Slashdot Mirror
Many People Still Don't Want To Ride in Self-driving Cars, Survey Finds (cnbc.com)
A lot of people may still have serious reservations about riding in fully autonomous vehicles, a new survey from Gartner indicates. From a report: The Gartner Consumer Trends in Automotive people surveyed about 1,500 people in the United States and Germany from April through May, and found that 55 percent of the people they spoke to would not ride in a fully autonomous car. However, just over 70 percent would ride in a car that was partially autonomous. Gartner defined partially autonomous vehicles as those that could drive autonomously, but allow a driver to retake control of the car if needed. Advocates of autonomous driving have said the technology will actually make driving safer, since statistics indicate human behavior is the major cause of most auto crashes. But many consumers familiar with the tendency of other electronic devices to sometimes malfunction or perform erratically still seem to have trouble accepting the idea of being held in a vehicle that could fail.
It's not just about safety, but expediency and what happens if something goes wrong. Do you want to risk being stuck for hours because there's no driver that can drive you around the branch in the road? Do you want to be delayed to a meeting because it will put safety above all other concerns and stop or slow down whenever in doubt?
Aircraft have well-defined places to take off and land, with very strictly enforced rules. In transit there is nothing like trying to weave through unmarked construction with cyclists and pedestrians trying to cross your path without warning.
There's more to worry about with the craft control itself, but that's where computers excel.
I'd get in a pilotless plane long before I'd be a passenger in a driverless car.
But many consumers familiar with the tendency of other electronic devices to sometimes malfunction or perform erratically still seem to have trouble accepting the idea of being held in a vehicle that could fail.
Don't forget the absolutely valid concern regarding not having control over your destination.
Government decides you're a nasty dissident? Off to the re-education camps with you.
Hacker got a grudge? Time to trap you in your car and charge a ransom.
Psychotic ex? How'd your car end up driving itself into that ravine, anyway?
An enigma, wrapped in a riddle, shrouded in bacon and cheese
The world will change whether people who enjoy driving like it or not. Just like nobody today insists on riding in an assisted elevator those people will have to adapt. Over time the pressures of technology will make human driven cars more expensive to own because of the higher insurance rates, having to add more mechanical steering components etc. Even just individual vehicle ownership may become very expensive because car sharing services will likely become extremely prevalent and efficient. Owning a vehicle outright will make very little economic sense. There will always be a market for manually operated cars but those cars are likely going to get relegated to race tracks and that market will probably be as big as the market for chariots today. And yes, chariots are still a thing and there is a chariot racing track not far from my place. And it's being used daily. But needless to say the people who go there don't ride chariots to the office in the morning.
Your pizza just the way you ought to have it.
At least until we have more data on the safety of self-driving cars.
But I would like to be able to drop myself off at the entrance of wherever I'm going and tell the car to go park itself.
Any sufficiently unpopular but cohesive argument is indistinguishable from trolling.
You'll be totally okay with self-driving cars. We have some of the highest insurance rates in the nation because of the number of old people, functional alcoholics, and drug users behind the wheel. I will totally trust AI over my fellow drivers here.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
I would bet that this is common for new technologies. I remember the early 1990s when a lot of people didn't like the idea of carrying a cell phone. I remember in the 2000s, few people saw the value of smartphones. I knew several people who weren't sure about Netflix streaming, and thought the idea of cord-cutting was absurd. A lot of those people have now cancelled their cable.
Of course people are unsure about self-driving cars. Give it enough time for them to be common, and to have a proven safety record. The results of that survey will change.
Arguing about the safety of fully autonomous vehicles is an exercise is futile theoretics. We know that if they work exactly as we imagine how they should work, it would be safer. But the question is *can* they be perfect? Or *will* they be perfect? Or even *when* will they be perfect?
Arguing about how safe 100% autonomous vehicles are is like debating if a Pegasus can fly faster than Griffin.
