Democratic Resilience or Authoritarian Rule? On Societal (Un)Certainty and (In)Stability in the Age of Artificial Intelligence

Against the backdrop of several contemporary crises and processes diagnosed as contributing to democratic backsliding and societal destabilisation – the legitimation crisis of democratic institutions, the erosion of liberal values and norms, the rise of populism – this paper starts from the assumption that late modern societies are currently confronted with a growing certainty gap which plays a role in all these developments. After decades of the scientization of many societal areas, science’s contribution to societal certainty and stability is decreasing in times of algorithmically mediated communication, “alternative facts” and conspiracy theories. At the same time the question arises as to what could (re)stabilise societies in the future. Drawing on the similarity that artificial intelligence currently destabilises former certainties and societal structures, much like science did during the Enlightenment, this paper explores the extent to which AI could contribute to closing the certainty gap and how this might stabilise societies in the future, both in a democratic and authoritarian context. By drawing on the history of modernity, it argues that the stability of modern societies mainly depends on three stability dimensions: (1) the social construction of a shared ‘objective’ reality with certain guiding paradigms, values and narratives (cultural stability); (2) the capability of policy makers to generate “passive legitimacy” (institutional stability); and (3) the widespread perception and/or prospect of material prosperity (material stability). From the perspective of these stability dimensions, it will be analysed how and why science is not (anymore) in the position to significantly contribute to societal certainty and stability and to what extent artificial intelligence, in the form of machine learning (ML) and generative AI (GenAI), could close this gap, both in democratic and authoritarian scenarios.