Exploring Hybrid Agencies: Distributing Epistemic Practices Among Humans and Machines in Citizen Science

A more transparent setting to investigate epistemological effects of extensive distribution of tasks among humans and machines is citizen science. This is a highly distributed practice where the general public volunteer to advance science by taking part in the research process. During the last decade this practice has developed new digital technologies to cater for big data sets, namely machine learning, that automatically learn and improve their capability after training by members of the public.

The aim of this four year project is to map the general consequences of these new forms of hybrid knowledge production in different scientific and humanistic fields. The significance of this is that distribution have been found to enhance cognition and be central to epistemic gains, but also introduce instances of radical complexity and opacity, increasing epistemological uncertainties as well as the cognitive distance among scientific experts. How can you trust scientific results when you are not in command of all instance in the distributed scientific practice?

Drawing on traditional and digital ethnography and on insights from Science and Technology Studies and Information Science the project combines qualitative and digital methods to analyse the activity traces left by the ML technology, volunteers and scientists on citizen science platforms.