How about we stop reporting on how people feel about non-existent/unproven technology and just report the testable advancements in said technology until they're at the point to where the technology is ready for the market... and then report on adoption and experience?
People are starting to realize what the real implications of this are, and they're having the perfectly reasonable, rational, and logical reaction to it: You're either in control of the vehicle that your safety and life depends on, or you are NOT, and if you're not in control, you can't cope with that -- and when you can't reason with, or even communicate in a meaningful way (i.e., talk to/have a conversation with) the machine that your life depends on? Then that's a dealbreaker. I'm sure this technology will be very useful to public safety in the future -- as a sophisticated 'cruise control' feature, and as a fail-safe collision-avoidance system, and perhaps maybe preventing you from going off the road if you fall asleep at the wheel or are otherwise incapacitated suddenly. But people WANT to drive themselves, even if they say they don't sometimes, because we need to be in control of the tool (vehicle), not the other way around. Human nature. So you can forget 'Level 5' autonomous vehicles, no one will accept them in the end.
even want to drive a car with an automatic transmission... don't tallk about self driving !
Ceci n'est pas une Signature !
The primary problem with the statistical evaluation of accidents is that the statistics don't account for the millions of accidents that are avoided each year because of a human driver. These near misses are never recorded.
It only seems to reason that humans will account for the majority of accidents because humans are the primary operator of almost all of the vehicles moving on the roads since the beginning of automotive time. Mechanical failures have been reduced to a reasonable level. Electronic control issues in cars have accounted for numerous deaths and serious injuries. The thing is that most of the time an electronic control system fails a human takes over and prevents the accident that would have been attributed to the control system, thus masking the actual risk values.
I won't ride in an autonomous vehicle unless it is in a closed circuit, highly controlled environment with mechanical safety over-rides.
Take the time to search up what happened when the mechanical safety over-rides were replaced with software/sensor only safeties in electronic devices that delivered radiation therapy. The software/sensor based safeties failed and/or were programmed incorrectly and numerous people died and/or were permanently injured because of it. This never happened when the mechanical safeties were in place and were monitored by humans. They had to go back to that model just to keep from being ruined by lawsuits.
Yes, there are some very nice automation systems and they are managed by humans where human life is at stake.
For fun, sit on your couch with an observer timing your reactions. Have someone randomly enter the room and randomly throw a rock at you while you are watching television or looking over your shoulder talking to the person behind you while your couch autonomously navigates the living room. Determine how much reaction time you need to recognize the risk, initiate action and avoid being hit by the rock. You'll be dead long before your hands and feet ever get back on the controls of the car after the computer yells "oh shit, I can't handle this, the car controls are all yours, good luc........"
When it comes to the feeling (perception) of safety, i'm usually more worried about other people's errors than my own errors. And i'm well aware i'm imperfect either, but the 'type' of collision seriously affects how much damage a human would get, as in - a side impact is way more dangerous than a head-tail collision. I rather be on the safe side and reduce risk, than assume everyone will drive perfectly and according the rules, which is a very flawed assumption.
On-topic. If the car gets confused, and i'm sure sooner or later it will, it's nice to have to ability to take control back. Common wisdom sais not to trust any machine, especially not moving ones. Factories have red emergency buttons everywhere for a reason. Machines can and will fail.
Also, the test and miles driven with autonomous cars do in no way relate to real-world circumstances. Yes, on highways it will work fine. On a rocky unmarked mountain road much less so.
People blindly believing in technology apparently never have seen it fail hard. And computer experts promoting safety of software should not be trusted, actually, should not be allowed to touch any code again the rest of their lives.
A glitch a day keeps the bugs away.
I would trust autonomous driving in a setting if most if not all other cars are autonomous, there are no stray animals or people or weather conditions on the road. In other words the domain is much more constrained.
And in fact I do, think of those trains in airport concourses or the Las Vegas strip.
Two wrongs do not make correct.
