skip to content

Histories of Artificial Intelligence: A Genealogy of Power


Friday 29 January 2021, 15:00–17:00 GMT

Co-facilitated by: Simon M Taylor (UNSW Sydney), Rachel Brydolf-Horwitz (UBC), Shreeharsh Kelkar (University of California, Berkeley)

Simon M Taylor, PhD Candidate UNSW Sydney, Research Associate @ Allen's Hub for Technology, Law and Innovation

Similarity as a Scientific Object: The Mahalanobis Distance Function and Operational Acts of Representation

This paper examines the history of a similarity measure – the Mahalanobis Distance Function – its computational mobility from colonial India into contemporary Artificial Intelligence, including facial recognition. I am interested in how a measure of 'similitude' at a particular period of colonial state-making, helps model wider categories of classification that proliferate in applications today. By tracing this mobility of knowledge I question how advanced networks for facial recognition are connected to older classifiers; how they operate in automated classification of images; and how these 'produce decision acts'. Is this a case of ubiquitous computational values being mobile? Scientific ideas circulating after creation? Or how certain values, become standards, as forms of 'rational discrimination' that challenge social progress – from colonial and non-western epistemes.

Rachel Brydolf-Horwitz, PhD candidate, Geography, University of British Columbia

A number of Israeli technologies companies are marketing sensors – laser or RFID – as benevolent monitoring systems for older adults in North America. These sensors come accompanied with proprietary AI systems that allow for massive data collection of the activities of daily living, including location in space, body position, duration of time spent in a certain room or section of a room, and more. The reach of these sensors can be extended through a variety of compatible wearable technologies, so even outdoors seniors are within range. Gathering this data and creating a digital profile, complete with predictive analytics and risk assessments, is presented as a form of care, a way for relatives and professional caregivers to better grasp and address the needs of the elderly, and mitigate risks.

Yet the private technology sector in Israel has close connections to the country's military and its mission to maintain a nation in a contested region, fraught with complicated histories of dispossession and dehumanization. My research looks at the genesis of certain technologies in Israel, with their possible military lineages, and their transformation into arbiters of care for homes and senior facilities in North America.

This research, in its very early stages, connects to the seminar's themes by exploring and emphasizing the genealogies of technologies, and their entanglement in the contexts – theoretical and physical – in which they emerge. I also focus on what is embedded and encoded in the technology itself, what Kalindi Vora calls the imaginaries behind and within technology, especially the imaginary of 'the human'. Following the lead of critical race scholars and activists, my research looks at who has power in relation to the design and uses of this technology, and how different groups of people are rendered human or less than human through their orientation around this 'care technology'.

Shreeharsh Kelkar, Lecturer, Undergraduate and Interdisciplinary Studies (UGIS), UC-Berkeley

In many organizations and scientific fields, there is a great deal of excitement around the practices of 'data science' and human-centered design, forms of knowledge production in which technical experts work fast, are practical and open-ended with respect to outcomes, and subject every decision to a data-driven experimental test. This paper argues that these knowledge practices reflect an underlying socio-technical assemblage that I call 'platform expertise' which is different from an older paradigm that I will call 'Programming'. Both assemblages reflect the perspectives of technical reformers, experts who seek to reinvent their institutions through the technologies of the digital computer. The Programming assemblage was constructed in the 1950s, 60s, and 70s by technical reformers who wanted to problematize the notion of decision-making in organizations and elevate the status of experts on information processing like computer programmers, operations researchers. The platform expertise assemblage is currently under construction by Silicon Valley technologists and their admirers across institutions. These reformers do not necessarily hope to reinvent organizational decision-making to make it more rational and more efficient; rather, they hope to use the new practices around networked computing, especially machine learning and user-centered design, to create new spaces of 'innovation' where new kinds of experts like data scientists and designers, drawing on 'domain experts', can govern 'users' in an open-ended fashion. I argue that the power of platform expertise is its ideological indeterminacy that allows reformers to attract others with different goals and political persuasions.

Join our Mailing List!

To receive further information on all our activities (and learn their online coordinates), please subscribe to the HoAI mailing list.

Email us at:

Join our Slack Channel!

To participate in conversations with scholars on topics related to your interests, please join our HoAI Slack Channel.