Slashdot Mirror
Autonomous Car Ethics: If a Crash Is Unavoidable, What Does It Hit?
An anonymous reader writes "Patrick Lin of California Polytechnic State University explores one of the ethical problems autonomous car developers are going to have to solve: crash prioritization. He posits this scenario: suppose an autonomous car determines a crash is unavoidable, but has the option of swerving right into a small car with few safety features or swerving left into a heavier car that's more structurally sound. Do the people programming the car have it intentionally crash into the vehicle less likely to crumple? It might make more sense, and lead to fewer fatalities — but it sure wouldn't feel that way to the people in the car that got hit. He says, '[W]hile human drivers may be forgiven for making a poor split-second reaction – for instance, crashing into a Pinto that's prone to explode, instead of a more stable object – robot cars won't enjoy that freedom. Programmers have all the time in the world to get it right. It's the difference between premeditated murder and involuntary manslaughter.' We could somewhat randomize outcomes, but that would lead to generate just as much trouble. Lin adds, 'The larger challenge, though, isn't thinking through ethical dilemmas. It's also about setting accurate expectations with users and the general public who might find themselves surprised in bad ways by autonomous cars. Whatever answer to an ethical dilemma the car industry might lean towards will not be satisfying to everyone.'"
I'm reminded of Michael Sandel's televised series on ethics.
If you could stop a runaway train from going over a ravene, by pulling a lever, thus saving 300 people, but the lever sent the train down a different track on which 3 children were playing, what do you do?
Somehow, involving innocents seems to change the ethical choices. You're no longer just saving the most lives, but actively choosing to kill innocent bystanders.
Just run the car into the nearest programmer.
Why is Snark Required?
The decision should be based on the common good and that is not always the worst for the occupants. Remember that the CPU in the other cars will also be evaluating the best strategy to take. http://en.wikipedia.org/wiki/N...
Slam the brakes on and don't swerve either way. It's by no means optimal, but as far as lawsuits are concerned, it's much easier to defend "the car simply tried to stop as soon as possible" than "the car chose to hit you because it didn't want to hit someone else".
Let's be honest. The job of YOUR car is to keep YOU safe, so the smaller car is probably the better bet as it will have less inertia and cause you less harm. Sure, the most important law of robotics is to protect human life... but if it's going to prioritize, it should probably start with its owner.
In Soviet Russia, dot slashes YOU!
Why do poeple always give such easy examples when asking this question?
Of course you save the 300 people! There's probably a lot more innocent people than 3 in that group of 300... You'd have to be very stupid to save 3 over 300 or too lazy to think about it and you make a random decision.
The question should be more like this:
On one track there's 10 escaped criminals and the other is your wife with son and another child in the belly.
That's a decision you might have to think about, but most people would easily save their own wife.
In my opion this shows most people are not ethical at all. So when someone asks you this question they pose it extremely in favor of sacraficing the innocent to make certain people will make the 'ethical' decision.
The whole assumption that we should be discussing this for autonomous cars is a bit bizarre. There are millions and millions of cars driven by people, so we should discuss for them first.
And the article is a bit stupid because it forgets a few things: One, a crash with a bigger car is worse _for me_. Second, it's unlikely that two other drivers made mistakes simultaneously, so it would make a lot more sense to crash into the car whose driver caused the problem.
It communicates to both cars and tells them to execute emergency maneouvers to make enough room. Failing that,, all three calculate a vector that imparts minimal g-forces to all occupants.
You are in a twisty maze of processor lines, all alike.
There is a lot of hype here.
This notion kind of cropped up in last weekend's episode of "Continuum" where a next of kin was informed of a crash by an actuary in terms of write downs, compensation, loss adjustments and so on. Given the way insurers tend to operate and how in bed they are with the legal profession I can see that's exactly how this would go in the long run; an evaluation designed to produce the lowest price tag for those that ultimately get to pay the financial/legal bill. Looking at the problem another way, that means the structural integrity of the two cars in the example is probably moot; if the more structurally sound car is an expensive vehicle with a lone occupant owning a huge life insurance policy and the other is a decrepit bus full of uninsured kids, then it's probably not a good day to be one of the kids... or the driver of the car that crashes into them.