I would love to see a mixed mode of self driving cars. When you enter the Interstate the computer takes over. This would make the Interstate much quicker and safer. When you exit the Interstate you have the option of taking back control or using automated mode. Interstate driving in large cities sucks. People are freaking idiots cutting in an out, no turn signals, looky looes., etc. I would gladly trust a computer over the idiot next to me on the Interstate.
> statistics indicate human behavior is the major cause of most auto crashes
Ah, but is OTHER humans that are the cause of accidents, not me. That's why it's better if OTHER humans use self-driving cars, but I'm better off in control because I'm better than OTHER humans.
Most drivers are an inherent risk to everyone else and they shouldn't be driving now. I don't care what they believe, its time to get them off the road and make it safer for all of us. Well, not maybe not quite the time yet, but the sooner as we can accomplish it the better. Autonomous cars will be safer, faster, cheaper and use a lot less road space. And people can text, drink and fall asleep to their hearts content without endangering anybody.
Create a traffic jam by hacking a rental fleet of autonomous vehicles?
Being so lazy sending my car to do errands?
Setting up a camera on my car and let it have its own adventure?
Writing software for my car so that it circles the block until a spot opens up?
The possibilities for fun are endless!
Elon will have to find another way to kill me.
Order from Wal-Mart, or Whole-Amazon Foods, and have an autonomous vehicle deliver my groceries to me.
I don't even trust myself when I'm driving, I constantly monitor all the traffic around me for the impending stupidity of drunk/unlicensed/douche drivers.
Oh... well in that case let's just toss out the whole idea. Don't want people to be uncomfortable or anything. Lord.
I could have sworn we sprayed for Luddites this month. What a bunch of shrinking violets. The future is happening with you or without you.
What I certainly won't do is becoming passenger in a car that is not really self-driving, but totally dependent on permanent online transmissions, hyper-accurate maps (which of course will never really be up-to-date) and other weird external influences.
If the industry can sell me a car that it really autonomous, as in: "Has sensors to look around and that is all it needs", then I'm buying.
Oh, and sure, I totally don't want a car that any half-competent script-kid-hacker can manipulate to drive into the next tree - which is equivalent to "I don't want a car that is "online", ever.
For a long time, i've contemplated my comfort level with riding in an autonomous car. I turned it over and over in my head wondering if i could ever be truly comfortable.
Recently, I found myself riding down the San Francisco highways in a coworker's model X on autopilot. I found i quickly became comfortable with my coworker answering emails while the car navigated the roads. It took me all of 10 minutes to just accept it as normal.
then the car lurched suddenly and veered wildly into the next lane. the driver quickly grabbed the wheel and put it back on course. He gave some explanations about the car losing sight of the vehicle it was tailing and deciding to pick up tailing the car in the next lane. We were totally safe he assured us. Maybe, but i realized something important.
These cars are already all around me. I don't really have a choice anymore on if i trust my safety to them. They are next to me on the highways. They are next to me in the city streets. It doesn't matter if i'm ready or not. I've been taking the risk the whole time they've existed.
.. the people who text and drive were too distracted to answer the survey.
The fact that people are more engrossed by their phones show they do not want to drive, period. Driving is a chore, and they're distracting themselves because really, they want to do anything but drive.
Of course in a lot of places if you don't want to drive, you don't. You simply take public transit and diddle on your phone the whole way
And those who love to drive, won't let the phone interrupt their pleasure. (And I'm sure those that love to drive hate distracted drivers too, because they just take away all the fun when you're dodging them.)
I am wondering how long will it be until police have the ability to remotely shut off cars?... While the cars are underway?
,,, "They turned off my battery!"
Obligatory existing movie reference--Idiocracy: "Why are we slowing down?"
https://www.youtube.com/watch?...
(and it's even an electric car too. Mike Judge is psychic)
Without trying too hard, I can imagine a few instances where an automatic-driving car could end up rolling down the road on its way to (somewhere) with nobody alive in it, or nobody old enough in it to possibly intervene if necessary.
