Organizations increasingly employ machine learning algorithms, whether as part of their original business model or in efforts to optimize and reform existing practices. For this study, we followed a Danish fintech start-up through its process of defining itself as a fintech, developing an algorithmic tool for screening investment portfolios, and launching its core product of sustainable investment opportunities for pensions savings. We found that the company follows a general trajectory of shaping the algorithm, then being shaped by it, but that this process involves a series of more specific sociotechnical arrangements, some of which develop consecutively while others co-exist for shorter or longer periods of time.
To gain deeper understanding of this process, we turned to the concept of figuration, as developed by Couldry and Hepp (2017) and Braidotti (2011), respectively. Beginning from their common understanding of figurations as relational arrangements of social and technical elements, which are to be ‘mapped’ rather than interpreted or otherwise explained, we follow Braidotti’s notion that figurations are living maps of organizations that can be identified as ‘conceptual personae’ or performative images of socio-political becoming. Returning to our case, we can now detail the figuration of our case organization as variously foregrounding the agency of human actors and of the algorithm and show how these agencies are predicated upon each other as well as on other aspects of the sociotechnical arrangement. The algorithm, initially developed to fulfil the organization’s purpose of sustainable investment, increasingly shapes the organization, shifting its purpose from normative potentiality to pragmatic realization. As one of human members of the organization observes:
I was attracted by it, like cool that it was a pension that would be good for a lot of things. But now it is like it has become more specialized in the direction of climate because it is quantifiable and measurable and easy to report. But the more parameters that have to add up, like human rights and women in management, the more difficult it will be to create a portfolio that still gives a good return. Because it should not cost our users any money to protect something good; that is our entire selling proposition.
Here, the agency of the algorithm interacts with economic imperatives of profit-maximization (for the organization as well as its users) to figure the organization as a conduit for turning input (investment portfolios) into output (optimal profitability, given set requirements of sustainability).
Empirically, then, our case is one of an organization that becomes figured more and more as its algorithm. Conceptually, however, this implies understanding algorithms in relation to the organizations they figure – and by which they are figured. An algorithm may be a procedure for turning input into output in accordance with set rules, but that is never all it is. Or rather, the ‘black box’ of that procedure is shaped from the outside as it were, through the relations between algorithmic and other organizational elements (be they social or technical). Understanding algorithms as organizational figuration, in sum, means providing living maps of their relationality, multiplicity and processuality, identifying the ways in which the give shape to and take shape from processes of organizing, establishing spatial multiplicity within a temporal trajectory.