Imagine this: You’re driving along, when all of a sudden, your brakes fail and you’re faced with an impossible situation. There are people crossing the road ahead of you. Veering left will cause you to hit an elderly couple while veering right will cause you to hit a mother pushing a stroller. What do you do?
Hopefully, it’s a situation that none of us will ever have to encounter. But researchers at MIT took this question to the next level. With all of the focus that’s been placed on autonomous vehicles, the researchers posed the question: How would a driverless car react if facing the same scenario?
In an effort to add some human logic behind the “thinking” of autonomous vehicles, the researchers created a simulation that created 13 scenarios with similar predicaments. Those who participate in the experiment are asked to play out these 13 scenarios and inform the driverless car of the best decision.
The scenarios swap out the variables, including the elderly, young children, animals, pedestrians and even people who are homeless. But each still poses the same questions: Who should live, and who should die? Which is the “best” of these horrible outcomes?
The experiment began as a localized survey but has since been expanded to other regions to see how the results change based on location.
One result that has caught the attention of researchers is that throughout the United States and Europe, a “utilitarian” approach is preferred. Essentially, those completing the experiment assessed each situation and made their decision based on which outcome would cause the least amount of deaths or injuries – regardless of age, gender and other factors each scenario presented.
Luckily, these types of dilemmas are extremely rare. However, this research does bring up a valid point over the trouble driverless cars could eventually get into. Although it’s a noble sentiment, the truth remains that no one feels comfortable getting into a vehicle that’s been programmed to sacrifice their life if the situation called for it.
For now, turning driverless cars into “moral machines” is still a long way off. Manufacturers are throwing all of their weight behind this emerging technology in an effort to be the first company that officially makes it to market.
We don’t know about you, but we’d rather wait for all of the T’s to be crossed and I’s to be dotted, before taking a ride.
What do you think? Let us know in the comments. If you’d like to participate in this study, and play through each of the 13 scenarios, click here to visit the Moral Machine website at MIT. And, in the meantime, check out this podcast to see just how far driverless cars have come!