Member-only story
The Ethics Of Autonomous Vehicles
Why we need a human-centric approach to make artificial intelligence acceptable
There’s been an enormous amount of discussion over the last decade regarding how best to implement software to manage autonomous vehicles (hereafter, for the sake of brevity, AVs). Ever-more-obscure edge cases are imagined in order to show that the ethics of AVs requires extremely deep and complex cogitation lest we release onto the roads the specter of non-human killing machines.
An example of an edge case is this: imagine the AV is confronted with a vehicle coming toward it in the wrong lane, driven by a young mother with two unsecured children in the back. Swerving into oncoming traffic would likely result in the AV’s passenger suffering injury and perhaps even death in addition to causing injury and perhaps death to whoever is in the vehicle that, unlike the car being driven by our hypothetical distracted young mother, is in its correct lane and therefore blameless. Meanwhile, if the AV chooses to mount the sidewalk in order to avoid impact with the rapidly approaching mother-driven vehicle, it will hit and almost certainly kill two elderly people who would be in the AV’s direct path. The dilemma is illustrated below:
