to politicize collective intelligence against computational capitalism.

[ e-flux ]

[ KADIST ]

1542 – a flood (The Unmanned series)

A machinic cinema trained on its own prior images folds colonial Silicon Valley’s origin myth into a closed feedback loop, staging an inhuman vision that searches for meaning amid the “death” of animal gods and the return of intelligence to its birthplace. By displacing human agency through autonomous cameras and AI editing, the work critiques technological acceleration and anthropocentric historiography, proposing a posthuman, non-teleological temporality where cinema becomes both scientific apparatus and speculative archaeology.

[ KADIST ]

All Watched Over by Machines of Loving Grace

Stages a triangulation between Brautigan’s cybernetic pastoral, Curtis’s critique of computational governance, and Cope’s algorithmic authorship to probe how AI mediates affect, agency, and the promise/failure of technological emancipation. Through glossy, floating CGI objects and synthetic sound/voice, the work reframes the everyday as liminal “transition” zones—an uncanny parallel reality where representation displaces function and the virtual becomes a metaphysical extension of consciousness.

[ KADIST ]

[ e-flux ]

Failing to Distinguish between a Tractor Trailer and the Bright White Sky

Stages a genealogy of “autonomy” as techno-illusion and structural opacity—from early radio-controlled spectacle to Tesla’s fatal misrecognition—arguing that automation’s real danger is epistemic and political: the transfer of agency into black-box infrastructures that naturalize corporate power, alienation, and the erosion of comprehensible control. Against this death-drive of networked capitalism, proposes a counter-aesthetics of legible, hackable, and communal machine intelligence (dérive, random walk, street-level traps) where artists and activists subvert and reimagine automated systems as sites of shared meaning, resistance, and utopian possibility.

[ e-flux ]