Not so long ago, the idea of a car that could (0) DRIVE itself belonged firmly to the pages of science fiction. Today, prototypes are already being tested on public roads, and several manufacturers expect fully autonomous vehicles to be commercially available within the decade. Alongside the obvious technical challenges, however, a set of difficult ethical questions has begun to the attention of policymakers.
Imagine, for instance, a situation in which a collision is genuinely unavoidable. Should the vehicle attempt to protect its passengers at costs, or should it prioritise pedestrians who happen to be nearby? How an algorithm resolves such dilemmas profound implications for the responsibility borne by manufacturers, owners and regulators alike.
Surveys suggest that ordinary drivers struggle to up their minds. Most respondents say they would prefer cars programmed to minimise overall harm, yet the same respondents also say they would be reluctant to buy a vehicle that could, in theory, sacrifice them. This apparent contradiction to the heart of one of the central puzzles of modern ethics.
The legal landscape is no less complicated. Traditional road laws it for granted that a human driver is ultimately in control. Once that assumption is removed, questions of liability become strikingly difficult to answer. Several governments are already up specialist committees to advise them on how existing rules should be adapted.
None of these issues is likely to be overnight. What seems certain, however, is that the march of autonomous vehicles will force us to confront moral questions that the motor industry has, until now, been able to avoid.