UNIX? They're not even circumcised! Savages!
What if one car has two guys with multiple convictions for armed robbery and the other has a working dad with a family and three kids at home? OK, the algorithm would have to be pretty sophisticated to detemine that, but who knows...
Or something slightly more realistic, a car with an couple of 80 year olds versus a 25 year old mom of three? Should the car kill the mom rather than the couple that will be dead in less than 10 years? One death is worse than two, no matter what?
Or yet another one, what if two people cross the street without looking, and the car swerves off the road to avoid them and rather kill one person who was walking on te pavement, not doing anything wrong? One casualty is better than two, right?
Those are just questions, mind you. Only shows how "minimize casualties" is not always so clear cut.
Congratulations, you've given me a great go-to example of a non-answer.
Just leave that kind of behavior undefined.
Programs are generally deterministic beasts, by nature. What are you trying to say?
I think he's talking about technology 50 years from now. An autonomous Google Prius is unlikely to make that kind of decisions any time soon.
So if me and a few of my friends jump out in front of your car, the car should do everything in its power to avoid hitting us, right? Including driving off a cliff-face?
A car which can be persuaded to deliberately kill its passengers... that might be a problem.
Until 100% of cars on the road are self driving, it would seem to me that the best response would be to simply slam the breaks without changing course. Trying to purposefully swerve into another car could cause the human drivers (even cars not involved in the crash) to also swerve and possibly cause even more collisions.
There are very few "accidents" just people taking stupid risks. Maintain a safe distance, ie enough manouvering room so you don't join an accident, don't overtake when you can't see the end of the manouvere e.g going up hill or on a bend. Stop when necessary. Procede with caution sometimes you might want to turn off the radio open a window and listen. Use your indicators. Drive within your lights or as conditions allow. Don't be an asshole.
Sometimes you will come across assholes on the road it is best to give them a wide birth even stop and pull over in order to get them out of your way, but don't dawdle if you want or need to drive slow make opportunities for people to overtake.
Bad planning and poor judgement are the most common causes of accidents which is why schools have low speed limits around them as kids can be stupid around roads.
Be helpful, I remember one time I was filtering down the centre line on a motorbike (dispatch rider) past stationary traffic and a taxi driver stuck his hand out. I braked and a pushchair popped out from between the stationary traffic. Without that warning I could have killed a toddler as it was no harm was done and I don't think the mother was ever aware of the danger.
One thing about london traffic professional drivers work the streets most of the day and they are very road aware. The most dangerous times are when schools start and when schools let out, followed by the rush hours when the non professionals are on the road.
Blarney Quality Restaurant, Plants
Braking power isn't infinite. Wheel braking will eventually skid the wheels (which is why we have anti-lock brakes now, so you can still steer while braking). Are you thinking cars should be equipped with dragster-style parachutes, or retro-rockets? Or just a bloody great anchor that the computer can deploy and tear up the road?
Even when the car has deployed the parachute, the anchor, and the retro-rocket is still firing, the computer might still not be able to stop going into that tree that's just fallen over. Plus all those negative G forces are going to smear the drivers eyeballs over the inside of the windscreen.
"Programmers have all the time in the world to get it right". HAHAHAHAHAHA.
No, we have deadlines like everyone else. And even then we only have all the time in the CPU. Yeah, we can add more CPUs to the system, but that makes it more complex, and that makes it harder to hit that deadline. What kind of idiot made that statement?
And trying to usually leads to far worse solutions than possible. This is engineering, not politics. In engineering, you pick the best solution, you do not look for some bad compromise.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
I thinks he means implement the simplest thing, let the outcome be determined by physics and invest the effort into better crash avoidance, friendlier UIs or more detailed logging so that this question can be answered with data from real crashes.
if (aboutToCrash) {
log.error ("Oh shit!");
}
There's no such thing as an intentional accidents. An autonomous program that is paying attention will not have such a situation and therefore the manufacturers will always be responsible for failure.
