Autonomous Vehicles Must Be Programmed to Kill
Self-driving cars are indeed a very hot topic right now. Tesla, Ford and other leading car manufacturers are racing to be the first to have their autonomous vehicles on the road. Countries, such as the United States, have even released a handbook of rules on autonomous vehicles. Features such as parallel parking, intelligent cruise control, and automatic lane transitions are already old news. However, when you look at the moral decisions that these cars will have to make, things inevitably become quite complicated.
What Would You Do?
The famous thought experiment, known as the trolley problem, may shed some light on these issues. In this dilemma, a person must picture a railway train racing down a track which has 5 people tied to it. Yet, you have the power to pull a lever which will change the direction of the train and avert it onto another track which only has one person attached to it. The question posed is: are you able to sacrifice the life of one in order to save the other 5?
Inevitably, autonomous vehicles will have the same decision to face. Take, for instance, a hypothetical case where you are sitting in your self-driving car cruising at 50 mph. Suddenly, a group of 5 people begin walking across the street. There are only two possible outcomes; either carry on driving straight into the crowd, or swerve off the road and kill the driver of the car. Not so simple.
Studies were performed at the Toulouse School of Economics in France, which attempted to form a public opinion on this topic by using experimental ethics. One of the approaches they found was to maximise life at all costs. That is, saving the greatest number of people possible. Yet, this thinking will undoubtedly lead to other consequences.
The Future of Driverless Vehicles
It may be very difficult for people to buy an autonomous vehicle knowing that it will sacrifice the life of its owner, based on the number of lives that could be saved. In this manner, more people will continue to own ordinary cars, and the extremely high number of car-accident related deaths in the world today will continue. Truly, it is a Catch-22 situation.
Do we sacrifice the life of one for the life of many? What if the many were truly in the wrong, by directly putting themselves in danger? All I know is that there is no simple answer. Yet, with experts predicting the self-driving market to be in the trillions of dollars in the future, the time to confront these crucial algorithmic morality issues is now.
*Adapted from the MIT Technology Review
This article has been edited by LinkedIn Campus Editor, Jaykishen Gokani. #StudentVoices
Dark by Design ZeroTrust Principal Executioner.
6 年Enter the dragon's den(s) of decades of endless litigation, who is the responsible party?
Digital consultant and business developer
6 年Is the study based on the MIT Moral Machine https://moralmachine.mit.edu/ ?
Chief Technology Officer at Tridentis Advanced Marine Vehicles LLC
6 年When you voluntarily choose to use an autonomous vehicle, you choose the right to be sacrificed. People around the vehicle did not. Makes autonomous people haulers a lot less inviting until the point when all vehicles are autonomous.
Senior Technical Writer for the mining and other industries.
6 年According to "reports" media or otherwise, the autonomous vehicle will be so smart that such a scenario will never arise. The analogy is also poor, as the first situation refers to captive "victims" and the second refers to "stupid people"putting themselves in harm's way. The driver should not be killed by the autonomous vehicle just because others are stupid!
The comparison should be the autonomous vehicle vs. the illogical, panicked human driver. If we had an infinite amount of time to consider the consequences of our possible actions before an accident, we could potentially determine the best course of action with minimal casualties. The reality is far from that. I've been witness to multiple accidents that could have easily been avoided if the human driver didn't panic.