It only stands to reason that the police will need to have a way to instantly and quickly shut the vehicle's motive power off, in some safe fashion.
And while the police having that ability doesn't worry me much, hackers learning how to do it will be able to cause quite a worry.
Any hardware or software process that is quick and simple (that can literally be done at the press of a button) won't be difficult to hack...
You gonna
I do not believe in karma. "Funny"=-6. Do good and forbid evil. Yours, Oft-Offtopic Flamebaiting Troll.
New ideas/developments take hold when the dinosaurs that oppose them die. Our children and grand-children will smile just as we do when we learn about those 19th century proposals to limit the speeds of trains on the grounds that air might be sucked out of them, thus smothering the passengers.
I am not yet convinced. It's not ALL about convenience and how much it protects those outside your car (that is some of it), Its not ALL about looking toward the future, it is ALSO about the safety of who is in my car and how well a car can be trusted along those lines. Until that trust is built, someone else can risk it. We back up our servers, computers and anything else your can think of because..They crash, Become corrupt, get hacked/infected, lightning strike, or what about those damn pixies that cause random explainable things to happen? These cars are imperious to these how?
Not having a backup plan is a ba....
*** A problem has been detected and your car will now shut down to prevent damage.***
A problem has been detected with the following file: left_turn_signal.sys
or maybe this:
**** Your car has encountered a serious kernel problem ****
Your car may become unstable now and might need to be restarted
Self-driving cars are still a long way from being good enough to put into widespread use. I imagine that people are judging based on the state of the tech as it is right now.
If that's the case, then I'm one of them -- I don't want to ride in a self-driving car as they currently exist. But once they get better, I won't have an issue.
And I hope they get better! My dream is that one day I'll be able to avoid owning a car entirely, but still have access to one on demand. Uber and taxis don't do it for me, but self-driving cars just might.
"Soylent Green is made from People!"
(squish)
-- Tigger warning: This post may contain tiggers! --
and I'll take a car that I can drive wherever, then have the thing park itself, then come back when I whistle.
Why the CVT? For one, it got banned from F1 for being "too good". And honestly, if I have a car that could drive itself, I'd let it if I had to get on the road drunk, dead tired, or something. I'd probably swear I'll never do that again, just like you do when hung over, and I certainly would not be comfortable with a car that's so self-driving I couldn't drive it at all.
And there are also a lot of us that just don't care about autonomous cars in the first place. The egos in California have been in overdrive (no pun intended!) inside their bubble for the past decade or so. Reality is a bitch!
Self-driving brings with it many, many forms of disruption that will add up to one thing, year-over-year-over-year reductions in the cost of daily transportation. Everyone not participating will see nothing but increases as their support structure gradually collapses.
The switch to fleets owned by the manufacturers has many potential benefits centered around a reversal in historic incentives to create planned obsolescence and the usage efficiencies that can be gained via fleet operations planning:
This tech is so disruptive that there is no reason why this list can't go to thousands of bullet points. And there is a whole other layer beyond this because developing a tech like this usually advances the science. Then areas that have no relationship will see benefits that are likely to surpass the direct ones.
And, btw, I've heard this same discussion with many other techs. In the end, it will become dominant and people will either adjust and learn to make the best of the new ways or become old, sourpuss grouches sitting around bemoaning how stupid the younger generations are, glorifying the days gone by, and crying for a regression to the way things used to be as the world marches on as it always has.
Re unreliable flesh drivers: rationally I agree with you, but part of me wants to see it road tested by actual riders for a couple of years. There are people who cannot drive for various reasons who want to be able to get around and will accept some risk to do so. These will be the "pioneers" or "guinea-pigs", depending on your perspective. They are doing us all a favor by helping the bot-car co's get the kinks out.