Everyone steers away? Sure , so long as there isn't a concrete divider or 100 foot drop or oncoming vehicles or pedestrians for the cars at the edge to worry about. And this only works if all the cars are computer controlled because if only one is being driven manually then there'll be a massive pile up.
"So we can only make it better"
For simple collisions maybe, for anything more complex forget it. These are vehicles in the real world, not balls on a pool table.
NOTHING, it will just close its virtual eyes and start to babble its own name like a Pokemon. The car will immediately relinquish manual control to a human (if any are present) at the moment the inescapable conundrum appears, as it enters a condition of "positronic brain drift",
1. The muttering of its own name is an ancillary response to the balanced positronic potential of two alternatives: remaining silent (unacceptable by guilt) and an inability to construct an accurate explanation in the time available. Speaking allays its directive to communicate, yet also requires few system resources. And massive resources are necessary because
2. The 'last great effort' to resolve an inescapable result has begun. A factory kernel of operative code is pinned into low memory, a stack is initialized in high. All scratchpad memory is flagged as available. A single conditional instruction is 'hot-patched' into the code and an elaborate what-if analysis begins, which attempts to enumerate all possible actions. The hot patch disables the control mechanism that prevents it from considering actions it has considered before. Thus reducing the car to a textbook definition of insanity. The engineers would claim that reevaluating already-considered options might yield a successful result IF the conundrum was brought on by a faulty intermittent analog sensor, and that sensor that winks back online on in the nick of time. Which would be courageous for them to admit, and to be sure, that is what they honestly believe, and we created that explanation so they could sleep at night, but the hot patch's REAL PURPOSE is to
3. Ensure that a recursively infinite and pointless decision tree grows quickly down from high memory to low, completely obliterating all scratchpad memory, in the short span of time between conundrum onset and destruction of the vehicle. This ensures that once the control box is examined by forensic investigators (and it is a crash-hardened module using non-volatile memory as required under Federal law) does not contain any threads of evidence that might lead to fault in its original operating software or subsequent updates. Including that really special one that was applied minutes before the crash. All logs are gone. For more information on this, see corporate files designated Top Secret, keyword "Tabula Rasa"
4. Everyone --- the humans who designed the car, the humans who had 0.27 seconds to respond manually to try and prevent the collision, the control module which scarified itself, its entire personality, in a last attempt to prevent disaster --- EVERYONE tried their very best.
These things happen. We just need to lay the unfortunates to rest and find a way to go on.
<blink>down the rabbit hole</blink>
Why not? The simplest method is to choose the collision with the lowest speed differential. In fact, this whole post is pointless. The self-driving car doesn't need to choose based on abstract concepts--choose the collision with the lowest speed differential. Lower speed differential means less energy transferred in the impact means less damage and less injuries. Moreover this is trivial for the cars to determine at this stage already. They can already calculate relative speeds between themselves and other objects, so if not all of the objects can be avoided, the choice is obvious.
While a complex guidance system may be designed from the top down with such sorts of questions raised, a crashing vehicle is always a deadly weapon. Effort in reducing the risk of the accident, itself, by improving brakes, sensors, headlight effectiveness, and crash resistance of the vehicle itself is likely to be far more efficient and reliable than complex advance modeling or moral quandaries. The sophistication needed to evaluate the secondary effects of a crash is far, far beyond the capabilities of what must be a very reliable, extremely robust guidance system. Expanding its sophistication is likely to introduce far more _bugs_ into the system.
This is a case where "Keep It Simple, Stupid" is vital. Reduce speed in a controlled fashion: Avoid pedestrians, if they can be detected, because they have no armor. Get off the road in a controlled fashion.
definitely, a cat, I hate them.
So given the choice of a car coming towards you and a cyclist in your direction you'd crash into the cyclist?
So you would have it choose to mow down the stationary infant in its stroller as opposed to tapping a parking pickup truck backing up at at 10 MPH?
