Will you trust 'driverless' cars?

Yes, I will probably trust the driverless cars. I see no reason not to.

They won’t replace driving cars, I can bet my ass on that.

I would need time to get used to it, though. I don’t trust machines that work on their own with no one to put things back in order when the system malfunctions.

You aren’t going to have any choice.

I trust them more than I trust distracted drivers watching their phones, or elderly drivers who can’t see beyond their hood.

I’ll trust it eventually.

No, but I don’t trust most drivers ether.

The “autopilot” cars that are out now are not driverless. They require a human in the driver’s seat paying attention in case of emergency. The stories going around trying to condemn driverless cars for the few accidents that have taken place always seem to overlook this.

Yes, when driverless cars are perfected and released I will trust them. Like many tasks, driving can be done much better by a computer that never gets tired, distracted, or emotional.

Much more so than the 16 yr new driver who is texting their friends while driving. Anyone texting and driving for that matter.

Our chauffeur is shitting bricks at the prospect…bless his white leather driving gloves.

No.

What are the currently available types, so I can be aware of them when driving?

I have programmed AI’s, and I’ve driven a lot. I consider my fear of AI drivers very well-informed and reasonable.

I think the argument that there are shitty human drivers out there so why not AI’s doesn’t hold much logical water. The problem with AI’s is that they are susceptible to a variety of new types of problems, for examples: Bugs. Malfunctions. Data errors compared to reality. Perception algorithms that fail in different ways than humans do. Situations they will interpret incorrectly. These new types of problems will be less predictable by good human drivers because you can’t use empathy to predict/respond to them. It may be less likely or cause less damage in the long run that having bad/drunk/drugged/mobile-addicted teenage humans driving, but that doesn’t mean it’s also not an entirely new type of problem, and one that will lead to all sorts of other issues (legal, control, privacy, repair, etc).

I like these answers so far. I used to ride a motorcycle for the majority of my life. Usually didn’t have another mode of transportation.

As the years passed, I was almost killed several times by distracted drivers. I sold my last bike about 2 years ago. (My mother was elated. )

I think driverless cars would be safer than distracted idiots.

But, I think it would be better if ALL cars were automated, not just a few. They could synchronize better, if all cars were driverless.

Like so many of my responses I’m going to have to respond with the word “context”.

In the context of, for example, major American, British and other European roads that most of us are familiar with, having a certain quality and style of markings, for example (I’m thinking of the interstate highway system in particular here, though many state highways and local roads have equally well-designed road beds, pavement, markings, turn radii, gradient, line-of-sight, etc.) we count on a certain level of “drivability of the road itself. (On the other hand, if you watch television commercials for new cars – particularly any that are marketed as any kind of off-road vehicle – or tires, for example, they always seem to feature immediate hazards springing up from or falling onto the road: avalanches, washouts, trees falling thither and yon, badly secured tarpaulins blowing off of loads of junk piled into truck beds, etc., and expert drivers with nerves of steel – and the sponsor’s super-duper vehicle and/or tires, swerving without care to avoid the obstacle. The primary thing that might be valuable about self-driving cars in those situations would be their ability to stop quickly. Because they would certainly NOT be designed for high-speed evasive and off-road maneuvering. So in that case I would trust driverless cars – which will stop more frequently when road conditions demand that – than I would trust any hotshot breezily driving over the median, swerving around fallen trees and rocks and plowing through streams “to make schedule”. Not that that sort of thing actually happens much, either; I’m just saying.

So: context. As driverless cars become more and more realistic and prevalent the demand will also increase to improve roads themselves.

GPS systems, nearly ubiquitous now anyway, will become even better than they are.

Control systems, including feedback mechanisms to warn operators (not “drivers”) of pending malfunctions, and, even better, to automatically warn other vehicle operating systems in the vicinity of traffic obstructions and disruptions (like Waze, for those who are familiar with the app, but car-to-car instead of driver-to-driver) will also become more vital. And more ubiquitous.

I will not only grow to “trust” driverless cars, but I enthusiastically look forward to the technological leap in transport that they will engender.

But when I travel to India in the foreseeable future, I will probably still want a car with driver, and preferably an older, experienced driver with a sense of sedateness and purpose, and who is more determined to get home to his family at the end of our trip than he is (because so far I have yet to see a professional woman driver in India, although I have used them in the States) to win some kind of land speed record for our trip.

We have become comfortable with commercial airplanes flying on auto pilot. Yes there are two pilots in the cockpit but the plane is flying itself. So right now we are just in the beginning of transition to driverless cars, and there’s a lot of uncertainty about the potential snags and so on until they get all the problems straightened out.