As far as being outright stuck when the software gets confused or crashes, I would hope the car co. will make it easy to fetch a taxi or Uber ride on their dime. Bleep happens. Even with regular cars, I've been 90 minutes or more late to work several times when there were nasty accidents (clogging routes), bad weather, flat tires, dead batteries, etc. If your career depends on arriving on time all the time, then bot-cars are not for you, but many job are not quite that tight. And retirees don't have such worries.
Table-ized A.I.
They can't drive into the sun, can't detect deer in the road out as far as I can with just my eyes, can't drive on snow, can't always drive in the rain, and they most certainly can't stop at the correct address because they can't read signs or residence addresses. How do you solve any single one of those? We're 30 years out minimum from that level of AI.
I'm much more willing to ride in a fully self-driving car, assuming the maker is confident enough to accept full liability, than a semi-autonomous system whose designers think they can rely on a human driver to retake control at a moment's notice. Either the driver is in control or the computer is in control. Sharing is a recipe for disaster.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
... if it is stable and rad(ical) as KITT.
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
I really wish we'd try & learn to walk before we run. If the goal is to increase safety theres much better, less expensive & easer deployed tech but it's ignored because it's not as flashy & the manufactuars can't charge a premium for it.
I would never feel safe with an autonomous car till V2V mesh commutation systems are adopt in mass. It's not that I don't think we can't teach a computer to drive safe in fact we already have. It's the unpredictable meat bags like me that worry me. We humans have a bad habit of doing incredible stupid things at bad times for reasons beyond comprehension. So unless we develop precognitive computers they'll need a way talk to the other cars on the road to deal with us. A V2V system would only make autonomous system work better and easier to deploy.
It's like the stupid NTSB backup cam regulation that set to go into effect next year where all new cars under 10,000 pounds must include one when ultrasonic sensors are better just about every way except the cool factor. There cheeper provide a larger range of detection & unlike a cam that becomes useless if you don't pay attention it's impossible to ignore the loud beeping they make.
Clearly I want to be in an autonomous car that has a chainsaw in the trunk - and maybe a Saiga 12. That should cover most of the videogame reasons why my car wouldn't manage to get me to my destination. Who wants to share the road with any vehicle regardless of how driven that does not prioritize safety?
First I don't trust any software (I'm a software engineer for 22 years now). Especially if they are remotely controllable (which will be the case for sure for these cars).
Second I really like driving.
Will $CURRENT_YEAR be the year of the Linux Desktop?
For me, car travel time is wasted time. Even when I'm stuck in traffic, I still have to be paying attention.
Bring on the automated cars, I say. I'd much rather be able to read a book, take a nap, or do some actual work while in the act of traveling.
Wanting to be use / own a driver-less car is like wanting to fuck with a prosthetic dick. Yeah, I "cannot" wait!
Insurance rates will fix this. They will charge people appropriately for the risk. "You will save X money if you get in this safer car." People will go for the cash and watch tv while they ride along.
"Ruthlessly pursuing the idea that the accordion is just another instrument."
anyway.
between their tailgating, checking their phone at stoplights and while they are driving.
the incredible amount of impatience and just plain idiotic, bad driving you'd think they'd love the opportunity to stay glued to their phone on the way to work.
i for one welcome our auto pilot overlords.
Absolute statements are never true
A car that's almost fully (almost always fully) autonomous will create a driver who cannot drive due to lack of continued experience. Every day it works, perfectly. Then one day there's a construction site with some unusual/complex routing. Car fails. There's no way to prevent all situations where it fails, because reality isn't Disney Land. And now the inexperienced driver has to master the complex situation.
No thanks. Automatic to take over boring cruise on a freeway, keeping the lane and distance - OK. But not anything beyond.
hacking cars and assassinating targets
Many people have negative thoughts about brake by wire systems and that is a relatively simple device compared to autonomous driving.
