Will my car be programmed to kill me?

Self-driving cars are one of the first products employing artificial intelligence that will be available for mass consuming. Whether, as our usual cars driving themselves, or as Uber or Google pods coming to our doors to pick us up and drop us wherever we tell them to, the future is presented without a steering wheel.

All this rises one very important question: how will cars behave upon an accident? What will happen when the artificial intelligence faces a sure human death?

This difficult issue is now being tackled by automakers, and the sad answer is that there is no right answer for this ethical debate.

Let’s give an example about a typical risk situation a self-driving car is going to face. Two children suddenly jump into the road.  The car is carrying one passenger, and its AI module can choose between two priorities: hitting the child or swerving and crashing another upcoming car. What criteria will it use to decide?

If the car’s brain is only guided by quantitative data, swerving would be the right decision, because killing two people is worse than killing just one (2>1).

Problem solved? If cars are to be programmed this way, who is buying a car that is not programmed to protect its passengers, but to kill them if necessary?

And if self-driving cars don’t reach success because of this ethical debate, are societies sticking to the use of our old-fashioned, human-controlled cars?

Some experts agree on the fact that the ethical debate might not be on how we program autonomous vehicles, but on how much longer we are using our dangerous, susception of human distractions, manual cars.

As you see, this question doesn’t seem to have a proper answer just yet. Please leave your comment down below on what you think an autonomous car should do in the event of  an accident.

What we can be sure about is that philosophers will be soon working aside computer engineers around Silicon Valley.