Since people are talking about self-driving cars in the previous Linky Friday thread, I thought I would share a post I wrote from a couple years ago that explores a potential future and the tradeoffs:
I have a hypothetical scenario and would be interested in your thoughts.
Imagine that a company creates something that they call the Antcar. The Antcar is a self-driving car where you basically put a destination into a GPS device and it will drive you there. By and large, it is not meant to be driven. There may be controls (like a steering wheel and breaks) for an emergency, but it’s all pretty limited. The entire point is that it will take you where you want to go.
The first question is… would you want such a car? Would you trade the ability to drive for the ability to do other things while the car drives itself? Would you pay $12,000 more for the privilege? $8,000?
The second question, though, is the real question. Let’s say the Antcar (and its competitors) work very well. They’ve been on the market for a while and now the premium is down to $8,000 per car. The initial hopes that it would cut down on automobile accidents was premature. They don’t always play well with human drivers. The Antcar manufacturers, meanwhile, start a pilot program in an eastern European city and then a few cities where Drivercars are banned and only Antcars are allowed on their roads. The results are astonishing in terms of safety improvements. Traffic engineers in the United States start proposing that the US consider doing the same.
Below are some factors to consider. Feel free to point off if I am way off-base about something, but for the sake of this hypothetical accept what I say as true. I’m less interested in how you think Antcars would really work and more interested in how you would evaluate the tradeoffs.
Pro: Safety! Car-against-car accidents reduce by 90%. Cars running off the road reduce by 98%. Pedestrian accidents reduce by only about 10%, but the victims of remaining accidents are caused by pedestrian error. Accidents where pedestrians follow the correct traffic signals are reduced by 98% (the remaining 2% are Antcar malfunctions). No drive drivers, no exhausted drivers, no reckless drivers results in considerable safety improvements.
Neither: Operation costs are roughly the same. The taxes to account for increased costs of road maintenance are offset by much lower insurance premiums.
Con: The cars are more expensive. In today’s dollars, you can add about $8,000 to the cost of any given car. However, the increased safety means that driving smaller cars becomes more possible. So while a 4-door economy car might cost $20,000, you can get a 2-seat Smart-size car for $15,000 and a one-seat bucket car for about $12,000. Nobody would have to buy the Antcar right away because there would be a ten-year transition period, but you would have to factor these higher prices into future cars purchased.
Pro: A more productive populace. People can (sorta) work in their cars. I say “sorta” because they can’t lay papers out everywhere or anything cause the car would be turning and breaking and even if there were some signals to alert the driver, it could be kind of tough in a lot of circumstances. They can unwind during the drive rather than when they get home. Cell phone calls are now guilt-free.
Pro: The makers of the (capital-A) Antcar enjoy a market advantage but not an absolute one. They’re willing to submit to standards so that their cars can cooperate with cars made by other companies. They already do somewhat, but they understand that the standards are going to become much more rigid. In other words, they would not have a monopolistic advantage.
Con: The government would have to administer these standards. A cynical person would point out that they may not always necessarily do so with the public interest in mind.
Con: Everywhere an individual driver goes becomes a matter of record. Law enforcement and courts can subpoena it the same way that they can subpoena phone records. A drug dealer is arrested and theoretically they can look through the records and see everybody that’s visited that house or street in the last thirty days or longer or whatever the records say. Divorce proceedings could unmask precisely where the husband or wife has been. And so on. Definite loss of privacy.
Pro: The ability to investigate where people have been would help the police solve crimes. It could also help innocent people establish alibis.
Con: No more driving. No more getting a turbo-engine car that you can rev up. These cars would still be available, but you really couldn’t drive them anywhere. The engines in the Antcar (and its competitors) would be pretty standard in terms of capabilities. Having a muscle car would not be nearly as advantageous since the cars would be navigating in a more cooperative manner.
Pro: Significantly reduced traffic times. No more accidents means no more accidents causing delays. No red lights at intersections where nobody is coming. Lane merges because much less painful. Eventually it will get to the point that traffic lanes themselves are no longer necessary, though the antcars are not yet ready for that.
