By: Airi Lampinen & Kristin Kaltenhäuser
The European Union announced recently that Europe should be a global hub and leader in the development of Artificial Intelligence (AI) that guarantees safety and fundamental rights. Under the hot sun of Portugal and in the cool rooms of a beautiful former church, we gathered for a ECSCW 2022 workshop on CSCW and Algorithmic Systems.
The aim was to investigate how we can approach this challenge from the perspective of Computer-Supported Cooperative Work (CSCW) – building on the kinds of conceptual insight and methodological approaches that our community is known for. Instead of considering algorithms or data in a narrow technical sense, our workshop focused on algorithmic systems and their increasing role in society: “‘It is not the algorithm, narrowly defined, that has sociocultural effects, but algorithmic systems – intricate, dynamic arrangements of people and code. Outside of textbooks, “algorithms” are almost always “algorithmic systems”’ (Seaver, 2019, pp. 418–419).
When it comes to the prior study of algorithmic systems, there has been a strong emphasis on widely used commercial platforms, with scholars examining what platform companies do and how platform users relate to them (Bucher, 2018; Gillespie, 2017; Lee et al., 2015; Rosenblat and Stark, 2016). More recently, increasing attention has been paid to the role algorithmic systems play in the public sector, how they shape civil servants’ work practices, and what implications they have for citizens and non-citizens (Flügge et al., 2020; Holten Møller et al., 2020).
In the workshop, we were particularly interested in algorithmic systems in and as organizations, and the questions that come up when investigating algorithmic systems as part of complex, cooperative work practices. These are issues where questions for computer science are fundamentally intertwined with those of social science – a combination that we consider to be at the heart of the CSCW community’s scholarly pursuit. How do we contribute to the kind of society we believe is best suited for human values of participation, agency, and accountability? What metaphors – either knowingly or unknowingly – shape how we imagine the future of algorithmic systems in, for example, public services?
We were particularly interested in algorithmic systems in and as organizations
The workshop advanced a CSCW perspective on algorithmic systems by bringing together a stellar group of researchers studying algorithmic systems within public institutions, such as health care settings, job centers, asylum decision-making, and libraries, but also in private companies, including both food-delivery platforms and knowledge work settings like Microsoft. The morning session was located in a beautiful old church in the Coimbra city center and organized around presenting and discussing participants’ ongoing projects:
- Kristina Popova and Cláudia Figueras shared their research on the lived experiences and emotion work related to balancing individual and collective responsibility in AI development.
- Juho Pääkkönen discussed the empirical case of library work automation in Finland, with a focus on the experience of the people working with systems and how work practices and routines change when new efforts at automation are made.
- Ida Larsen-Ledet’s presentation focused on knowledge bases and how they scaffold the work of knowledge workers.
- Riyaz Sheikh and Kalle Kusk both presented empirical studies on how workers navigate food delivery systems in different places in the world.
- Hubert Zając shared the case of an algorithmic system at work in collaborative clinical work in Denmark.
- Claus Bossen discussed research on how text processing algorithms are used to support workers in coding clinical documents.
- Kristin Kaltenhäuser and Naja Holten Møller’s presentation introduced research on the role of data-driven technologies in asylum decision-making.
- Asbjørn Amnitzbøll Flügge and Naja Holten Møller also shared research on algorithmic decision support in job centers.
After lunch, we ventured outside for a walk-and-talk through the old town of Coimbra and the Botanical Garden to continue discussions and explore shared interests. This gave participants space to discuss empirical cases, methods, and theoretical stances in meaningful ways. The workshop concluded with reflections, asking critical questions. We are sharing some of them here to allow others to join us in addressing them:
- How might we best develop the CSCW perspective to the study of algorithmic systems from a historically informed standpoint? The notion of algorithmic systems is relatively new, but the CSCW community has been working on related themes since its inception.
- What kinds of organizations do algorithmic systems bring into being? Or, to what extent do these systems create the very problems they set out to solve? Such questions also can lead us to the politics of algorithmic systems, that is, if different systems bring into being different realities, who benefits from the use of these systems, and who suffers the costs?
- What concepts, cases, and methods can help us advance this body of research? Are we interested in technology or rather the things outside or at the boundaries of the technologies we study? In some cases, it is not straightforward what the “work” is or where “the workplace” is located. With increasing datafication, we are also called upon to consider how participatory design methods or ethnographic approaches can be adjusted to fit the empirical realities we are encountering, and to explore possibilities for advancing human-centered data science.
- Finally, how do we as scholars see ourselves as contributing to the European project (if at all) with the commitments of a CSCW perspective? What are our relationships or responsibilities as researchers towards politics and the broader public? What work do notions of “Nordic” or “European” do for us and how might they be harmful?
References:
- Taina Bucher. 2018. If… then: Algorithmic power and politics. Oxford University Press.
- Asbjørn William Ammitzbøll Flügge, Thomas Hildebrandt, and Naja Holten Møller. 2020. Algorithmic decision making in public services: A CSCW-perspective. In Companion of the 2020 ACM International Conference on Supporting Group Work. 111–114.
- Tarleton Gillespie. 2017. Algorithmically recognizable: Santorum’s Google problem, and Google’s Santorum problem. Information, communication & society 20, 1 (2017), 63–80.
- Min Kyung Lee, Daniel Kusbit, Evan Metsky, and Laura Dabbish. 2015. Working with machines: The impact of algorithmic and data-driven management on human workers. In Proceedings of the 33rd annual ACM conference on human factors in computing systems. 1603–1612.
- Alex Rosenblat and Luke Stark. 2016. Algorithmic labor and information asymmetries: A case study of Uber’s drivers. International journal of communication 10 (2016), 27.
- Nick Seaver. 2019. Knowing algorithms. DigitalSTS (2019), 412–422.