Artificial Intelligence and Autonomy: Self-Determination in the Age of Automated Systems

Despite a long tradition of effectiveness in laboratory tests, normative messages have had mixed success in changing behavior in field contexts, with some studies showing boomerang effects. To test a theoretical account of this inconsistency, we conducted a field experiment in which normative messages were used to promote household energy conservation. As predicted, a descriptive normative message detailing average neighborhood usage produced either desirable energy savings or the undesirable boomerang effect, depending on whether households were already consuming at a low or high rate. Also as predicted, adding an injunctive message (conveying social approval or disapproval) eliminated the boomerang effect. The results offer an explanation for the mixed success of persuasive appeals based on social norms and suggest how such appeals should be properly crafted.

In Nudge (2008) Richard Thaler and Cass Sunstein suggested that public policy–makers arrange decision–making contexts in ways to promote behaviour change in the interest of individual citizens as well as that of society. However, in the public sphere and Academia alike widespread discussions have appeared concerning the public acceptability of nudgebased behavioural policy. Thaler and Sunstein's own position is that the anti–nudge position is a literal non–starter, because citizens are always influenced by the decision making context anyway, and nudging is liberty preserving and acceptable if guided by Libertarian Paternalism and Rawls’ publicity principle. A persistent and central tenet in the criticism disputing the acceptability of the approach is that nudging works by manipulating citizens’ choices. In this paper, we argue that both lines of argumentation are seriously flawed. We show how the anti–nudge position is not a literal non–starter due to the responsibilities that accrue on policy–makers by the intentional intervention in citizens’ life, how nudging is not essentially liberty preserving and why the approach is not necessarily acceptable even if satisfying Rawls’ publicity principle. We then use the psychological dual process theory underlying the approach as well as an epistemic transparency criterion identified by Thaler and Sunstein themselves to show that nudging is not necessarily about “manipulation”, nor necessarily about influencing “choice”. The result is a framework identifying four types of nudges that may be used to provide a central component for more nuanced normative considerations as well as a basis for policy recommendations.