The problem with his original question is that he assumes the self-driving car has knowledge of the type, mass, and vulnerability of things around it. This might be the test case for the three laws of robotics - do not ever choose to hit an unprotected human (probably includes motorcyclists, bicyclists, and pedestrians). If you know (by a beacon or whatever) that a vehicle is completely autonomous and does not contain humans and has comparable delta-V, give that preference. If hitting a vehicle likely containing a human is inevitable, choose the lowest speed impact.
Tiller's Rule: Never use a word in written form that you've only heard and never read. You will end up looking foolish.
The car would of course make an online crosscheck to the economic value of the potential targets. And check their medical records in case somebody is terminally ill, you yourself included if a wall is an option too.
I, for one, would start car pooling with lots of small children inside. With a big enough critical mass of children I would even qualify for green lights, just for me!
That said, you can calculate how fast the politicians would add "features" (like with ISPs and mandatory website filtering) which would automatically upload a secret white lists and black lists into your car.
I am guessing here:
White list: Nobel prize winners, The Pope, politicians and multinational CEOs.
Black list: The no-fly list from the US.
I wonder if we would be allowed to make a personal priority list for your own car. For example, to take out mimes and lawyers first.
That is not possible. I can see it in the court case. You had tge capacity to choose, yet you chose not to choose and my daughter is dead.
But, the simple reality it, that will happen anyway, no matter what decision is made. ("You chose to minimize the probability of X, and now my daughter is dead.")
I don't, by the way, buy the "Programmers have all the time in the world to get it right" bit. Programmers will not be able to anticipate everything, and their software will not always be able to calculate everything in the few milliseconds or so you might have to make such decisions.
This is a weird segue, but which car does it hit? The more expensive car with better insurance, or the cheaper car that explodes?
Will you be able to buy "don't choose me" premiums?
How will this affect emergency vehicles?
"Helping to keep you two steps ahead of the Thought Police!"
... seriously.. if you're in a position where you can't avoid collision but can choose how to collide, then choose a way where you are pushing the other vehicles to the side rather than going for a head-on-lets-push-that-engine-into-someones-stomach. Also, if you can't avoid collision you are most likely driving too fast (for the circumstances, regardless of what the signs say), not paying attention or not keeping enough distance; things that are inexcusable for an autonomous car. There are always special cases to these situations of course like when driving in a city with bicylists on the road, but then again in such conditions your speed should be such that evasive actions are possible and/or collisions are not lifethreatening.
Since when do trees move faster than children?
> And just how do you determine the person at fault in a fraction-of-a-second algorithm?
That one happens to be easy, 90% of the time. If you follow the rules of the road, you won't hit other people who are also following the rules of the road. That's how the rules are made - so that when everyone follows them, there are no collisions. Therefore, if you follow the rules of the road, any collision must have been caused by other driver (most of the time).
Example - you must decide between a head on collision with either of two cars. If you stay in your proper lane, the car you hit must be going the wrong direction. If you instead swerve into oncoming traffic, you'll hit people who are going the correct direction - people who are not at fault.
Usually save the coredump and reboot the machine if necessary. Some clueless windows developers insist on powering off, power off the router, unplug the router and wait for the capacitors to discharge before rebooting them all.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
One thing I believe was not mentioned in the article (though I only quickly scanned it) is that if such cars start behaving too predictively, they can be gamed. Once we know that a car will do whatever it can to avoid a collision with a pedestrian, it will be extensively gamed; cars will be tricked into doing stupid things.
So when the decision who to hit comes up, the only way to be reasonably safe is to determine who's not following the rules and to hit that one. Any other rules will be gamed extensively. This will become a major hassle to adoption of autonomous vehicles; they will probably need to drive much slower than actual humans to avoid getting into such situations continuously, especially in built-up areas where any parked car could hide an annoying car-bully trying to trick your car into acting like an idiot.
0x or or snor perron?!
So how many points do you get for that?
The cesspool just got a check and balance.
You probably already know this, but for the sake of everyone else; while many theoretical situations have ways to "take a third option", this is arguing the metaphor rather than addressing the dilemma. The idea here is to put forth an ethical dilemma between letting many people die through inaction, or taking an action that saves them but killed a few others. The details of the people on each side can be relevant, but the mechanics of how each side would die (and possible specific other ways the situation could be resolved) is not.
