I’ve been thinking, reading and writing about pre-data/digital histories if we are to take seriously the task of decolonizing, then how do we go about it? While I am remotely not there yet, reading a poem from a book on the criminalized tribes of India and a mention of the “automated” responses by some of the Chhara community members to policemen questions got me thinking about a certain level of automation with which we approach places and people of power. Depending on who that “we” is, I am thinking back to my mother’s constant reminder to me as we approach the passport office, the biometric ID office, the police station for passport verification, the bank for account verification. Her reasoning (to my very Brahmin outspoken female self) is “why create trouble?” Why create complications and extra questions when a certain automated performance (not a fake one yet a deliberate gestural confirming one) can achieve the desired result? These are pedagogies of compliance, keeping the larger goal in mind.
As I build my thoughts out more I want to save this piece (http://spheres-journal.org/the-difference-that-difference-makes/) by Maya to read for later. I guess what I am trying to work out is: apart from locating the knowledge/power nexus sedimented in colonial disciplinary formations as well as administrative tools, how do I offer a theorization of the postcolonial state and governmentality and lay out a responsive, alternate program for decolonizing/re-orienting all things datafication that attacks and disrupts extant forms and local configurations of social power in different national, regional, local contexts? Big ask but at least we have to start by unpacking the liberal, rights-based discourse that assumes (and also works) in Western contexts at least when the focus is on contemporary AI systems (or, counterquestion: “is the data reliable at all?”). The other benefit to such a shift (towards re-examining institutions and social, political, racial power that prepares the ground for certain kinds of classificatory practices to flourish or realize goals) is that it allows for a reframing of the AI agenda for the global South. In some sense, the imagination of current AI research is to look for automation, datafication, signs of what “typically looks like AI” but in contexts beyond the North. To me, that is also a kind of extractivism or a narrow approach that refuses to attend to social geographies unless and until they are enrolled in what we have already decided “AI looks like”. So, a shift towards 1) colonial/colonizing legacies in administrative apparatus (as pre-data histories), 2) assessing the potential for exacerbating violence and a speculative, politically committed investigation of how any AI implementation can ameliorate, not simply be attuned to harms or “bias” and hence then 3) extending and reworking the agenda for AI justice or Justice after AI in global South to include these things in the agenda is important.
So the two things I want to write going forward: 1) pedagogies of compliance, automated responses for smooth living before and through digitalization…2) articulating the differential relationships of colonial and post colonial state-subjects to arrive at the harder question of “what AI can do for India” for instance, that is OBVIOUSLY not predicated upon some imagination of the nation as an economic machine (https://www.outlookindia.com/magazine/story/india-news-who-owns-the-republic/302681)…
Other things I want to read before I write more:
https://civichall.org/civicist/world-without-wizards-facebook-cambridge-analytica/