Will I be comfortable in a driverless car? Not yet but eventually when the system has proven itself.

Probably not in my lifetime.

They will be common within ten years. First for long haul trucking, and then for Uber and the like.

Yes, if they were all that way.

No driverless car for me.

I think they have a long way to go before they become the norm.

I would trust them more readily than one with a drunk driver at the wheel.

What does a driverless car do when it’s on the highway in traffic at 60mph and suddenly a problem occurs (wiring, computer issue, broken/blocked sensors) such that it is effectively blind?

^^^ It stops, same as a car does now.

When was the last time that happened all of a sudden while you were cruising at freeway speeds?

I had a car stop at highway speed. I was in the center lane on a curve.
There was nothing I could do in that situation. Driver or no, the car was dead.

Trucks and cars were rushing up in my rearview mirror honking and barely veering away. I was literally trying to decide if getting smashed from behind by a truck or making a run for it across 3 lanes was a better way to die. Most frightening moment of my life.

Somehow, a guy in a big pickup understood the problem as soon as he saw me, nudged up to my bumper and pushed me to safety. Then he said, “Be careful!” and drove away.
What I found later was a main electrical cable had burned through, cutting all power and thus the engine.

@zenvelo , from my decades in auto insurance I can tell you it happens more than anyone would like to think.

I’ve had various interesting malfunctions and other dangerous situations in cars myself. Many of them electrical, with cascading effects. Often involving onboard computers, or sensors. I don’t expect robot drivers would be reliable to handle many of those situations, and the risk of a robot driver doing something wrong in a way that a human wouldn’t, seems like a new kind of road risk for drivers to face. One I’d much rather not have to face. Driving is already dehumanizing enough with human drivers.

Being able to program a computer to rely on electronic sensors to tell what all the driving conditions are accurately and respond appropriately seems like a very ambitious thing to try to do, and one that will clearly lead to new kinds of failure situations that we are not used to. And personally, I’d rather deal with road situations involving humans than ones with computer drivers relying on sensor data interpretations. Errors with humans have a lot of layers of human thinking behind them as backups against insane moves. Computers use logical rules, math, and digital data to choose what to do, and an error in those could really do just about anything if it breaks in an unexpected way. Like failing to brake, turning hard, swerving, accelerating full speed, failing to notice cliffs or realize they ought to be cautious, etc. Yes humans do stupid crazy shit and sometimes go berserk, but computer drivers will do them in new ways/circumstances, which to me seems to add more types or crazy to look out for.

Also, it’s already possible for a malicious person with a cell phone or lap top to hack into some cars and get them to do deadly maneuvers. The more automated and online these things get, the more there are, the longer they’re out there, the more chances for malicious attacks to control cars and make them into land torpedoes. Possibly even smart land torpedoes. If someone could hack one of these car computers and load up some Carmageddon AI into it (a game where cars run down people and ram other cars for points), for example? Not just accidents but homicidal cars? None of that is even a possibility if we’d just not put computers in control of cars.

Oh, and also, yes you can program a car to stop when its sensor data stops, which may “only” be as dangerous as a car’s driver being unable to see or perceive any of his surroundings and putting on the brakes. But what if the computer itself fails, and doesn’t do a programmed safe stop OR respond appropriately to the situation? (Like the example where the robot car mistook what it saw and swerved and hit someone.)

@Zaku I would still trust it more than I would a car with an elderly driver who confuses the accelerator for the brake, or an impaired driver on a Friday evening after “happy hour”.

Who is at fault when the vehicle is involved in an auto accident??
@zenvelo I totally get where you are coming from now understand where I am coming from my big truck is just under four years old and has over half a million miles on it and the last month alone we have seen 4 computer sensors crap out on it, costing a lot of down time and very costly to fix.
These things have every body thinking oooooooh great but how are they going to stand up in the long haul with say a million miles on them, with what I have seen with the trucks I am exposed to not very well.

I would wonder about the ethics of life or death decisions. Would the car be able to kill the driver to avoid hitting a pedestrian?

@zenvelo But what conclusion follows that? Dangerous human drivers, whether due to senility or drunkenness or incompetence are all not supposed to be on the road. Someone else should be driving them places they need to be driven to.

@RedDeerGuy1 And who will be held responsible when a programmer accidentally introduces a bug that gets into an update the tests don’t catch, causing cars to start turning the wrong way in certain circumstances, causing who knows what damage?

@Zaku So, the question is: Would you trust driverless cars? and your answer is no. You (and @SQUEEKY2) think that because cars fail now, that they will fail more when programmed.

