From driverless dilemmas to more practical commonsense tests for automated vehicles

For the first time in history, automated vehicles (AVs) are being deployed in populated environments. This unprecedented transformation of our everyday lives demands a significant undertaking: endowing complex autonomous systems with ethically acceptable behavior. We outline how one prominent, ethically relevant component of AVs—driving behavior—is inextricably linked to stakeholders in the technical, regulatory, and social spheres of the field. Whereas humans are presumed (rightly or wrongly) to have the “common sense” to behave ethically in new driving situations beyond a standard driving test, AVs do not (and probably should not) enjoy this presumption. We examine, at a high level, how to test the common sense of an AV. We start by reviewing discussions of “driverless dilemmas,” adaptions of the traditional “trolley dilemmas” of philosophy that have sparked discussion on AV ethics but have limited use to the technical and legal spheres. Then, we explain how to substantially change the premises and features of these dilemmas (while preserving their behavioral diagnostic spirit) in order to lay the foundations for a more practical and relevant framework that tests driving common sense as an integral part of road rules testing.

From Driverless Dilemmas to More Practical Common-Sense Tests for Automated Vehicles

For the first time in history, automated vehicles (AVs) are being deployed in populated environments. This unprecedented transformation of our everyday lives demands a significant undertaking: endowing complex autonomous systems with ethically acceptable behavior. We outline how one prominent, ethically-relevant component of AVs—driving behavior—is inextricably linked to stakeholders in the technical, regulatory, and social spheres of the field. Whereas humans are presumed (rightly or wrongly) to have the ‘common sense’ to behave ethically in new driving situations beyond a standard driving test, AVs do not (and probably should not) enjoy this presumption. We examine, at a high level, how to test the common sense of an AV. We start by reviewing discussions of ‘driverless dilemmas’, adaptions of the traditional ‘trolley dilemmas’ of philosophy that have sparked discussion on AV ethics but have limited use to the technical and legal spheres. Then, we explain how to substantially change the premises and features of these dilemmas (while preserving their behavioral diagnostic spirit) in order to lay the foundations for a more practical and relevant framework that tests driving common sense as an integral part of road rules testing.

No Correlation Between Ethical Judgment in Trolley Dilemmas and Vaccine Scenarios for Nurse Specialist Students

We tested whether responses to trolley problems by nurse specialist students correlated with their responses to hypothetical vaccine problems, as a follow-up to a similar study on ethics committees. No statistically significant correlation was found between the trolley and vaccination scores. These results confirmed and strengthened the finding of a very weak correlation (possibly zero), and the point estimate was even lower than for the ethics committees. Hence, the nurse specialists’ responses to the trolley problems cannot be used to indicate any direction for their responses to the vaccine problems, although there is a common core issue of sacrificing some for many. The respondents reported a relatively high willingness to push one man in front of a trolley to save five. They also reported a high willingness to act in trolley dilemmas compared with vaccination dilemmas, although the dimensions of risk–reward ratios and consent heavily favored the latter.

Trolley Dilemmas Fail to Predict Ethical Judgment in a Hypothetical Vaccination Context

We investigated whether the responses of 68 ethics committee members and staff to trolley dilemmas could predict their responses to research ethics problems concerning vaccine trials. Trolley dilemmas deal with the issue of sacrificing some for the benefit of many, which is also a core issue in the vaccination trial dilemmas. The subjects’ responses to trolley dilemmas showed no statistically significant correlation with their responses to our vaccination trial dilemmas. We concluded that, if there is a component of transferable intuition between the contexts, it must be small and dominated by other factors. Furthermore, the willingness to sacrifice some for many was larger in the trolley context, despite a more favorable risk/reward ratio and the voluntary participation of the subjects at risk in the vaccination situations. We conclude that one’s general willingness to trade lives in the trolley context may be an artifact that is due to its unrealistic setting.

An Asymmetric Moral Conformity Effect

The present study investigated whether and to what extent people’s judgments on trolley-type moral dilemmas are subject to conformity pressures. Trolley dilemmas contrast deontological (principled) moral concerns with consequentialist (outcome based) moral reasoning. Subjects were asked to respond to trolley dilemmas in a forced choice format and either simultaneously received bogus information about the base rate of consequentialist and deontological responding for each dilemma or received no distribution information. In the information condition, the bogus distributions showed that either the consequentialist or the deontological choice option was favored by a majority of previous participants. In a set of two independent studies, we showed that subjects exhibit little conformity to a consequentialist majority opinion but strongly conform when confronted with a deontological majority opinion. We suggest this asymmetric conformity effect demonstrates that subjects are less willing to appear consequentialist than deontological, and we explain these results through mutualistic partner choice models.