Funhouse Mirrors: Institutional Expertise As Representation in the Digital Age

This paper examines institutional expertise in the digital age through a case-study of the participation of non-scientific organizations in the debate on the environmental footprint of artificial intelligence (AI). It investigates the expertise of institutions making epistemic claims through gray literature and provides a framework to investigate what three existing concepts can teach us about institutional expertise, namely “communities of practice”, “epistemic communities” and “international knowledge institutions”. Each concept allows for an investigation of specific aspects of institutional expertise along two axes, one separating tacit and explicit expertise, and another distinguishing expertise as a network versus an individual property. Communities of practice focus on the tacit knowledge of the network of individual experts that are part of institutions and brings to light the values they embed in their methodological decisions. In contrast, epistemic communities emphasize experts’ explicit knowledge, revealing potential discrepancies between what experts claim to represent and what they represent. International knowledge institutions’ attention to the institutional level underscores the work institutions must do to realign what organizations say they represent and what they represent. We conclude by stressing the need to study the tacit knowledge of institutions in order to improve the quality of institutional mediation in public debates about technical decision-making.