I’ve been interested in driverless cars ever since I saw Uber’s prototype, Johnny Cab, in the original Total Recall when I was 10. It’s a fascinating thought experiment to sit down and extrapolate the implications of mainstream adoption of this technology. Let’s take a dive down the rabbit hole and see what a future with driverless cars might look like.
Without a doubt, the advent of mainstream automated vehicles will usher in a new era of safety and leisure. Automated vehicles will crash less often–as they never tire, blink, or sneeze–resulting in an increase in safety for passengers and pedestrians alike.
Automated vehicles won’t need steering wheels or pedals and will instead be designed around the comfort and safety of passengers. This will lead to much more than additional legroom as much of a vehicle is designed around the drivers need not only to control the vehicle but to maintain awareness of her surroundings through windows and mirrors. Both the interior and the exterior of an automated vehicle can be redesigned without visibility in mind.
The possibilities for new vehicle designs are nearly endless but it isn’t all sunshine and rainbows. Automated vehicles come at a price that isn’t strictly monetary. They introduce as many issues as they do possibilities. Let’s delve into what the seedier side of our future with automated vehicles might look like.
Manual Cars on Automated Roads
Most people with a driver’s license who have spent any amount of time on the highway are familiar with the jerk in the sporty car that weaves in and out of traffic at an absurd speed. This individual is intent on getting wherever they need to go a few minutes faster, safety be damned. Not only do they endanger their own lives, but all the lives in all the vehicles they careen past.
Perhaps you can relate. Ever been driving along, minding your own business, checking your mirrors like you’re supposed to and still be startled when a vehicle zooms by you unexpectedly? When that happens to me I can’t help but jerk the wheel a little to the right or left, a defensive maneuver, like shying away from a raised fist or throwing your hands up when a ball is flying toward your face. This reaction of mine could cause an accident, either by taking me out of my lane or by causing the vehicle in the lane I’m shying into to initiate their own jerky motioned collision avoidance.
Automated vehicles won’t get startled like this. They will, however, be defensive, perhaps even submissive, drivers. Imagine a two-lane highway with an automated vehicle in the left-lane and a manual vehicle in the right lane. The two vehicles are driving side by side. There is another car in front of the manual vehicle in the right lane, is dutifully obeying the speed limit. The manual vehicle can’t pass this vehicle with the left hand lane because the automated vehicle is there, and it is also dutifully obeying the speed limit. What if the manual vehicle simply shifts into the left lane, the one with the automated vehicle? Due to it’s programming the automated vehicle either slows down or shifts out of its lane and into the breakdown lane, avoiding collision with the manual vehicle.
The automated vehicle has no idea this was a deliberate action on the manual vehicles part. The manual driver could be asleep at the wheel or not have noticed the automated vehicle. It has no choice but to react in a defensive manner to protect all the lives involved.
It’s not a big stretch to imagine young adults making a game out of this and seeing how far they are able to push the automated vehicles. This could give way to bugs being found in the way the system handles collision avoidance, bugs which could cause the automated vehicle to act in an unpredictable and unsafe manner. Perhaps under certain circumstances, it hits the brakes and comes to a dead stop on a busy highway. Perhaps the bug results in the car taking too sharp a swerve and flipping over. What if this two-lane highway runs along a sheer cliff?
Automated vehicles that aren’t modded, will not be good getaway cars in the future. However, manual vehicles that can bully automated vehicles, may make for excellent getaway cars and be quite difficult for police to deal with.
Imagine a road filled with automated vehicles, and a single manual vehicle comes careening down the center of the road, not obeying any lanes. All of the automated vehicles might just… get out of the way. Dan Wells explores this idea in his novel Bluescreen:
The first lane of cars saw them coming, the swarm intelligence registering their presence and passing it along to the cars behind. Trajectories were calculated and courses were corrected, and the cars moved to avoid the girls before the passengers even knew anyone was there. Marisa ran along the edge of the freeway, trying to catch up to Anja’s position, gripping Sahara’s hand as the giant metal monsters rushed past, buffeting them with wind and noise. She saw each vehicle’s passengers in a strobe-like slide show, smiling and laughing, oblivious to anything out of the ordinary. Their headlights caught the desperate girls, lighting them up in an almost subliminal flicker of leg and face and glittering minidress, but by the time the speeding passengers’ eyes had relayed the information, and their brains had processed the sight and its deadly implications, the cars would already be half a mile down the freeway, restored to their place in the lane and the danger now safely behind them. For Marisa, stuck in the middle, the danger seemed to blot out the entire world, leaving her blind and disoriented.
