Mazda Finds 71% of Europeans Don't Want Full Self-Driving Cars

Just like the computer :)
A computer isn't capable of being distracted, and can multitask to a far greater degree than a human - and on more relevant things. A computer won't be mixing driving with what it's having for dinner, wondering why their spouse hasn't text back and thinking about changing the radio station.
Computers also aren't immune to mistakes all the time :)
If a computer makes what we'd consider a "mistake" it's more down to misinterpretation. A computer can make an error based on processing the wrong kind of data, and it can fail to operate based on some kind of glitch, but it won't really make a "mistake" in the same way a human driver would.

The high-profile fatal accident involving a Tesla on autopilot wasn't a "mistake" by the system, it was a lack of programming allowing it to deal with a particular situation. The recent incident where an autonomous bus was involved in an accident on its first day of operation wasn't a "mistake" - the mistake was made by the driver that crashed into it, whereas the bus lacked the programming that would have allowed it to account for another driver's mistake (reverse out of the way, in this instance).

Notably, these are all areas where AVs can be improved upon, whereas the average human driver is unlikely to change.
So all that is needed is for the person to decrease his distractions as much as possible to at least being able to do excellent on almost all evasions on critical situations on the road :)
This is, to all intents and purposes, impossible.

I consider myself a fairly competent driver and I mitigate potential flaws and distractions to ensure I drive to the best of my own abilities - I never make or take phone calls, or text, or eat when I'm driving, and most of my driving over the last decade or so has been in the arena of either economy or performance driving, both of which put emphasis on things like observation and correct operation of a vehicle.

But in the past and almost certainly in the future I've misjudged situations on the road and been distracted by things. I'm sure similar applies to even the best drivers on the planet, so the chance of the average, disinterested driver getting to a level that they're able to cope with every situation presented to them is overwhelmingly unlikely.

The unfortunate thing is that autonomous cars probably won't entirely skim the bottom layer of worst drivers off the roads, because there are plenty of terrible drivers who will still want to drive themselves.
 
A computer isn't capable of being distracted, and can multitask to a far greater degree than a human - and on more relevant things. A computer won't be mixing driving with what it's having for dinner, wondering why their spouse hasn't text back and thinking about changing the radio station.

If a computer makes what we'd consider a "mistake" it's more down to misinterpretation. A computer can make an error based on processing the wrong kind of data, and it can fail to operate based on some kind of glitch, but it won't really make a "mistake" in the same way a human driver would.

The high-profile fatal accident involving a Tesla on autopilot wasn't a "mistake" by the system, it was a lack of programming allowing it to deal with a particular situation. The recent incident where an autonomous bus was involved in an accident on its first day of operation wasn't a "mistake" - the mistake was made by the driver that crashed into it, whereas the bus lacked the programming that would have allowed it to account for another driver's mistake (reverse out of the way, in this instance).

Notably, these are all areas where AVs can be improved upon, whereas the average human driver is unlikely to change.

This is, to all intents and purposes, impossible.

I consider myself a fairly competent driver and I mitigate potential flaws and distractions to ensure I drive to the best of my own abilities - I never make or take phone calls, or text, or eat when I'm driving, and most of my driving over the last decade or so has been in the arena of either economy or performance driving, both of which put emphasis on things like observation and correct operation of a vehicle.

But in the past and almost certainly in the future I've misjudged situations on the road and been distracted by things. I'm sure similar applies to even the best drivers on the planet, so the chance of the average, disinterested driver getting to a level that they're able to cope with every situation presented to them is overwhelmingly unlikely.

The unfortunate thing is that autonomous cars probably won't entirely skim the bottom layer of worst drivers off the roads, because there are plenty of terrible drivers who will still want to drive themselves.
If that's what we are afraid of in case of human mistakes - the assists like Auto Emergency Brakes - Smart Cruise Control or lane keeping assist- Blind Spots Detector - 360 cameras should do the trick ... no ?
Also you're correct when it comes to pure driving alone.
But shouldn't those mistakes can at least be rid off if the available assists are around ?
 
If that's what we are afraid of in case of human mistakes - the assists like Auto Emergency Brakes - Smart Cruise Control or lane keeping assist- Blind Spots Detector - 360 cameras should do the trick ... no ?
What if the systems are hacked and deaths result because inattentive drivers expected intervention that didn't occur? Or one imagines if a car can keep itself in a lane, it can be tricked to veer out.
 