Con: Riding in inclement weather can become difficult or impossible in some circumstances. These things are directed by satellite so things that disrupt a satellite signal would make the car not work. When signals are lost and are cutting in and out, the car can let you direct it (you tell it to turn right ahead and then you tell it when to turn left and so on), but it’s a real hassle.
So… what do you think? Here are some options, though I’d like you to elaborate if you have any further thoughts.
a) I would absolutely support banning drivercars. Safety is a premium consideration. Not just the lives saved, but the freedom from fear on being on the road after 2am would absolutely make it worth it.
b) I would probably support banning drivercars. I’m concerned about some aspect of it or another, though.
c) I couldn’t support banning drivercars on libertarian grounds. People should never lose the freedom to drive (and conceal where they’ve been) even if it results in the loss of life and a significant reduction of accidents.
d) I can’t support it because I don’t trust our government to play fair with standards and not play favorites.
e) It’s hard to answer your question because you didn’t explore what I would consider to be a significant factor and/or your prediction on some aspect or another is so far off-base I can’t suspend my disbelief that far.
f) What’s an Antcar? I’ve never heard of that. I don’t think this technology exists. I also don’t understand the meaning of the word “hypothetical.”
g) I refuse to recognize your hypothetical because that’s not how it would happen and I would much prefer to litigate that issue rather than put any thought into the post you wrote.
While G of course has good potential just be cranky my answer would split B and C down the middle. I don’t think i could say people shouldn’t ever drive or have that ability. Partially because it would force a lot of people to buy new cars even with the phase in/out period you state. But the increase in safety would be enormous and especially valuable for certain subgroups. The handicapped and elderly would likely get immense benefits from the mobility the Antcar would provide. It could be virtually life changing for some. Serious alcoholics and addicts would benefit, as would the rest of us, if they had a safe ride home. I could see making it law that after a 2nd or 3rd DUI you are only allowed to “operate” an Antcar. I can rebuilding our traffic infrastructure around Antcar’s as they grow in popularity. Being able to easily talk or read or eat or whatever while driving could be real easy to get used to.
I don’t see why there couldn’t be a Tor like app for privacy or the ability to wipe records although there could likely be limits to that given the Sat connection.
I wonder about parking. I’m picturing various trail heads where parking can be tricky or unclear. I’d think people might have to takeover for certain kinds of parking.
Something to keep in mind regarding the expense: With antcars, owning a car would not be the necessity that it is now. You could have one drive to your doorstep where needed.
Keep in mind, self-driving cars are available with or without the antcar-only model. So the advantages for the elderly and whatnot would still be there, either way. You just wouldn’t get the safety improvements.
Having rentalbe Antcar’s would indeed be great either for people with temporary but serious injuries/illnesses or for people who don’t need a full time car Heck rental car companies would love them since it would cut on people treating the cars like crap.
I can imagine there would optimizer programs to plug in all your errands and have it figure out the quickest, cheapest route sort like UPS has.
Hmmm… currently driverless prototypes like the GoogleCar operate in auto mode with a human driver always at the ready to take over if the autopilot goes screwy. Just turn that around and you could have the car allow manual operation with the autopilot looking over your shoulder, ready to take over if you operate outside a certain envelope.
Given computers can land planes in many situations i think the bigger challenge is improved mapping and getting autopilots to deal with multiple cars, walkers, bikes, etc.
Oooh, I hadn’t thought about bikes. That would really throw a wrench in things as I do not think we will ever have antbikes.
I think given an increase in computer and sensor capability Antcar will be able to deal with bikes, walkers, dogs, moose, etc. It seems like doing the first 90-95% of the Antcar is easy, its the last bunch of details that will take time to work out.
Or we just eliminate bikes.
And dogs.
What you describe is a certainty. It will happen. Probably soon.
The next step after that will be the common adoption of the GoogleCar or something like it.
Step three is the car that doesn’t need a passenger in it. It can come and pick you up. Or go into a drive-through and pick up your groceries for you. I’m not positive this will happen, but I think it will. The benefits are just too great to ignore, once the technology is there and sufficiently reliable.
