The point here is to emphasize that more and more of our interactions in digital “environments” are programmed to produce data at unprecedented scales and speeds. This refers to the three central handling stages—distinguished here for analytical purposes, but empirically much more blurry—of “algorithmic governmentality,” put forth by Antoinette Rouvroy and Thomas Berns. The first step is data warehousing, whereby massive quantities of data are systematically collected from a plethora of practices and interactions and stored, regardless of their utility or significance; after which comes data mining, effectively extracting correlations and patterns through algorithmic calculations, the inductive self-evidence of which lies in the sheer quantity of data that is handled; and finally there is profiling, which aims not so much at identifying specific persons, but at correlating scattered occurrences together so as to reduce behaviors themselves to necessary and temporally successive algorithms.
-
Reigeluth, Tyler Butler (2014): Why data is not enough: Digital traces as control of self and self-control. In: Surveillance & Society 12 (2), S. 243–254.