By Lance Eliot, AI Trends Insider
Human drivers are opting to bully nearby AI-based autonomous self-driving cars, according to increasing reports. Actions being taken by these overbearing drivers include cutting off a self-driving car while in traffic, nudging toward a self-driving car to get it to veer away, and otherwise opting to be devious toward these emerging autonomous driving systems.
Some human bullies are doing this for sporting purposes, relishing the thrill of causing the AI to take evasive action. Other bullies act in this manner because they figure all is fair in love and war, as it were, and believe that if the AI is going to be readily intimidated by other drivers than by-gosh it ought to be pushed around at will.
A few bullies seem to simply hate AI that drives a car and therefore perhaps are making these moves to showcase that humans are divine and superior to AI.
Is it legal for these human bully drivers to get away with this practice?
Time to consult the driving regulations. Here in California, the official Department of Motor Vehicles (DMV) Driver Handbook provides prescribed driving practices that everyone is supposed to comply with while driving on our state roadways.
Topics include safe driving methods when the roads are wet from rain (admittedly, we don’t get much rain, but when we do, California drivers are known to freak-out and drive crazily), and driving when there is a tough curve or when on a steep hill, plus what to do when driving nearby animal-drawn vehicles or coming up to railroad tracks. There are over 130 pages of crucial material in our DMV Driver Handbook, which licensed California drivers get tested on and presumably need to understand and are expected to obey (alright, I acknowledge that many don’t, but without those explicitly “you are on notice” regs, I think we’d agree that there might be chaos or at least even worse driving exploits than we already experience).
Furthermore, the booklet is chock full of stern warnings about how you can get a moving violation ticket or get charged with other kinds of criminal infractions, and potentially have your license revoked, if you don’t drive properly as codified in the laws and regulations stipulated in the California Vehicle Code (CVC).
Here’s a question for you to consider: Should the state handbooks on licensed driving include prescribed practices for human drivers to abide by when they are driving nearby to self-driving driverless autonomous cars?
Some say yes, namely that we need to update our myriad of state driving handbooks, and correspondingly each of the state sets of vehicle codes, to include specifics about expectations for human drivers when traffic-wise encountering autonomous cars. The logic being that since there are already explicitly stated expectations about driving nearby to motorcyclists, nearby to bicyclists, nearby to horse-drawn carriages, there ought to be a similar inclusion about driving nearby to driverless cars.
Those that say no, meaning that they don’t see a need to include driving practices depicting autonomous cars as a special roadway consideration for human drivers, argue that an autonomous car should be treated like any other car being driven on our streets and highways. In essence, they insist that the existing driving regulations are sufficient and that human drivers are to drive in the same manner that they drive when around other human drivers; thus, in this viewpoint, it makes no difference whether a car happens to be driven by a human or an AI system. Treat them all the same, they would contend.
For my framework about AI autonomous cars, see the link here: https://aitrends.com/ai-insider/framework-ai-self-driving-driverless-cars-big-picture/
Why this is a moonshot effort, see my explanation here: https://aitrends.com/ai-insider/self-driving-car-mother-ai-projects-moonshot/
For more about the levels as a type of Richter scale, see my discussion here: https://aitrends.com/ai-insider/richter-scale-levels-self-driving-cars/
For the argument about bifurcating the levels, see my explanation here: https://aitrends.com/ai-insider/reframing-ai-levels-for-self-driving-cars-bifurcation-of-autonomy/
What The Fuss Is All About
You might be somewhat perplexed on this subject because there are pundits that keep predicting that we’ll have only autonomous cars on our roadways and the human driving act will become as extinct as the dodo bird.
Imagine a world in which all cars are driverless cars. No need to worry about the wild antics of human drivers. No need to license human drivers. No longer pine away about distracted driving, drunk driving, and the other sad and dangerous driving larks that contribute to our annually disheartening car crash death rates and injuries. Instead, AI “drivers” will be whisking us to-and-fro, and presumably communicating with each other, via V2V (vehicle-to-vehicle electronic communications), ensuring that they don’t step on each other’s toes or get into traffic bogs.
Let’s set the record straight on this aspect.
For the foreseeable future, we’ll have a mixture of human-driven cars and AI-driven cars, and it won’t be some overnight transformation that we suddenly have all and only autonomous cars on our roadways. In the United States alone, there are over 250 million conventional cars, and they aren’t going to disappear simply due to the emergence of self-driving cars. Presumably, driverless cars will gradually be introduced, slowly and gradually increasing in numbers, which will likely lead to a gradual decreasing of human-driven cars, all of this taking place over many decades to come.