My prediction is that autonomous driving will be a safety feature that augments normal manual driving. A fully autonomous car will not be accepted by the majority of the population. Also autonomous cars will kill the car industry. What's the point in having all these different car manufacturers when we we are producing basically an appliance. If it's not an extension of our selves who cares what it looks like or how it drives. Perhaps the only thing that would matter are interior design and luxury level. Lame future if your a car enthusiast.
Here are my thoughts about autonomous vehicles... Firstly, they're decades away. Not just because of the technology, but also because of the legislation. There's no way I foresee governments letting those things loose on our roads until the technology is proven to be perfect. How is the tech proven? Look at another country where they are approved. Until one country has the balls to go first it's impossible to get things off the ground. Secondly, I've got a more 'human' concern. A fully autonomous car will need to be able to decide how to act in certain circumstances... like... there's a toddler who has just run into the road. There's not enough time to come to a controlled stop, but the car CAN choose to crash into a wall and likely break the passenger's leg, but NOT kill the toddler. And many people would agree that's a fairly sensible and ethical trade-off. But what happens if that toddler was just a plastic bag floating in the wind that the car mistook for a toddler. With the current sensors used, that could easily happen. So now you've got a broken leg because of a plastic bag and some wind. Then you look at the fine print and it says "We're totally not responsible for your injuries if your car decides to break you legs for no reason other than a plastic bag floating in the wind". Are you going to step into that car again? I just think the technology, specifically the *sensors* and the *central decision-making capabilities* are so far away at this current point in time. As I say above, I think we're talking 30+ years, not 3 years like some people would have you believe. J
And what if there is nothing left of the car "after something goes wrong"?
I'll never get into a self driving car. For that matter, I'll never use Uber, or Lyft. So I guess that shoots your theory down from all angles.
Please redo this survey when at least 50% of the cars drive autonomously.
Loudoun county VA, home to mae east, Washington dulles airport, and top 5 richest counties, has 300 mi of unpaved roads. Folks commute on these roads daily. Most are so narrow that you have to pull over to let opposing traffic pass.
Edge case? Maybe. Useful for sdc? No way.
https://www.washingtonpost.com/local/loudoun-wants-to-keep-its-historic-potholed-gravel-roads-unpaved/2014/02/14/ccdda5e0-94c8-11e3-84e1-27626c5ef5fb_story.html
I’m not sure that we are anywhere near the ability to switch to autonomous cars. The projected cost and maintenance of them seems to me to be prohibitive for the current consumers of cars. Selling an autonomous car to the general public means that they will have to make sure that required diagnostics are done per the car’s schedule, like cleaning the laser and/or camera lenses as per instructions, besides maintenance of tires, brakes, etc. It’s not unlike selling someone an airplane. It may be possible to make a car like that available as a lease vehicle, with strict agreements with the dealership regarding the maintenance agreement. In the end, lack of or avoidance of maintenance will be the undoing of any automated system. I am also concerned about what happens in the event of failure. An engineer may think that they can provide the correct instructions to avoid problems in any situation, but those instructions are only as good as the hardware that they’ll run on. If there is a memory or storage issue that introduces an unforeseen bug, or there is faulty information given to the system by a bad sensor or worse, an almost-working sensor, the human costs could be devastating. I would assume that if the car fails self-check, either before a journey or during it, it will stop until required repairs are completed. There would be no limping it to the shop. You’re getting it towed on a flatbed. I don’t know how I feel about paying for what feels like a personal mass transit device controlled by the vendor. Why not just expand mass transit if the majority of people don’t want to drive?
This precedence may be overruled by grouping expressions between pairs of sparks (') or rabbit-ears (").
Most people can drive themselves, thank you. The self-driving is a marketing gimmick, and till conclusive safety tests are done, I can only assume that it is MORE DANGEROUS to be in a self-driving car than in a human-piloted one
... try asking people if they want to sleep on the way to work and I'm sure more support will be found for fully autonomous cars.