Purpose of Self-Driving Cars
The purpose of the self-driving car is that with technology, you remove the factor of human decision making, which means that you no longer have human error. The car is able to calculate all possible decisions in order to get out of an accident, and execute on that decision within a fraction of a second, whereas it will take a human much longer. Therefore, the car is supposed to be much safer in terms of an accident. However, what is supposed to happen in terms of a situation such as one like the Trolley Problem?
Summary of the Trolley Problem
The Trolley Problem comes from the Philosopher Phillippa Foot, and it conjures up a scenario where there is a trolley on a track and it is heading towards a switch in the tracks. At that switch, there is one person on one side of the tracks, and 5 people on the other side of the tracks. The question is: which side of the track should the trolley go to?
One might say, kill one and save five.
But then a question is posed: What if the one person is your family member, and the five people were random strangers. Then, the choice is a lot harder to make--then the decision to save five and kill one may not be so clear.
The Trolley Problem is also extended to a scenario in a hospital--where five sick patients come in, and one healthy patient comes in. The doctor can kill the one healthy patient, harvest the organs, and use them to save the lives of the five sick patients.
Here, it is clear that killing one to save five is not acceptable.
One might say, kill one and save five.
But then a question is posed: What if the one person is your family member, and the five people were random strangers. Then, the choice is a lot harder to make--then the decision to save five and kill one may not be so clear.
The Trolley Problem is also extended to a scenario in a hospital--where five sick patients come in, and one healthy patient comes in. The doctor can kill the one healthy patient, harvest the organs, and use them to save the lives of the five sick patients.
Here, it is clear that killing one to save five is not acceptable.
The Problem with the Trolley Problem and Self-Driving Cars
The Trolley Problem has a wide variety of scenarios that can be applied to it--for instance, what if the five people on the track are track workers who were aware of the risks, and the one person on the track was an innocent civilian, or it can be extended to other scenarios such as the doctor who is harvesting organs. This is why self-driving cars and the trolley problem are controversial issues.
Human beings can make conscious moral decisions based off of emotion and logic at the time an incident is occuring. Computers can not do that--computers do not have a moral compass and are not able to calculate the risks like a human can in terms of philosophical questions or questions of moral behavior. This means that whoever is programming a self-driving car, will have to think about the Trolley Problem ahead of time and all of the possible scenarios and decide ahead of time who gets to live and who gets to die. However, the car at the time of an incident will not know if the people on the "tracks" are a relative of the person, if they are a track workers, or any other situation that may be at play in accordance to the trolley problem--most likely, the car will be programmed in a Utilitarian form--which is to maximize the total amount of goodness in the world, and minimize the total amount of suffering.
Human beings can make conscious moral decisions based off of emotion and logic at the time an incident is occuring. Computers can not do that--computers do not have a moral compass and are not able to calculate the risks like a human can in terms of philosophical questions or questions of moral behavior. This means that whoever is programming a self-driving car, will have to think about the Trolley Problem ahead of time and all of the possible scenarios and decide ahead of time who gets to live and who gets to die. However, the car at the time of an incident will not know if the people on the "tracks" are a relative of the person, if they are a track workers, or any other situation that may be at play in accordance to the trolley problem--most likely, the car will be programmed in a Utilitarian form--which is to maximize the total amount of goodness in the world, and minimize the total amount of suffering.
Critiques of Programming Self-Driving Cars to Deal with the Trolley Problem
As mentioned above, a car will not be able to know who the people are in front of it, it will simply know the numbers and therefore work in a Utilitarian view in order to maximize pleasure and minimize pain in the world. If this is the case, this means that even if we were placed in the situation mentioned above about the hospital and the five sick patients and the one healthy patient, the Utilitarian point of view would justify killing the one healthy patient in order to save the five sick patients. This is not okay, because as a society, we cannot fear for our lives every time we walk into the doctors office, thinking our organs may be harvested in order to save sick people.