There’s an incredible amount of potential riding on — excuse the pun — driverless cars. As a culmination of significant technologies, you can’t do much better: They incorporate GPS, radar, lidar, camera systems, pathfinding algorithms and more, all into an asset that transports us safely, more efficiently and with less of an impact on the environment.
Granted, a breakthrough this huge isn’t going to happen overnight, and it won’t happen without some growing pains. Below are three major unanswered questions and imperfectly solved problems that need attention before the self-driving car movement rolls out to the mainstream.
1. Meeting Transportation Demands Without Harming The Planet
Ironically, the advent of self-driving cars will probably worsen our influence on climate change before it improves it. It’s not hard to imagine why: Self-driving technology is just one piece of a larger puzzle that also contains electric vehicles and the sharing economy.
With such an explosion of options to choose from, not to mention the unprecedented convenience of traveling in a vehicle we don’t have to pilot or even outright own, our collective demand for transportation will likely only increase once driverless cars arrive in force. The biggest challenge right now is keeping all of these roaming, self-driving and shareable automobiles powered up without putting a huge drain on the nation’s electric grid.
Transportation has been the largest emitter of greenhouse gases in the U.S. since 2016. But until we rethink how our grid is structured and make better progress on rolling out clean power plants, we’re only going to shift the environmental burden to a different kind of locomotion: electricity versus gasoline. By 2050, say, experts, electric self-driving cars will have increased the demand on our electrical grid by up to 26 percent.
2. Empowering Machines To Make Life And Death Decisions
The “trolley problem” is a classic by now. The dilemma goes something like this: A trolley is barreling toward five bystanders and cannot stop in time. You have the choice to divert the trolley onto a secondary path where only one bystander is located. Do you pull the lever or not?
The artificially intelligent algorithm operating from inside the “black box” in a self-driving car probably won’t wring its hands over making this decision. In the grand scheme of things, some people think it isn’t much of a choice at all — you just take the path that results in the smallest amount of harm. It’s already a dark topic, but it gets darker when you think about some of the preemptive laws countries are already coming up with to ensure the autonomous vehicle expansion doesn’t bring any unwanted human baggage with it.
For example, Germany now has laws governing the decision-making capabilities of autonomous vehicles. No vehicle may prioritize one human life over another based on superficial characteristics, for example. And given a choice between preserving property and preserving life, the car must always choose life.
None of this means autonomous cars won’t, after a fashion, have the power to choose to protect one person’s life over another should the worst come to pass. But our job now is to ensure that none of our usual human prejudices make their way into the AI black box — nor our habit of prioritizing material possessions and capital over human life.
3. Sorting Through The Liability Dilemma
One of the biggest unanswered questions about self-driving cars is: What type of insurance will I need to carry once my car can drive itself? It seems straightforward enough, but we’ve found ourselves at a delicate point in this technological rollout where none of the parties involved want to be named liable, even though certain realities apply.
The most significant change we can expect is that liability will shift from drivers to automotive companies. That’s leaving insurance executives understandably worried. Over the short term, experts predict insurance rates will climb for early adopters of driverless cars. Automobiles are getting more and more complex, not to mention pricier to fix — and having them drive themselves only adds to that complexity.
But this trend probably won’t last. At least, not for very long after cars achieve “level 5” autonomy — that is, a degree of autonomy where the human occupant should never be expected to intervene. After sufficient market penetration, insurance rates should drop again. By some estimates, the year 2040 should see about 95 percent of new cars sold with autonomous functionality built right in.
Drivers will still need some comprehensive insurance coverage for fallen tree limbs and the like, but collision insurance is becoming a different beast right before our eyes. When it comes to an accident which a reasonable person would attribute to a miscalculation by the vehicle’s onboard pathfinding equipment, we can expect liability to gradually shift from drivers to the developers of the built-in software.
When a self-driving Uber struck a pedestrian, “questions were raised” about the liability in that particular incident. Uber owned the vehicle and was developing the software controlling it, after all. And yet, multiple lawyers gave multiple accounts about who might be named a defendant, including:
- The vehicle’s operator
- The vehicle’s owner (Uber)
- The vehicle’s manufacturer (Volvo)
It’s an open-ended question right now, although Volvo, in 2015, committed itself to shoulder the responsibility and the liability should one of its self-driving modules result in an accident or fatality. Volvo has since sold 24,000 such modules to Uber for its nascent venture into a self-driving for-hire fleet of cars.
What’s clear is that autonomous cars represent one of the most exciting, democratizing and cost-saving technological advancements to come around in some time. But before they enjoy their time in the mainstream, we’ll all have to put our thinking caps on and figure out how many of our existing legal precedents have to change before we can live in proper harmony with the machines.