The Mercatus Center at George Mason University held a panel discussion on Wednesday, April 27 regarding the ethical, legal, and public policy implications of driverless cars.
One of the main points of discussion was the ethics of “crash optimization.” This refers to the debate around what choice a driverless car will make in an unavoidable crash, and if it might sacrifice its driver to avoid hitting a person or multiple people in the road. For example, if one or more people step out into the street unexpectedly, will the car hit them, possibly killing one or more, or will it swerve and hit a car, pole, or tree on the side of the road, possibly injuring or killing its occupant or occupants?
This ethical dilemma gets even more thorny when you introduce multiple variables, such as the presence of a child or infant in the car, the number of people in the road, the nature of the obstacle that the car would hit, and numerous other possibilities that one could imagine. This is naturally a huge concern; drivers would likely be uncomfortable putting their life in a machine’s hands, even if they are much less likely to get in accident in an autonomous vehicle than if they were driving themselves.
Jesse Kirkpatrick, the Assistant Director of the Institute for Philosophy and Public Policy at George Mason University, said that if driverless cars are programmed to sacrifice the occupant, then most likely they will see a much lower rate of adoption. Kirkpatrick emphasized the need to balance the ethical concerns with the overall usability and safety benefits of self-driving cars.
Adam Thierer, a senior research fellow at the Mercatus Center, pointed out that even with this ethical dilemma, self-driving cars are not capable of driving drunk, distracted, or drowsy as humans can (and often do).
Panelist Tammy Trimble discussed the potential of occupants who are doing “non-driving tasks” to become engrossed in that task; this would make the theory that occupants could take over driving responsibilities in extreme conditions quite dangerous. Also discussed was the multiple ways in which people have been abusing autonomous driving tech, such as drivers taping water bottles to a Mercedes-Benz steering wheel. Mercedes’ autonomous system will only work as long as the drivers’ hands are touching the wheel (or in this case, their water bottle).
Despite the potential for abuse today and the potential for hacking tomorrow, the panelists were optimistic about the future of self-driving cars. Thierer pointed out that no system is unable to be hacked; he also stated that self-driving cars have the potential save families thousands of dollars in operating costs per year (even more if people eventually switch to a car-sharing over car-ownership model).
The panelists also briefly discussed the big “if,” the elephant in the room for those who enjoy driving: will human-driven cars, and indeed car ownership, eventually be banned or heavily restricted? There seemed to be some agreement that it could happen eventually, but not for many years and not without significant resistance.