As did Eliot Peper in Culumus:
Lilly took the south on-ramp and merged. The Fleet cars heading in the same direction automatically adjusted their speed and spacing to give the Land Rover a wide berth. She snorted. The algorithms governing their movements had little trust in fallible human drivers. Only the extremely vocal minority of muscle-car enthusiasts and their pet lobbyists had kept human-operated vehicles legal through their Right-to-Drive campaigns.
On the bright side, police and ambulances will be able to trigger masses of vehicles to get out of their way with emergency systems, since the automated vehicles will all be networked.
It’s fact, not science fiction, that a fully automated vehicle will have to make decisions that could purposefully kill the passengers inside the vehicle. Imagine an automated vehicle with four passengers traveling down a road. Ahead of the vehicle is a parade or gathering of some sort with numerous people lining the street. The vehicle attempts to brake and realizes it’s brakes are not functioning (or that it is currently on a patch of ice, or oil or some other slippery substance). It assesses it’s options and the return value from its assessment is only two possibilities. It can careen forward, crashing through dozens of people, more than likely killing many, or it can swerve and crash itself into a cement wall, more than likely killing everyone inside the vehicle.
What does it do? This debate is more than theoretical. This very thing is being discussed in the labs that are developing the automated vehicles that will someday be a large part of our society. If you think about this situation for a while you may come to the conclusion that these machines should put the greater good first. Killing the passengers, while terrible, is better than killing dozens of different innocent people in a crowd.
Let’s extrapolate though. Are car manufacturers going to advertise that your new (and expensive) automated car may decide to sacrifice you to save a group of strangers? Doubtful. Will lawmakers implement a greater good law for the governance of automated vehicles? In first world countries, this seems likely. Perhaps it will be something that comes out of a terrible accident that causes public outcry, an accident that would have been avoided if the greater good law had been in effect.
Let’s assume this law exists and extrapolate further. You are now driven around in a vehicle that at any moment, might kill you. The chances are slim but still greater than dying in a plane crash. Most everyone accepts the new greater good law, but what about the folks that do not?
In first world countries, car manufacturers will not be given a choice. They implement the greater good law in the decision-making functions of automated vehicles they manufacture for sale in countries with the law. Some car companies may take the moral high ground and make all of their vehicles employ the law as a best practice, regardless of if it’s being sold in the first world or third world. Some will not.
Companies are always looking for a competitive edge. Telling car buyers a vehicle will prioritize lives above all else, well, that sounds like a hell of a competitive advantage. It also sounds like a luxury that a company could charge extra for. This allows the rich an opportunity to pay a premium to protect their own lives. An opportunity that the middle and lower classes simply won’t have access to.
These cars won’t be available in first world countries though, right? Well, automated vehicles are governed by operating systems. Operating systems that can and will be hacked or modded to support after-market, perhaps even illegal, software and hardware. This software will enable the owner to prioritize the lives within the car. It would be quite cyberpunk of me to point to rich corporate salarymen and say that these are the only people one might expect to procure these kinds of illegal, greater good ignoring, modifications to their vehicles. But what about the mothers and fathers that can’t accept that their children aren’t the most valuable and important things in the world– it would be hard to begrudge them wanting the safest possible vehicle.
Laws Rule Everything Around Me
Automated vehicles will at some point be required to abide by all traffic laws. This will involve obeying police directions to pull over. The police, trust them or not, will have the ability to stop your vehicle cold for whatever reason they want.
If your vehicle is impounded, they will not need a tow truck. They will simply remove you from the vehicle, override the default control on your vehicle, and tell it to drive to the nearest impound. If you park on a street that has construction or street cleaning, something many folks from big cities can relate to, your car won’t be towed. It will be told to move itself or impound itself if it’s unable to find a suitable place to park.
Since automated vehicles will be networked, it would be possible for the police to cordon off and evacuate the streets of a particular area, by rerouting all automated traffic out of that area.
Beware the Repo-code
Gone will be the days a repo person would show up at your home or place of business and stealthily take back the banks car because you failed to make your car payment.
Repossession of a vehicle upon failure to make payments will be an automated process that doesn’t involve humans. The vehicle will wait until it’s in an area that it can safely drive out of (IE: not a garage) and when you get out, it will drive away on it’s own, with the destination being some corporate reclamation center. Indeed, we have already taken a step in this direction with GPS transponders and networked kill switches in some high-risk car loan vehicles.
Hack The Planet
Hackers find ways to do things every single day that they aren’t supposed to do. Remotely hacking a vehicle and stealing it without ever leaving their desk will become a reality. Theft is just one of the mischievous things a hacker might do with a networked, fully automated vehicle.