And I will restate that a properly programmed driverless vehicle will be safer all around for everyone than the current state of affairs. And insurance companies recognize that, and since money talks, that’s a sign that they have demonstrated that it will be safer.

About a year ago I read some newspaper letters and articles that some archivist had compiled (no, I can’t link) from around a century ago, about the dangers of so many people using motor carriages/automobiles instead of horse drawn carriages. This all sounds very similar.

@canidmajor This all sounds very similar.

Yep. Horseless carriages were the driverless cars of the 19th century.

There were laws requiring automobiles to be preceded by a pedestrian with a red flag warning the carriage drivers and horses.

@RedDeerGuy1 . Interesting point. Would an automated vehicle run you into a tree,rather than hit a pedestrian?

Would it instantly assess the numbers? i.e. 3 in your car should die rather than 4 people who ran out in front of the car.

@canidmajor . Good point. That’s why I brought up old people who don’t trust elevators.

A horse is quite different from a AI or some computer though. Horses are skidish. They might be more risky than driverless cars.

But I’ve also heard of horses carrying intoxicated people back home ,while the rider was unconscious.

@zenvelo I prefer the type of danger that involves humans driving vehicles, even if some of them shouldn’t be driving, to the new types of danger (and other circumstances/considerations) that will become possible with computer-driven cars, as well as the new types of responsibility situations around them.

For example, I don’t want it to be possible for a car to go on a rampage due to bugs, malfunctions, or malicious hacking. I also don’t want algorithms making life-or-death (or who dies) decisions. As a human driver, I don’t want to have to anticipate new types of behavior by computer drivers. As a programmer, I do not want to be expected to ride in a computer-driven car, because I am all too aware of the types of errors that could theoretically occur. It really freaks me out, just as it would freak me out if there were humans with computer chips in their heads that could theoretically make them go berserk even if they weren’t driving cars. Computers may seem reliable, and be statistically reliable, but they can theoretically malfunction or be reprogrammed in malicious ways. I do not trust corporations and technologists to be wise nor to share my values when programming and connecting such things.

The idea of a car pilot AI virus that makes hundreds of thousands of cars suddenly aim for pedestrians is not that far-fetched an idea, if/when we have a whole fleet of such cars and they’re all on a network and accept updates.

Even without computer drivers, the computers already hooked up to the control systems and to wireless communications are already problems already being exploited. Adding robot drivers would multiply the potential for abuse. Not only could you cause a car to turn or accelerate or fire airbags, but you could cause a car to seek and destroy evading targets, such as actively trying to run down pedestrians, or driving to specific targets and charging at them at top speed, aimed by your nice skillful driving robot.

Examples of just a few current issues:

“Markey cited studies showing hackers can get into the controls of some popular vehicles, “causing them to suddenly accelerate, turn, kill the brakes, activate the horn, control the headlights, and modify the speedometer and gas gauge readings. Additional concerns came from the rise of navigation and other features that record and send location or driving history information.”

“Drivers have come to rely on these new technologies, but unfortunately the automakers haven’t done their part to protect us from cyber attacks or privacy invasions,” Markey said. “Even as we are more connected than ever in our cars and trucks, our technology systems and data security remain largely unprotected.””
– Report: Cars are vulnerable to wireless hacking

- Hackers Remotely Kill a Jeep on the Highway—With Me in It

- Car Hacking Report Refuels Concerns About Michael Hastings Crash

@zenvelo While I will agree that if the vehicle is programmed perfectly it would out preform a human driver.
But these components fail , regularly ,and if the system gets a glitch it could become for more unsafe than a human driver.
You stated at an above post when was the last time you had to do a panic brake at highway speeds,uh lets see when you round a bend and see an over turned RV blocking the road, or a moose standing there, or a rock slide, want me to go on?
Hell I came into this one curve one night and out in the middle of nowhere there was a drunk standing on the centre line waving at me.
And believe me it will never totally take the person out of the cab(notice how I didn’t say driver) because they will still want someone with the load in case of a breakdown.
And how will these wonder trucks do at backing into docks ( you wouldn’t want to try and back your honda into some of the places they expect a driver to back an 83 foot truck into.
The things may indeed be common place on the roads some day but they have a very long way to go.
SURE they may prove right now they can build and and prove it will do what they say, but lets see how they stand up in the long haul, I think the initial cost and the maintenance cost alone will keep them from taking over North American highways for a very long time, but their day may indeed my come but they still have a super long way way to go.
They have proved it can be done, great first hurdle accomplished many more to go.