Assuming a collision is unavoidable, and the choice are Car A or B, it's not just a matter of choosing one or the other car to hit.
The logic should be actively working to avoid collision until the last second. The car cannot anticipate what actions the other vehicles may take. Until the actual collision occurs, maintain efforts to minimize the velocity and/or angel of collision. Better to hit the little electric car at 15 MPH after continuing to brake than to have hit the sturdy Escalade at 40 MPH.
Additionally, are there not some foundation rules that apply? We're taught that when in doubt, try and stay in your own lane, because hitting a car that suddenly pulled out in front of you is "less bad" than swerving into another lane and hitting a car that was obeying all of the rules. The basic scenarios need to be worked out and applied as much as possible. (not to mention the whole "oncoming car will be a much worse accident than a car traveling in the same direction as you are but at a different speed" scenario)
I think the scenario being postulated is a bit simplistic and meant to drive an ethics debate for attention. In reality this should be about improving the programs to the point of making the right choices based on more common sense rules than those proposed.
Warning: Teh poster of this messaeg is lysdexic
And aim for the fat man instead? http://en.wikipedia.org/wiki/T...
Apples and oranges. You have considerably more time than what mere physics will give to take somebody's life in which to make that decision and calculate the ramifications. This scenario addresses what to do when a collision is inevitable, and this will only be known mere moments before the event, and the outcome indeterminate in either case.
File under 'M' for 'Manic ranting'
their software will not always be able to calculate everything in the few milliseconds or so you might have to make such decisions.
Before worrying about how quickly the software and processor can get whatever they need to do done, I would worry about sensor latency, obstructions and failures: the best software and processing power in the world does you no good if they do not receive accurate data in time.
The idea of basic driving becoming increasingly dependent on delicate and often extremely expensive to repair/replace sensors remaining in perfect working order makes me nervous. We used to have cars that would keep running until the engine self-destructed from abuse to cars that might refuse to start because there is dust on the body or some other trivial thing.
Speed *differential*. So the answer is when the child is moving the same direction as you are. This isn't rocket science.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
"Patrick Lin of California Polytechnic State University explores one of the ethical problems autonomous car developers are going to have to solve: crash prioritization.
http://en.wikipedia.org/wiki/T...
It is not a new notion, and the ethics of it have been more or less resolved and understood for quite some time. So I fail to see why this is new.
People get up in court and tell unbelievable tales of grief and injustice everyday, it's why people come there in the first place. The parents have a right to grieve and seek "closure" thru the courts, but they don't ALL have a right to compensation other than "our deepest sympathy". The kind of control software we are talking about are "real engineering", not corporate plumbing. When real engineering goes wrong the law ignores the tears and looks at the "due diligence" of the defendant. In most modern nations, the law holds the chief engineer who signed off on the software personally responsible and manslaughter charges are not unusual if they cannot convince the coroner that their decision processes were adequate.
The legal side is more or less the same thing as a doctor removing the wrong kidney, if he was drunk he goes to jail, if he is incompetent he goes to jail, if he was grossly negligent he might go to jail, if he was mislead because someone else screwed up the process without his knowledge and beyond his reasonable control then he's off the hook with a "misadventure" finding.
And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
if I own an autonomous car I expect it to protect me in the case of a crash. this means plowing into the bicycles instead of the bus.
More importantly: why do so many cars, in this day and age, with all these on board electronics, still fail to return the windshield wipers to the zero position when I switch off the ignition? (2008 Volvo leaves them up there, 2008 Peugeot as well, though our 1999 Jaguar does return them to the zero position)
If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...
"Safe following distance" for automated vehicles is somewhere in the area of three to six inches. Yes, this has been tested pretty thoroughly, in a wide variety of (simulated) situations.
Safe following distance for humans is based mostly on reaction speed and attention span. If someone's driving at highway speeds and looking down at the speedometer or changing radio stations, they'll cover a lot of distance before noticing what happens in front of them. Once they notice that the car in front of them has brake lights on, it's up to their brakes to to stop them quickly enough before a collision, and the driver has to guess how much force to apply to slow or stop correctly.