What if the systems are hacked and deaths result because inattentive drivers expected intervention that didn't occur? Or one imagines if a car can keep itself in a lane, it can be tricked to veer out.
I see where you're getting at.
You initiate it not by itself.
Smart Cruise doesn't work on it's own.
Also i don't fully trust Smart Cruise.
Neither i care that much for Auto emergency Brake.
Those systems should be initiated by us - not by themselves ( except auto emergency brake which i don't care that much about )
So if they fail - you took the risk on initiating it.
 
If that's what we are afraid of in case of human mistakes - the assists like Auto Emergency Brakes - Smart Cruise Control or lane keeping assist- Blind Spots Detector - 360 cameras should do the trick ... no ?
Also you're correct when it comes to pure driving alone.
But shouldn't those mistakes can at least be rid off if the available assists are around ?
They help, but again, they're unlikely to prevent all mistakes.

And eventually you get to a point where so many systems are assisting the car is driving itself anyway.
 
They help, but again, they're unlikely to prevent all mistakes.

And eventually you get to a point where so many systems are assisting the car is driving itself anyway.
But that still doesn't make AVs 100% safe in case of glitches which is why i don't trust it.
I won't take the risk on riding fully automated vehicles.
 
But that still doesn't make AVs 100% safe in case of glitches which is why i don't trust it.
Vehicles driven by people aren't 100% safe--even those capable of intervening.

Back to the subject of hacking. If it were so easy and such a problem, an individual or group with malicious intent could cause unimaginable disruption without even resorting to ending a single life if they simply tricked vehicles into not operating at all. Why don't they do that now since AVs aren't the norm?
 
Vehicles driven by people aren't 100% safe--even those capable of intervening.

Back to the subject of hacking. If it were so easy and such a problem, an individual or group with malicious intent could cause unimaginable disruption without even resorting to ending a single life if they simply tricked vehicles into not operating at all. Why don't they do that now since AVs aren't the norm?
I'm not saying people are 100% safe.
Neither AVs are safe 100%.
If you're aware - you would evade accidents caused by you.
About malicious people - what do you mean by why they don't do it now ?
You make it sound like they never did in the first place.
If you think they never did in the first place - this is based on the available data on the media.
 
About malicious people - what do you mean by why they don't do it now ?
You make it sound like they never did in the first place.
If you think they never did in the first place - this is based on the available data on the media.
All that you've offered so far is:
There was a word which i read somewhere that there was a car assassination which was implemented by an unknown or maybe expected source.
... and I'm still waiting for the evidence.
 
If you think they never did in the first place - this is based on the available data on the media.
Yeah, the media withholds information, but Twitter has been around since 2006 (a certain someone sure makes it seem longer). Surely if hundreds of thousands of cars didn't start one morning, someone would have tweeted about it.
 
Lack of evidence doesn't always mean it didn't happen.
But it does mean it's a claim you can't make until you present the evidence.

In this particular instance you're claiming that you've seen, somewhere, news that someone assassinated someone by hacking their car. All you need to do is show us where that somewhere is.

As someone who's been writing automotive news for the last five years, I'm pretty sure I'd have seen it too... if it had happened. After all, we can't get enough of Tesla autopilot crashes, and "hacked autonomous car kills occupant" would be right in my particular pigpen.
 
Person adamantly opposed to autonomous cars because reasons advocates for auto-driving assists. More at 11.

Currently, I'm without a vehicle, so in the context of using Uber and things of the like, the only issue I can think of for myself is just knowing how bad some of these pre-determined routes can be when people use their GPS. A live driver has the option intuitive about it if they know the general area, where as autonomous vehicles will likely be following a stricter route. Although, that probably is a moot point since it seems most people I get paired with have a hard time even following the verbal directions that are given by their own GPS.

The big (potential) change with autonomous cars is car-to-car communication. We already have apps that take in real-time traffic situations and re-route as necessary. If that were handled by the car's self-driving systems, it'd know of any potential bottlenecks and avoid them.

...