Also, there is an open question about whether we’ll ever fully get rid of human driving, which some argue that human driving should always be retained for those that wish to drive (claiming it is a right, but I’ll gently point out it is actually considered a privilege and not a right per se), plus concerns that if we don’t allow human driving then humans will become deskilled at the driving task, apparently alluding to the idea that we will become dependent and subservient to AI systems (a conspiracy theory that driverless cars are part of an elaborate AI-takeover plot).
Anyway, let’s, for now, agree that there will be a period during which human drivers and AI self-driving cars will be on our roads together. Will they play nicely with each other?
For AI developers, they are struggling with getting autonomous cars to gauge what human drivers do. There are AI developers that whine about the pesky human drivers and the sloppy and perilous driving tricks that humans employ while at the steering wheel of a car. It’s a lot easier to set up an AI system to drive a car when there aren’t human drivers in the mix of traffic. If we could somehow ban all human driving, it is presumed that we’d already had autonomous cars galore on our streets (or, at least be getting much closer to doing so).
For why remote piloting or operating of self-driving cars is generally eschewed, see my explanation here: https://aitrends.com/ai-insider/remote-piloting-is-a-self-driving-car-crutch/
To be wary of fake news about self-driving cars, see my tips here: https://aitrends.com/ai-insider/ai-fake-news-about-self-driving-cars/
The ethical implications of AI driving systems are significant, see my indication here: http://aitrends.com/selfdrivingcars/ethically-ambiguous-self-driving-cars/
Be aware of the pitfalls of normalization of deviance when it comes to self-driving cars, here’s my call to arms: https://aitrends.com/ai-insider/normalization-of-deviance-endangers-ai-self-driving-cars/
The Exploitation By Human Drivers Of Nearby Driverless Cars
There are two facets of the nature of human drivers that need to get considered in this matter:
- Human drivers driving in their everyday questionable ways, and then perchance are doing so while nearby autonomous cars.
- Human drivers realizing that an autonomous car is nearby, and purposely aiming to maneuver in a manner intended to play with, confound, overrule, or otherwise exploit the driverless car.
Many of the automakers and tech firms that are developing and starting to field public roadway tryouts of their autonomous cars are generally trying to deal with the first facet already. Earlier versions of driverless cars might have been coded in a fashion that the AI assumed there weren’t human drivers around, but the latest versions tend to be aiming to deal with human drivers that are nearby and that are (hopefully) driving in a normal human way, including being a commonplace greedy driver, being an uncaring driver, being a reckless driver, and so on.
The second facet is a bit of a twist on this topic, and one that has not yet gotten much attention.
For the second facet, human drivers are at times changing their driving behavior specifically and explicitly when they encounter an autonomous car. You might liken this to what some human drivers do when they see a novice teenager driving a car that’s marked as a driver instruction vehicle. Believe it or not, some human drivers do rather dastardly things toward such a vehicle.
For example, some “seasoned” drivers try to cut off the novice teenage driver, wanting to see how the neophyte will handle it. I suppose you could say it is a kind of experiment, perhaps driven by curiosity, wanting to gauge the reaction of the novice teenage driver. Other times a weathered driver might zip around the novice teenage driver, doing so because the novice is going no faster than the speed limit, and the zipping past driver has no time to waste and gets exasperated at trailing behind the so-called slowpoke. And so on.
In case you weren’t already aware, human drivers are now starting to do that same kind of “triggered” driving whenever they see that an autonomous car is nearby.
It’s usually pretty easy to spot an autonomous car, either due to the sensors protruding from the car, or sometimes the car has a branding message on the side that tells you it is a driverless car, or it could be that there’s no human in the driver’s seat (there is nearly always so far a human back-up driver in the car, sitting at the driver’s seat, thus, it would appear to be human-driven, though sometimes the back-up driver might be seated in the backseat and be using auxiliary driving controls).
During your first-ever encounter with an autonomous car, doing so when you are driving nearby it, your reaction usually involves a kind of open-mouthed gaping fascination of what you are seeing, but, since these tryouts are typically occurring in the same geofenced areas, over and over again, you eventually get more accustomed to seeing these driverless cars.
Apparently, with familiarity comes a bit of disdain.
Some like to cut off the driverless car and see what happens, akin to the same trickery pulled on a novice teenage driver. Others are annoyed that the self-driving car is abiding by the speed limit and do the zip around maneuver to leave the autonomous car in the dust. Pranks are being played on self-driving cars, such as coming up to a four-way stop, and the human driver seizing the right-of-way by rolling through the stop sign, meanwhile, the driverless car waits patiently because it has been coded to only take its dutiful turn when so permitted (even if it reached the stop sign first).
