We May Be Able To Teach Cars To Drive, But Can We Teach Them To Be Ethical?

Can artificial intelligence learn genuine ethics?


Modern technology is always evolving, which can be exciting but also a bit scary. There are many examples of an advancement that did more harm than good, though it was not understood at the time. One new dilemma facing people today is the rise of the driverless car.

What is the Driverless Car?

The driverless car or self-driving car has been a long held dream for people in the car industry. Car companies started working on the technology for these types of vehicles in the 1920s, though there was no real progress until the 50's and 80's. The first self-driving vehicle that was successfully created was in 1984 by the Carnegie Mellon Robotics Institute under the NavLab project.

This type of vehicle uses a number of different sensors like radars, lidar, GPS, odometer, worldwide satellite coverage, and computer vision to navigate the car without worrying about a person doing anything.

Why are Some Leary About Driverless Cars?

The idea of the driverless car is exciting and could do some good, but that's not to say there are no issues to consider. The main problem deals with ethics. Ryan Calo, who is an expert on robotics law and works at the University of Washington is not ready to give the okay on these driverless cars. Calo simply does not think that self-driving cars are sophisticated enough to understand ethical factors. In essence, he is not sure if a driver-less car can make a choice between hitting a car with two grown ups or one with two children.

The other problem that driverless cars have are the pre-programmed ethical choices. It is possible to instill some foreseeable ethical choices into a driverless car, but the problem boils down to who decides what is ethical? Should legislators, lawmakers, people, or the car manufacturer's themselves decide what is ethical?

Some also worry that insurance companies will attempt to make sure that evasive maneuvers favor them, which will cost them less money. This might end up being morally wrong given the situation. For example, should you drive into a grocery cart carrying a child or an empty car? The insurance company will want to avoid the car, but the car's intelligence may not know what is at stake. Sure, listening to insurance companies might provide inexpensive policy costs, but what will be the price?

Are People Paying Attention?

These issues are being played out in courts, in the minds of robotics experts, and in manufacturer's panels, but the people are still in control. Still, most people seem to be in agreement with transitioning into self-driving cars. This is likely due to the possibility that accidents due to human error should significantly be reduced. Liability for accidents should be reduced as well.

The most recent survey regarding support or opposition for self-driving cars has shown that 90% of Americans favor a mixture between self-driving with the option to switch to manual drive if necessary. The survey also showed that three-quarters of the survey-takers are in favor of a fully-automated vehicle without the option to switch. This may indeed be the future of cars. There is no denying that the switch towards driverless vehicles is going to change American life in a significant way. For one, generations of people who know how to drive will start to disappear. Drivers who base their livelihood on driving will have to learn another skill. Laws will have to be changed to accommodate driverless driving.

The question is will it remain a choice to switch to a driverless car, or will it become mandatory at some point? There is a lot that will be asked regarding driverless cars, but these questions have not stopped the majority of people from being excited about the transition. Hopefully, these questions and dilemmas are resolved in a way that affects the future in a positive way.


Read next:

2016 State of Crowdsourced Innovation