I feel like a natural progression in terms of autonomous cars would be for auto-only lanes on highways. To be honest, the biggest hurdle IMO is that it isn't something that can be gradually introduced. Having autonomous cars sharing road space with other drivers will cause issues for both. Autonomous cars would be capable of making moves that could be deemed dangerous to human drivers (because of things like reaction times — though I suppose programming could minimize that), and humans could easily mess with the folks in auto-driven cars.

Say I see a car in my rearview doing its own driving. The person behind the wheel is about to take a sip of their coffee. Oops, I brake-checked him.

I foresee that happening a lot when/if autonomous cars start showing up in bigger numbers.
 
The big (potential) change with autonomous cars is car-to-car communication. We already have apps that take in real-time traffic situations and re-route as necessary. If that were handled by the car's self-driving systems, it'd know of any potential bottlenecks and avoid them.
Car-to-car communication is something about autonomous vehicles I've considered for some time now. Suppose your car could inform nearby AVs of its intentions, such as lane changes or turns, they could respond by making this maneuver easier to execute without their passengers--or you--experiencing any significant interrption in travel.

As AVs replace cars driven by people, this communication opens up yuuuuge opportunities in lessening the life-suck that is traffic. Let's face it, a significant element of traffic is people putting their own needs ahead of the needs of others. People make maneuvers--regularly without even indicating they're going to do so--and because these maneuvers aren't expected until moments before they happen, traffic flow experiences a hiccup. Even worse, these maneuvers regularly cause collisions that create bottlenecks and hinder traffic flow further.

Not only would AVs make an individual's on-road experiences more relaxing and an opportunity to safely accomplish other tasks, but they would very likely decrease the duration (even if they travel slower).

Say I see a car in my rearview doing its own driving. The person behind the wheel is about to take a sip of their coffee. Oops, I brake-checked him.

I foresee that happening a lot when/if autonomous cars start showing up in bigger numbers.
Yeah, people are dicks.
 
Say I see a car in my rearview doing its own driving. The person behind the wheel is about to take a sip of their coffee. Oops, I brake-checked him.

I foresee that happening a lot when/if autonomous cars start showing up in bigger numbers.

When autonomous cars start showing up in really big numbers, you won't see anyone behind you in your rearview - no windscreen. They'll be facing the other way so that they can talk to the other passengers, and so that they face away from any collision from the front.

Suicide by AV will probably be a thing, and people walk out in front of traffic already to be rude. It'll only be worse with AVs when they know that the car will stop and do nothing unsafe. But AVs may get surprisingly good at telling when they're going to be messed with - especially if they network through a bunch of machine learning for millions of hours of driving experience with other AVs. They'll have seen every trick.

Of course... networking all of the AVs together to learn from each other's experiences (in any fashion) opens them further to software manipulation, so that's a bit of a trade.
 
But it does mean it's a claim you can't make until you present the evidence.

In this particular instance you're claiming that you've seen, somewhere, news that someone assassinated someone by hacking their car. All you need to do is show us where that somewhere is.

As someone who's been writing automotive news for the last five years, I'm pretty sure I'd have seen it too... if it had happened. After all, we can't get enough of Tesla autopilot crashes, and "hacked autonomous car kills occupant" would be right in my particular pigpen.
How can i show something i forgot where i've seen it ?
It's fine to think that it's not trustful without evidence but we all know that there are millions of things which can happen without us knowing about.
That's my point.
I have confidence that it's possible to hack AVs to cause disaster, so it's debatable.
A debatable thing can happen even if the chance is 1% if it's logical enough.
If a certain kind of personal logic doesn't meet others logic that's fine!
I'll make a last statement to share my honest thought about the future which i'm afraid of, but please take it into account that this is a general opinion ( not directed to any individuals ) which could be a little off topic :
Acknowledging AVs is the same as saying humans aren't suited to ride horses because they can cause accidents by killing unnoticed pedestrian by having horses running them over.
Later people in the future will say :
Doctors aren't needed - robots are more direct in operations.
Firemen aren't needed robots can put out fire better.
Writers aren't needed - robots can listen & write better than humans.
Restaurants workers aren't needed - robots will make & take the delivery.
Pilots aren't needed - robots can do better.
Then next day - robots will feed us because we make mistakes when we eat.
What kind of future is that ?
It's like saying humans are bad at almost everything and we aren't suited for working!
I'll never acknowledge that future.
Making mistakes is what makes us evolve since ancient times.
 
Back