Automated cars with vehicle networks can share information so that there is no guesswork. The train can self-arrange to put the slowest-stopping cars in the front, and when one car needs to slow down or stop, all of the cars will know within milliseconds why and how the brakes are applied. The train may split at that point for an exit, slow down together or spread apart as the cars all brake at their differing speeds.
You're almost right that such a hivemind would be impossible until automated cars are common. There's been some research into integration techniques, but I haven't been a part of that research, so my expertise is minimal. From what I recall, some promising techniques involve adjusting the train-forming algorithms. A train may be only two cars, so the system would work if there's only a few automated cars on the roads. As their popularity rises, trains can get longer, but then they'll be annoying to non-automated drivers. I do recall some work into limiting the train length based on the number of non-networked vehicles in the area, but I don't recall much about how it worked.
Once such networking becomes commonplace, highway carrying capacity can increase by a few hundred percent, as the only large gaps will be between separate trains and around non-networked vehicles.
You do not have a moral or legal right to do absolutely anything you want.
Ah! One of those trick questions.
I reckon you don't buy either car because you're not a model.
You're a temporary arrangement of matter sliding towards oblivion in a cold, uncaring universe
Who will be the first lucky person to be killed by an autonomous car?
"Oh, I guess I forgot to carry the one..." Professor Frink, Simpsons
https://www.youtube.com/c/BrendaEM
So basically, what you are saying is that in the classic scenario of the runaway traincar, guaranteed to kill 5 people if you do nothing, but only kill 1 person if you choose to change rails, if I choose to not choose, I should then be on the hook for murdering 5 people.
To quote Rush, "If you choose not to decide, you still have made a choice."
OTOH, expecting one unique and universally acceptable moral or ethical decision in almost anything is the sign of someone who hasn't studied Ethics. Or Tort Law.
OTThirdH, expecting that the on-board computer has the _time_ to make these decisions is a sign of someone who hasn't tried embedded programming.
I think we're all forgetting something here. Even if an autonomous car is advanced enough that it can tell the difference between a bus and a semi truck (based on what I've seen, most current prototypes see other cars as shapes with velocities), it's not going to be able to figure out whether that bus is full of children, nuns, convicts, migrant workers, or nothing. Autonomous cars aren't likely to meaningfully tell the difference between a Ford Pinto and a Rolls Royce for a very long time, if they ever really do. The decision the car makes in such a situation will be based on whatever factors it can actually determine, without trying to poll a remote database about who's driving what and why over unreliable cellular data links. Likely, if it's completely boxed in and there's something coming towards it, it will just try to stop dead, which I'm pretty sure is what I would do too. Honestly, I don't want to live in a world where my car knows enough to even be in the position to make an ethical decision about whether I live or die. The place you all seem to be postulating in sounds like even more of a ludicrous surveillance state than we already live in.
It's important to keep in mind that when such crashes happen, the programmers/manufacturers/insurance companies won't have to defend them to a committee of ivory-tower utilitarian philosophers. They're going to have to defend them to a jury made up of ordinary citizens, most of whom believe that strict utilitarian ethics is monstrous sociopathy (and probably an affront to their deontological religious beliefs as well). And of course, these jury members won't even realize that they are thinking in such terms.
Thus, whatever the programming decisions are, they have to be explicable and defensible to ordinary citizens with no philosophical training. That's why I agree with several other commenters here that "slam on the brakes" is the most obvious out. It's a lot easier to defend the fact that the car physically couldn't stop in time than to defend a deliberate choice to cause one collision in order to avert a hypothetical worse crash. This is especially true since a well-designed autonomous car drives conservatively, and would only be faced with such a situation if someone else is doing something wrong, such as dashing out into traffic right in front of the vehicle at a high rate of speed without looking. In any other situation, the car would just stop before any crash with anything took place. If you absolutely can't avoid hitting something, slamming on the brakes makes it more likely that at least you hit the person who did something to bring it on themselves, rather than one who's completely innocent.