Vehicles As Weapons
We’ve seen vehicles used as weapons and weapon delivery systems before. A modified automated vehicle could be programmed in such a way as to crash itself in a manner that would result in a large loss of life.
An automated vehicle with explosives on board could be directed, probably without (or with just minor) modifications, to drive to a certain destination and wait.
These could be powerful weapons in the wrong hands. Huge pains will need to be taken to insulate automated vehicles with firewalls and fail-safes. However, it only takes one bug or exploit to cause a disaster. High-profile luxury vehicles will certainly have their due diligence done with regards to security, but what about low-end bargain vehicles?
Good, Or Bad?
Automated vehicles will save lives at the cost of control, privacy, and freedom. They will solve many of the issues we have with our current transit systems, such as accidents and traffic. Unfortunately, they will also create a new slew of problems that will need to be addressed with laws, security systems, and a change in the mindset of how we view vehicles.
Even with the potential for hacking and the loss of privacy and freedom that come with utilizing an automated vehicle, they could be a good thing. Only time will tell how well we integrate this new transportation method into our societies, and to what level we rise in order to ensure that the best practices these vehicles operate under ensure the protection of ‘the greater good’ and not just corporate greed.
If you’d like to delve into dystopian visions that have dealt with our coming driverless dystopia, check out some of the following:
Total Recall / I, Robot / Minority Report / Cumulus / Bluescreen
Some of the links included in this article are Amazon affiliate links. If you would like to purchase these items, consider using the links provided and help support Neon Dystopia.
Thank you for this piece. However, I might have to disagree with you in terms of self-driving vehicles only being limited to two choices in terms of dealing with problematic situations. In fact, I had just recently written an article for the U.S. Transhumanist Party on this very subject.
You can read it here: http://transhumanist-party.org/2017/06/12/the-trolley-problem-and-self-driving-vehicles-article-by-b-j-murphy/
Let me know your thoughts on it. Would love to discuss this more.
Was a fan of your podcast and have enjoyed both of your articles on here so far. Hope to see more of your stuff on here in the future.
I have an issue with the “kill the passengers or kill pedestrians” scenario in that the problem give is unrealistic. These automated systems have internal and external sensors that would allow it to regulate it’s actions unlike a human driver. If the brakes failed, the automated car would know before applying them via sensors that detect brake pressure, wear, etc., unlike a human who has to pump the brakes to notice a failure. Once detected, the car would cut acceleration and drift to a stop and notify police or a tow before even getting into a situation like that. Same goes with ice patches or water: those are detectable via infrared and thermal imagers, so much that road crews use them as non-destructive testing now. An automated car would sure to have an IR sensor array, along with LIDAR, ultrasonic, radar, and/or visual sensors for driving. Also, programmed behaviors would help with avoiding scenarios: human drivers forget road safety rules or justify breaking them to make for a quicker commute, as long as they’re no law enforcement to remind them. An automated car would have all that stuff programmed into it’s code to be kept legal, so an automated vehicle WOULD reduce speed under weather conditions, entering certain roads, around stopped cars, in populated areas, etc. It comes down to that an automated vehicle would basically never enter the whole “trolley problem” scenario to begin with, because it’s going to be a much more attentive and careful driver than a human.
That said, you missed the elephant in the room when it comes to automated vehicles and that’s automated interstate commerce. Uber’s end goal is largely intimately private public transit, where instead of being in a bus or train with two dozen other people, you can be alone or with a small group of friends or family. The big change, and it’s coming, is replacing truckers, which is the most common job in many states and is one of the highest paying careers for someone who doesn’t have a college education that’s not in mining or oil. And replacing truckers also effects the entire ecosystem underneath them that serves them. Everything from roadside diners, truck stops, and motels to illicit economies like prostitution and meth dealing is going to be vaporize, much like the ghost towns built around railroad water stops after the introduction of the diesel engine, or those not lucky to have a highway routed near it when the federal highway system expanded.
Isaac L. Wheeler (Veritas)
You are on point with your observations.
Your observations are valid, and they give rise to a point where people typically freak out. Knowing that the theoretical situation in the original article is still technically possible (what if the vehicle were electronically sabotaged, or a jagged rock cut the brake lines JUST before the crowd emerged?) even if hundreds of thousands of times less likely, it will be cited as “evidence” that these new vehicles can’t be trusted. Luddites will always fight our advance into the future.
I have a background in nuclear reactor operation, and so I am often faced with a situation where I have to explain that just because something exists, is detectable or is technically possible, does not mean that it is a legitimate threat or concern.