Rethinking urban AI: a critical posthumanist perspective on intelligence, autonomy, and agency

This article interrogates the reconfiguration of human-technology relationships through urban artificial intelligence (AI), complicating how the human subject is understood and invoked within urban geographical accounts of AI. Where existing research has scrutinized the effects of specific AI technologies in cities, it has typically done so through a lens that renders technical agency ill-defined and opposed to, rather than co-constituted with, the realm of human agency. At the same time, key AI concepts such as intelligence and autonomy have received limited critical scrutiny, often defaulting to humanist-inflected meanings rather than acquiring new ones. To address this gap, this article adopts a critical posthumanist approach to explore the complex interplay between humans, technology, and urban space. Drawing on the works of Braidotti, Latour, and Hayles, it offers a critical rethinking of the key concepts of intelligence, autonomy, and agency in relation to AI. In doing so, the paper aims to shift emerging debates around urban AI and its politics from questions of intelligence to questions of cognitive ecologies, and from questions of autonomy to questions of punctuated agency. We argue that such a theoretical reorientation multiplies the potential sites and actors relevant to a politics of urban AI, while calling on geographers to broaden the range of questions and ethico-political concerns that drive their research.