With the diffusion of artificial intelligence technologies (AITs) come grand narratives about their
transformative effects in the political economy. Some claim these transformations to be good news,
some as bad news, and yet others contend that AITs are simply the latest re-articulation of an
entrenched—so no news at all. This debate is based on a faulty premise, however: arguing for or
against an all-encompassing reconfiguration of national economies, let alone the global political
economy, overshoots the analytical target. Instead, we build on critical political economy and
economic geography to develop a more nuanced and precise account of variegated spatial and social
relations under the “rise” of AITs. AITs, we argue, create significant transformations in contemporary
capitalism, but not necessarily of it. As a novel heuristic, we propose to capture variegated AI-related transformations across two key
dimensions of capitalist relations in the GPE—the organization of production and class relations—and
for two key measures of reconfiguration (spatial and agential). With that heuristic, we find selective
cross-spatial integration of data- and AIT-based production chains and value creation. This integration
strengthens hierarchical relations among firms and countries, beneficial to a small number of them
and detrimental to most others. At the same time, AIT diffusion promotes a disintegration of
traditional class identities both globally and within countries. The net effect of increasing integration
in the relations of production and disintegration in class relations is that both states’ capacity to
intervene in markets and the room for intra-class solidarity shrink. Both dynamics damage citizens’
ability to organize collectively and effectively to promote their interests vis-à-vis entrenched economic
interests.