It’s step four I wonder about. All of the above will reduce accidents, but you could reduce it hugely if you took out the wildcard of human drivers. On the other hand, it’s here I think the freedom argument will win out to some degree or another. I am conflicted as to whether this is a good thing or a bad thing.
I don’t see personal cars going out to pick up groceries. A car designed to carry humans would be overkill for this purpose, and a waste of gas. It makes more sense to have a miniature vehicle designed specifically for this purpose. Or for the store send out a car to deliver to many homes. Really, I’m not sure why this isn’t more common now. Probably because people can stop on the way home from work instead of making a separate trip.
Some cities have a grocery delivery service, but I was always skeptical of it. If I ordered bananas, were they going to choose the same bananas I would? Probably not. Assholes.
We all tend to look at things through the lens of our own experience. Let me share mine…
I believe that commercial vehicles will eventually prove to be the real driving force for adoption of this tech. Consider:
1. Whereas the average car owner puts something like 12-14k on their vehicle in a year, I do close to that in a month.
2. As a long-haul driver, I spend perhaps 80-90% of my time on the Interstate system and other limited access highways.
3. Those highways are precisely the roads that are most amenable to operating a driverless vehicle. Everybody’s going the same direction and you don’t have to deal with nearly as much crap like stop lights, pedestrians, slow vehicles, etc.
4. I’m subject to hours-of-service rules limiting the number of hours I can drive each day. This also puts a limit on the ability to utilize some very expensive capital equipment. (A semi tractor runs about $100k and goes up from there. Trailers are in the $20k – $40k range depending on type.
The argument is essentially economic. Automated vehicles, even if they still required an on-board “operator” (there’s a lot that I do that goes beyond driving — safety inspections, loading, unloading, fueling, etc.) could result in some serious efficiency improvements. It would require robotic systems good enough that I could set the “super-cruise” and take a nap, but it wouldn’t necessarily require them to be good enough to drive the rig anywhere and everywhere on secondary roads and streets.
Road Scholar, I presume you’re our long-gone Ramblin’ Rod?
I hope you’ve been well, for I haven’t seen you commenting in some time. Welcome back.
One and the same, my Good Lady. And I’ve been fine, thank-you kindly. Unfortunately, my laptop can’t say the same and so while I can follow the League fairly well on my phone posting a comment with it is enough of a hassle that I don’t do it much.
You should check out MD. We’ve been doing some recreational math you might enjoy.
I know, Mike. I love that kind of stuff.
As someone who is sadly math-phobic (seriously, and my life is LITERALLY poorer for it), I have to say that “recreational math” is a nearly-inexplicable phrase to me.
Kind of like “good superhero movie”.
I don’t think that drivecars will be completely banned. Too many people enjoy driving. What I see happening is more stringent requirements for getting a driver’s license similar to that of a truck license.
I would foresee that drivecar enthusiasts would be able to rent time on a track to pursue their antiquated, dangerous hobby.
Like go-carting for adults?
Basically, but you can drive a Ferrari.
This is a thing now, by the way. I used to have a coworker who had a sports car that he would take to a track, where he would pay to drive around on it at speeds that would be illegal and (more) dangerous on a public road.
It’s pretty much the equivalent of those places where you can rent a fully-automatic firearm and shoot it at a range.
I’m going to add my own option: Tax the externalities via insurance. And none of this $300,000 coverage nonsense. If you want to drive, you have to insure yourself against the actual cost of any damage you might do, including noneconomic losses in the millions if there’s a body count or a severely debilitating injury.
Actually, I’m not sure why we don’t do this now. I guess it’s a reciprocal thing. We’re willing to tolerate sharing the road with drivers who have $50 thousand insurance policies because we don’t want to pay for $20 million insurance policies ourselves.
On second thought, that’s kind of a cop-out, if it turns out that even small numbers of human drivers on the road dramatically reduce the safety of the robot drivers. If that’s the case, I’m willing to go with the ban on human driving on public roads. It’s not inherently more illiberal than a ban on speeding or reckless driving.
