Machine learning political orders

Abstract A significant set of epistemic and political transformations are taking place as states societies begin to understand themselves their problems through the paradigm deep neural network algorithms. machine learning order does not merely change technologies governance, but is itself a reordering politics, what can be. When algorithmic systems reduce pluridimensionality politics output model, they simultaneously foreclose potential for other claims be made alternative projects built. More than this foreclosure, actively profits learns from fracturing communities destabilising democratic rights. The transformation rules-based algorithms models has paralleled undoing social international orders – use in campaigns UK EU referendum, trialling immigration welfare systems, COVID-19 pandemic with becoming reconfigured problems. Machine decouple attributes, features clusters underlying values, no longer tethered notions good governance or society, searching instead optimal function abstract representations data.