Content area
Algorithms and artificial intelligence (AI) promise unprecedented predictive power, which many associate with greater efficiency, objectivity, and control. They also imperil many cherished values, such as justice, equality, and freedom. This dissertation investigates how the predictive power of algorithms and AI threatens freedom. It neither counsels the wholesale rejection of algorithms and AI nor overlooks their potential benefits. Instead, it (1) assesses how negative, positive, and republican conceptions of liberty have been invoked in prevailing discussions of algorithms and AI; (2) advances interpretations of Hannah Arendt and Charles Taylor on science, technology, and the exercise of freedom to clarify the political impact of algorithmic prediction; and (3) applies these theoretical insights in two empirical case studies that employ predictive tools at planetary and personal scales—the Intergovernmental Panel on Climate Change (IPCC) and the Quantified Self (QS) movement.
Traditional approaches to freedom, the negative, positive, and republican conceptions of liberty, help us understand the consequences of algorithms when they malfunction or are used to facilitate political interference, inequalities, and unaccountable decision-making. However, they inadequately address the full impact of predictive technologies in our political and personal lives. Arendt’s theory of political action and Taylor’s concept of self-interpretation provide crucial supplements, revealing the political and personal impacts of predictive technologies and the prospects for addressing them. An investigation of the IPCC and QS movement demonstrates that predictive technologies are neither inherently oppressive nor inevitably liberating. Members within this institution and movement illustrate how predictive technologies can fruitfully coexist with democratic and reflective practices. In this manner, algorithms and AI can be guided by wisdom. And human freedom, with all of its uncertainties, can endure.