Algorithmic prediction is disrupting how society manages an uncertain future. Whereas traditional forecasts distribute risk over a population, algorithmic systems predict the riskiness of particular individuals. I explore this transformation in the fields of insurance, medicine, and policing. Giving up the principle of shared uncertainty may come with a heavy cost for solidarity and the social fabric.
Recent advances in algorithmic prediction promise to provide a predictive score for individual persons or singular events, thereby introducing a new way to manage the uncertainty of the future. But knowing the future in advance is not only advantageous. In fact, for our society, uncertainty about the future is also a resource. Since modernity, with the support of probability calculus, various social institutions have developed means of coping with ignorance of the future by starting with the one thing that we all share – uncertainty. The system of insurance, for example, distributes the risk of uncertain future damages over a pool of individuals; medicine treats everyone’s illnesses, not knowing who will need care; public policing manages and controls the widespread and undetermined possibility of criminal acts. My research suggests that giving up the principle of shared uncertainty may come with a heavy cost for solidarity and the social fabric.
Tags: algorithmic prediction, shared uncertainty, opacity, future, discrimination.