My only reservation about the Antcar is “Everywhere an individual driver goes becomes a matter of record. Law enforcement and courts can subpoena it the same way that they can subpoena phone records.”
Although I’m not entirely certain that bothers me all that much either — when you’re out in public, people can see you and you have no reasonable expectation of privacy. You can’t complain if someone takes a photograph of your car and that fixes your location at a particular time. I’m not entirely sure how an Antcar GPS register would be different than that, although obviously there would be a lot more detail and thus it’s creepier.
The extra expense would be mitigated by the much lower insurance premiums. Big insurers would almost leap at the chance to insure an Antcar, with such a vastly diminished risk of collision. My suspicion is that compared to the lifetime of a status quo drivecar, the Antcar’s lower insurance premiums would outweigh the increase in purchase price.
If Antcars were mandated, I’d struggle with the privacy issue. If they were optional, it’d still be somewhat of an issue, but less so. So I guess I go with C.
Now, a few caveats. I don’t particularly like driving. If I could take a train or bus or plain somewhere and read or otherwise be productive, I like it. However, I struggle being in the car with other drivers, unless I really, really trust them, especially if I’m in the front seat. I think it is a control/fear issue. I get uneasy, especially if they drive like crazy Californians and break at the last possible second every goddamn time (Zazzy!). So, I would have to learn to trust the Antcar. But, I probably could get there, especially given the safety record.
Also, regarding your point on working in the car, I assume you wrote this several years ago because… what the hell are papers? Screens, bro.
There is a really bad anime OAV called Ex-Driver about driverless cars. The basic plot is that in a future of driverless cars, the government employs people who know how to drive when something goes wrong with the system. Interesting plot mired by poor execution.
Have you ever read the Asimov story The Feeling of Power?
My feeling is that if the tech is still at “can’t function in bad weather”, the tech is not ready for banning human driving just yet.
I would absolutely trade the driving experience for other things I could do with that time. I love traveling, even the pure transportation part of it, but I really don’t get any actual enjoyment out of the act of driving. But I don’t know how much of a premium I’d pay. I’m a used-car buyer, and hate the idea of spending hundreds a month on car payments. The more I drove, the higher a premium I’d pay, but my commute is only a mile (and I only drive it because of taking kids to school and picking them up), and once my kids are grown we won’t be driving them around to swim meets, etc. So as much as I love driverless cars, I’m unlikely to get one unless it’s an aftermarket kit I can put on a used car.
But I would oppose banning driver cars, at least in the short term, because it would impose excessive costs on the lower-middle and lower classes. It’s the same reason I dislike mandating new auto safety improvements on a short time schedule. Experience has shown that auto safety improvements are readily adopted by the market because consumers who can afford it want the safety. Then it works its way down to the middle and lower classes through 2 mechanisms: used cars and diminishing cost of the technology. Mandating it too quickly can just mean that the poorer folks are priced out of the safety market, and end up driving even less safe cars than they otherwise might be able to get, a perverse outcome.
I say stay out of the way (other than for regulators to keep any eye out for unexpected problems that may arise) and let it develop naturally.
And I agree with those who said to price insurance for people who choose to drive at a proper level; although that, too, should not happen too quickly, since, again, it’s likely to be the poorer folks who will be able to adopt the driverless technology last. It would be a really vicious public policy that says “you must use this technology you can’t afford, and if you don’t we’ll hit you with an insurance cost you also can’t afford.”
I like driving sometimes, but if I was car shopping and there was a driverless car (that could not be driven manually) that cost roughly the same as a “traditional” car, I’d buy the driverless car in a heartbeat, even if it meant never driving manually again (which as I said, I sometimes enjoy, but just as frequently find a hassle). Think of all the extra free time to work, and read, and eat/drink, and sleep, and make phone calls and answer e-mails and comment on blog posts…
That’s it exactly. I have better uses for my time than watching for potholes and people changing lanes without signaling.
I would gladly give up driving, given the option.
I might remark that there are millions of drivers who can’t afford a car that costs $8,000. I can’t see adopting any of the policies that effectively price the poor out of ability to own and use a car.
Car ownership would be less required in an age of antcars. Companies Like Zipcars would take off.