For more details about ODDs, see my indication at this link here: https://www.aitrends.com/ai-insider/amalgamating-of-operational-design-domains-odds-for-ai-self-driving-cars/
On the topic of off-road self-driving cars, here’s my details elicitation: https://www.aitrends.com/ai-insider/off-roading-as-a-challenging-use-case-for-ai-autonomous-cars/
I’ve urged that there must be a Chief Safety Officer at self-driving car makers, here’s the scoop: https://www.aitrends.com/ai-insider/chief-safety-officers-needed-in-ai-the-case-of-ai-self-driving-cars/
Expect that lawsuits are going to gradually become a significant part of the self-driving car industry, see my explanatory details here: http://aitrends.com/selfdrivingcars/self-driving-car-lawsuits-bonanza-ahead/
How shall we as a society contend with these human drivers that want to prank the AI systems of driverless cars?
Merely telling people to stop their hunger-games efforts is not likely to do much. I think we all realize that human behavior on a societal scale is arduous to adjust and shift.
Here are some more armed options that have been voiced:
- Make it the law.Put into the official driving regulations in each state that any such driving acts by human drivers against or toward an autonomous car are against the law.
By explicitly naming autonomous cars as a kind of protected class, as it were, this might get human drivers to be more accommodating. At least it would allow for clearly having provided a forewarning (otherwise those offending human drivers might say they didn’t know it was wrong to do such acts toward an autonomous car if you buy into that kind of rather a vacuous argument), and make those human drivers legally susceptible to getting a ticket or losing their human driving privileges.
- Catch the scofflaws.Since human drivers know they can get away with these dreadful driving antics and not likely get caught (you can’t have traffic police on every corner), consider using the sensory data collected by autonomous cars as a means to legally pursue the untoward human drivers.
Keep in mind that autonomous cars will have cameras capturing whatever is visually happening around the driverless car. Perhaps the recorded video could be uploaded to your local police station that would then examine it and if a human driver did something wrong, a ticket would be issued, and ultimately after enough violations, their driver’s license would be revoked. This approach though has a lot of privacy concerns and other contentious elements to it, so don’t place your bets on this option for happening anytime soon.
- Divide up our roads.Perhaps it is best to not have human drivers mixing with autonomous cars, and thus we could divide up our roads, having some streets or highways, or particular lanes, declared as for human-driven cars only and others for autonomous cars only.
The notion of dividing the roads is fraught with all kinds of problematic issues. The odds are that the cost to do this would be enormous. Is that really the best way to spend our limited infrastructure dollars? Also, some worry that this is a short-term fix not worthy of investment and that if indeed human driving is going to winnow down, time will solve this conundrum.
- Toughen-up the AI.This viewpoint says that autonomous cars ought to be able to drive in the same way that other humans drive, fighting fire-with-fire, so to speak.
If a human driver is going to be aggressive toward an autonomous car, the driverless car shouldn’t necessarily back down (bowing to the human bully), and instead needs to showcase that it too is willing to play the dog-eat-dog driving game.
This will presumably “teach” human drivers not to mess with autonomous cars, and overtime society will adjust to allowing autonomous cars the same driving “courtesies” as they do among other humans.
Some worry though that if we toughen up the AI, we will have self-driving cars that will start driving in the same untoward manner that humans do. We’ll wake up one morning and find ourselves confronted by “angry” AI systems that dare human drivers, in the same way, that human drivers do, which could indeed be a consequence of Machine Learning, whereby the AI has merely mimicked driving behavior patterns based on the collected driving of how human drivers antagonistically maneuver.
One wonders, would bullying by AI driverless systems get those human drivers that we all already despise to become reborn as civil and polite drivers? If so, maybe it would be worth having slightly aggressive autonomous cars. Or, would we only end-up with lots more car accidents, road rages by human drivers against those in-your-face AI systems, and fail to gain the hoped-for reductions in car crash-related deaths and injuries?
We’re in the midst of placing a new set of “drivers” onto our roadways, and we need to figure out how the existing drivers will deal with these strangers, perhaps welcoming them with open arms or instead exploiting them like they are newbies that deserve a hazing.
Copyright 2020 Dr. Lance Eliot. This content is originally posted on AI Trends.
[Ed. Note: For reader’s interested in Dr. Eliot’s ongoing business analyses about the advent of self-driving cars, see his online Forbes column: https://forbes.com/sites/lanceeliot/]