Click the star to add/remove an item to/from your individual schedule.
You need to be logged in to avail of this functionality.
Log in
Accepted Paper:
Paper short abstract:
We theorise web APIs as political technologies implicated in the configuration of an algorithmic rationality, which is world-making as it sorts the world in specific ways, and tend to reinforce the position of power of a few corporations.
Paper long abstract:
Technologies can be conceptualised as world-making, that is, as political actors participating in configuring the world (Latour, 2005; Suchman, 2007; Winner, 1980). Following this perspective, we theorise web APIs as political technologies implicated in the configuration of an algorithmic rationality (Aradau & Blanke, 2022), which participates in the circulation of power (Foucault, 1980).
Our everyday practices are increasingly mediated by AI systems and digital platforms that deploy algorithms, e.g. to personalise, recommend or rank. The programmability of platforms involves making data platform ready (Helmond, 2015; Plantin et al., 2018) so that algorithms can process it (Bucher, 2018). This involves categorising, tagging, etc. according to specific rationalities. When a platform offers a web API to offer a service or exchange data, this web API structures and sorts data in specific ways. This rationality embedded in web APIs follows a business logic.
This is particularly problematic given that these technologies are increasingly controlled by just a few corporations. Various factors explain this concentration of power—network effects (Rochet & Tirole, 2003), protectionist regulations (Rikap & Lundvall, 2021), and the level of investment, data access, and capabilities needed. Web APIs play an important role as well. On the one hand, they are important governance mechanisms, which in turn force third party users to adapt to its rules and rationality; on the other, web APIs enable the access not only to data, but also to real-time insights that reinforce the capacity of platforms to improve their analytic capabilities (Pujadas et al., 2024).
Opaque APIs: biases, blind spots, and instability
Session 1 Wednesday 17 July, 2024, -