Yeah, I think that rent-by-the-hour becomes significant in the long term, although for other reasons. You have to wonder, though– like taxis today, will there be parts of town that shared antcars would consider off limits at certain times of the day? When I am on my rural-vs-urban energy soapbox, I regularly note that many of the components of an energy-efficient transportation network that will work just fine in urban/suburban areas (eg, a combination of light rail and rent-by-the-hour) don’t work at all in many rural settings.
Which is why we shouldn’t have rural settings.
In rural areas you can get away with shooting cats.
But not fighting dogs… see Vick, Michael.
I’ll make a point to have the next 80-meter wind turbine installed in your backyard, then :^)
You’ll have to fight with the developers that want to build condos back there.
From what I understand, what they got Michael Vick on was not “animal cruelty” but “running an illegal interstate gambling operation”.
It was the animal cruelty that made the situation so lurid.
If we fix the urban (and the dangnab surburban areas), we can afford the rural. At least, in reasonably thought out ways (Big Trips! Go to town once a month!)
Here’s how the Antcar will evolve. It will be pretty much exactly the same evolutionary path as the automobile itself. Consider the early autos out there among the pandemonium of horse and carriage traffic. We’ve all heard of those hilarious laws where motorists had to put a walker with a red flag out in front of the automobile, all sorts of complaints about how they’d scare the horses.
Same with the Antcar. At first, self-steering vehicles will require special designations, precisely because they’re never going to play well with everyone else on the road. But out on the interstate highways, Antcars will be in their element. Antcars are just one step up from power steering, automatic transmissions and cruise control.
All robots have handles which fit a human grip. Every one of them. And they always will. They’re tools.
But out on the interstate highways, Antcars will be in their element.
Having just done a couple of long drives on rural interstates, I’m strongly in favor of this. And Anttrucks (or would that be antrucks?) especially. With software that keeps one truck from taking 10 minutes to pass another truck, at least in situations where there’s other vehicles around.
Where yall gonna be when the machines “decide” that to protect us from ourselves, they need to go all “i robot” (the movie) on us?
I’ll keep my driver car. That way i’m not limited to the speed someone else thinks I should go..for my own good. And if I run over your dog or your grandmother, well, she shouldhave looked left twice. 🙂
Part of the evolution that I picture would involve Antcar only lanes, similar to carpool lanes. They would have significantly higher speed limits, enabling drivers to get places much faster. That would incentivize the driving of Antcars. Over time, as more people drive Antcars, the availability of manual driver lanes would shrink.
Then, we get some sort of disruptive technology that enables travel in a manner to make the discussion moot, and Antcar and driver car manufacturers both work to legislate it out of existence.
Designated lanes aren’t necessary, strictly speaking. The Antcar won’t smarten up all at once: it will be a gradual thing. The implications are already in view: the GPS guidance systems will meet up with these new-fangled (and unstandardised) distance detection and braking thingummies such as we see on Mercedes-Benz cars now. Standards will emerge, soon enough. Pretty much all the moving robots I’ve seen have them. It’s coming into focus: if it becomes an open standard, we’ll see it sooner rather than later. It won’t be too much different than our interfaces to anti-lock braking and skid avoidance stuff. We sorta forget how many subsystems are already in place between our perceived and actual control of our vehicles.
I do not see the Antcar lanes as necessary. I see them as a way to encourage Antcar ownership. The Antcars might be able to drive on the freeway with regular cars, but they have to obey the speed limit because of the other traffic. In the Antcar only lane, they are able to open up, because they do not have to compensate for erratic behavior of manually driven cars.
Designated Lanes, such as the diamond HOV lanes, are nothing but a pain in the ass. The Antcar must evolve responsibly. If I were in charge of implementing the Antcar on the Interstate highway system, I’d start with a series of RFID slugs in the roadbed. “Ah” says the Antcar, “I’m now in an Antcar-compatible roadway. I can integrate lots of information into my guidance system on this basis. Weather reporting indicates I’m on an icy road and should take this into account.” Lots of other considerations might come into play — but the RFID slug can at least keep the Antcar in the lane.