Greater levels of autonomy are making their way into contemporary vehicular technology, and research on multiple fronts is being done to engineer fully autonomous vehicles that will eventually be capable of safely executing parking-to-parking tasks for their owners. The large number of self-driving vehicles expected to ply the roads of the future raises the possibility that a vehicle driving in autonomous mode may be faced with a mechanical or environmental issue that will have it face a situation of unavoidable harm, and force it to choose among human lives to spare or to expose to a lower risk. The complex interplay of factors to consider in such a situation are staggering, including those of fairness, liability, responsibility, utilitarianism, intervention, survivability, and the social roles of potential victims, as well as the differing relationships of each of the humans to the vehicle. The way these factors will be dealt with will influence the public’s trust and their reaction to this technology. In order to achieve this trust, we need a multidisciplinary approach in which scientists, engineers, lawmakers, and car manufacturers exchange ideas, and work together to propose potential solutions.
To promote such approach, MIT Media Lab, Berkman-Klein Center, and Institute For Advanced Study in Toulouse (IAST) organized a symposium on trust and ethics of autonomous vehicles at MIT Media Lab. The workshop aims at initiating new collaborative projects through interactions, and at creating potential mechanisms for influencing regulations and creation of industry standards.