skip to primary navigationskip to content
 

Noopur Raval

ravalNoopur Raval is a PhD candidate in Informatics at the University of California Irvine and a fellow at the Center for Technology, Society and Policy (CTSP) at UC Berkeley (2020). Her interdisciplinary research combines approaches from critical software studies, new media studies and post/de-colonial histories in order to understand the enduring effects of classification projects as they spillover into AI systems. Her writings on gig economy platforms, automation and challenges for decolonizing data science have been published in Human Computer Interaction, Anthropology and Media Studies venues.

For more information on her work, please visit her website.

Project: A deep history of datafication to inform postcolonial AI

Responding to the agenda of the Histories of AI seminar, my project asks how we might shift and reorient AI research to make it responsive to historical struggles around enumeration and classification in the majority of the world? Further, what might current AI and Data Science research learn from anticolonial and post-colonial struggles around quantification, enumeration, classification exercises and their roles in fixing subject positions? Also, what epistemic categories and relational ontologies are missing within critical scholarship on AI research and how might we mobilize non-Western, anti/post-/colonial histories to inform problem-framing in AI?

My project proposes a comparative deep dive into the colonial and post-colonial histories of enumeration and classification in India by reviewing primary sources as well as historical scholarship on colonial administrative censuses and the criminalization of nomadic tribes in colonial and post-colonial India. By reading these two projects alongside and through contemporary critiques of the revival of phrenology and anthropometry in AI, I will offer genealogies of enumerated surveillance with an eye to the political and social life of 'caste' and 'denotified tribes', also two classification categories that have been understudied within information studies.