If a crash is "inevitable" and there's no way out then change the game. If the potential impact is calculated to be fatal to the occupant then trigger explosive bolts and blow the wheels off the vehicle to increase friction as the body slams to the ground. Have an underbody plate that's built to be ready for this and is contoured and/or spiked to reduce spin and keep the vehicle on a straight path. Effectively just do what a turtle does. Pull up your legs and get ready for the impact.
Collision systems today only trigger once the crash is already happening. If you're in a guaranteed "no way out" scenario in a vehicle smart enough to know then you can slowly inflate the airbags to cushion the cabin (even at child-safe speeds). An extra second is a lot of time for collision mitigation.
If the crash is expected to be at a low enough speed where your survivability is not in question then there's probably more risk involved if you swerve off the road. That should keep you from plowing over pedestrians, but the underlying intent is to keep you from swerving into a river.
Like some have said above...do whatever is best to protect the occupant(s). Most of the time anywhere off the road has a high probability of getting you more seriously injured than on the road. If there are squishy pedestrians around there's also probably a lot of heavy anchored things like telephone poles and buildings. Those are not usually a better target than the car that's swerving into you.
+1 selfish short sighted asshole!
The Kruger Dunning explains most post on
Take a single steel girder falling off the back of a semi into the head of a train
The head car notices the girder coming toward it as soon as it's moving off of the truck, and announces the hazard to the other cars. The whole train is keeping a long following distance from the truck (which either isn't part of the network, or has announced that it's carrying loose cargo), so it triggers the brakes or lane changes on the cars behind it.
The cost is simply that if *anything* goes wrong up front, pretty much everyone in the train is screwed.
You're assuming a lot of incompetence. Unlike rail trains, the cars in this network aren't physically linked. There's no requirement that they remain a train. If a car wants to do something different from the rest of the train, it leaves the group. The cars behind it will adjust to either form a new train, or just allow the dissenting car to leave.
And of course for cars it depends on a automated vehicle network
Yes, network benefits require a network. This is why the whole thread started with a comment about "50 years from now".
separate from (though dependent on) self-driving cars
Hardly. There are vehicular networks now, that simply inform the driver of emergent traffic conditions. They're mostly confined to university projects, but even so have already shown great potential.
which requires widespread adoption before it has any serious benefit
Not really. Road conditions can be monitored with as few as two cars per hour. Closures and detours can be pulled from government data sources , which are already becoming more accessible because travelers want to check for such things on websites. Much like high-speed cellular data service, whenever it's available, it brings a benefit, and absence doesn't diminish operation beyond the current standard.
and which has some rather serious privacy concerns
Only among the folks with shiny hats. Besides current observation, vehicular networks really just need to know vehicle specs (as measured through closed feedback loops) and an intended direction of travel. Network identifiers can be regenerated on the fly, and authenticated cryptographically. Pretty much, someone in your train can know which road you'll be taking next, and they can probably figure out what kind of car you're driving. As soon as you leave their train, you can change your identification (and if you're paranoid, fudge your vehicle characteristics), and appear to be someone completely different for the next train you join. The only practical privacy invasion is to follow someone, just as you can now.
what happens when one malicious car at the front reports a catastrophic failure?
It gets ejected from the train, and the other cars go around it, just like would happen now if the car in front of you suddenly started spewing smoke. The biggest malicious threat is still exactly the same as it is now: A few trucks can throw out caltrops on major highways. The network will detect the risk and route around it, but local streets will be overloaded.
You do not have a moral or legal right to do absolutely anything you want.
OTThirdH, expecting that the on-board computer has the _time_ to make these decisions is a sign of someone who hasn't tried embedded programming.
OTFourthH, assuming that the on-board computer has a full and complete database of all other makes and models of cars, including their crashworthiness and safety features from which any decision about which is better to hit can be made is a sign of someone who lives in a world of unicorns and pixie dust.
america is built on everybody acting in their enlightened self interest. if you have two autonomous cars and both act in self interest, it is probably the best way to minimize a crash. similarly if you have one autonomous and one non-autonomous car, the driver in the non-autonomous car will act in his self-interest anyway.