Artificial Intelligence in the form of different machine learning models is applied to Big Data as a way to turn data into valuable knowledge. The rhetoric is that ensuing predictions work well—with a high degree of autonomy and automation. We argue that we need to analyze the process of applying machine learning in depth and highlight at what point human knowledge production takes place in seemingly autonomous work. This article reintroduces classification theory as an important framework for understanding such seemingly (...) invisible knowledge production in the machine learning development and design processes. We suggest a framework for studying such classification closely tied to different steps in the work process and exemplify the framework on two experiments with machine learning applied to Facebook data from one of our labs. By doing so we demonstrate ways in which classification and potential discrimination take place in even seemingly unsupervised and autonomous models. Moving away from concepts of non-supervision and autonomy enable us to understand the underlying classificatory dispositifs in the work process and that this form of analysis constitutes a first step towards governance of artificial intelligence. (shrink)

The rise of Big Data in the social realm poses significant questions at the intersection of science, technology, and society, including in terms of how new large-scale social databases are currently changing the methods, epistemologies, and politics of social science. In this commentary, we address such epochal questions by way of a experiment: at the Danish Technical University in Copenhagen, an interdisciplinary group of computer scientists, physicists, economists, sociologists, and anthropologists is setting up a large-scale data infrastructure, meant to continually (...) record the digital traces of social relations among an entire freshman class of students. At the same time, fieldwork is carried out on friendship relations amongst the same group of students. On this basis, the question we pose is the following: what kind of knowledge is obtained on this social micro-cosmos via the Big and Small Data, respectively? How do the two relate? Invoking Bohr’s principle of complementarity as analogy, we hypothesize that social relations, as objects of knowledge, depend crucially on the type of measurement device deployed. At the same time, however, we also expect new interferences and polyphonies to arise at the intersection of Big and Small Data, provided that these are, so to speak, mixed with care. These questions, we stress, are important not only for the future of social science methods but also for the type of societal knowledge that may be expected from new large-scale social databases. (shrink)