While I don't agree with Tesla's strategy of playing it fast & loose, I'm afraid that risk aversion will lead us to demand perfect instead of the good enough.

A car manufacturer might not be willing to accept 100% liability, but their system might still be a lot safer than a drunk driver returning home from the bar.

Realistically we're not going to be able to pick "well, don't drive drunk then!", it'll happen. By demanding that these system are that reliable before they're in common operation we'll have incurred a net loss in terms of safety.

I don't know what the right solution is, probably some mixture of the two. Clearly you don't want manufacturers to be incentivized to ignore the risk associated with their systems, but you also don't want these systems not to be offered when they'd be a net gain in safety.

The point at which they're a net gain is not the same point that they become financially viable to the manufacturer from a liability point of view.

I don't care what they're willing to accept. It's a total cop-out to not "accept 100% liability" for a self-driving system they built, and instead avoiding responsibility by insisting on an inhuman-level of human monitoring. If they're not willing to build a reliable system (or buy enough insurance to pay out the claims when their system crashes), then they simply shouldn't be selling production self-driving cars, period.

The "drunk driver returning home from the bar" problem has been solved. It's called a taxi. No need for defective self driving cars for that case.

With a system like that, the market would ultimately work out an optimal solution. Vehicles with safer technology and greater manufacturer liability would be subject to lower insurance rates, while on the other end of the spectrum consumers would be responsible for paying for the externalities of their choices. No one is going to use a self-driving mode that costs more than the vehicle itself to insure, so in theory manufacturers will be more strongly incentivized to get this right.

Edit: A follow-up thought: self-driving car insurance should be independent of manual car insurance. They could be bundled together for savings, but someone who doesn't intend to drive manually shouldn't be required to purchase two types of insurance. Eventually, as the technology improves, self-driving insurance will become cheaper than manual insurance.

This model also neatly solves the current issue of enforcing insurance requirements. If self-driving mode can require a valid insurance token, then so can manual driving mode.

That's... technically true but not at all an acceptable explanation for anything. We don't just go ahead and do whatever we want to "speed up innovation". All kinds of abuse constantly gets bundled under the "only way to innovate" umbrella.

If a manufacturer advertises a feature and even charges money for it then it certainly should take responsibility for the feature working as expected and for the fallout from any failure outside of user control. The customer insures against things that are their own doing, or simply cannot be assigned to someone (weather?). The company pays for it's mistake directly, not via "market forces". That's why companies don't get out of such lawsuits by just claiming "innovation" and "the market will deal with us".

"Driving" is absolutely core to a car. It's not a GPS where the manufacturer can or should be able to just say "we take no responsibilities for the outcome of a GPS or (self)driver malfunction".

If the insurance on your car suddenly triples because the stuck accelerator pedal caused accidents, you're now left with a car which is not only terribly expensive to operate but also impossible to sell because the expense follows the car. All because of something the manufacturer caused. And that will also stifle innovation in all kinds of ways.

Of course, the insurance company can simplify this by just not charging you more for future premiums, rather than paying you for future premiums and taking it back later. And by not charging other people more for future premiums, they preserve the resale value of the car.

This is an automobile equivalent of "why health insurance should cover preexisting conditions"--any increase in future premiums caused by your condition is an expense caused by the condition, so your insurance should pay for it. If you switch companies, technically the first company should pay for the increased premiums from the second company, but averaged over the whole market this is still equivalent to just not increasing premiums.

I actually have (some) personal experience with that. When Toyota has the defective accelerator pedal issue:

1. I got a check from Toyota from a settlement compensating me for the reduced resale value of my car due to the issue.

2. Insurers sued Toyota for compensation for the claims they had to pay out: https://www.reuters.com/article/retire-us-toyota-insurance-i...

So does food safety, regulation of harmful chemicals, etc. Do we have a shortage of tech development? For self-driving cars? Are these companies so fragile?

The free market's efficiency requires a balance of power and information. Similarly, in law, the validity of contracts depends on the relative position of the parties; the court will frown on a company that tries to enforce absurd contractual clauses on naive individuals.

A car manufacturer has incredible power and resources (including lawyers, actuaries, etc.) compared to an individual consumer buying a car. What will a consumer do, hire their own legal team to call the general counsel's office of Mercedes and renegotiate the agreement? It will cost more than the car. Have you ever done that with any agreement with a corporation?

We use our government to regulate where the market fails. Otherwise, we still have lead in gasoline, monopolies, etc.

The owner of the vehicle got the cheapest package and nobody will cover the victim.

This is the reason why we demand drivers to have an insurance and don't leave it to the owner to figure this out. Without regulation this doesn't work as not everybody who in the end is affected is involved in the negotiation when buying the insurance.

Insurance payout goes to the victim.

>The owner of the vehicle got the cheapest package and nobody will cover the victim.

Like regular insurance, 2nd party liability is mandatory.

The manufacturer (or their insurer) will find the sweet (market) spot between safety and costs, where costs include not just manufacturing but the legal repercussions.

Putting the onus on the end-user is messy and wasteful from a market efficiency stand point. It may benefit Elon to make the user at fault, but it's just as "socialist" (or more-so) than putting the onus on the manufacturer.

I think the bigger concern is how to assign fault. If you allow the end-user to override the software then you'll need an objective 3rd party to determine blame (not the manufacturer or the end-user).

Regulation works.

Standards works.

Consumers get rug pulled.

    > If they're not willing to [...] buy enough insurance to
    > pay out the claims [...] they [...] shouldn't be selling
    > [these] self-driving cars, period.

Western countries have mostly decided that society should simply eat that cost.

If a child gets killed by a negligent driver the driver or their insurance is almost certainly not going to be "fairly" compensating anyone, to the extent that such a thing is even possible when it comes to losing a human life.

There was a case in The Netherlands[1] where a grossly negligent driver killed 2 year old and her grandparents by rolling their car over them in a turn they were speeding through. The driver got probation, and a bit of community service.

On reflection I don't even think that's a bad outcome per-se. Very few people could truthfully report that they've never broken the traffic law in a similar manner even once.

It's mostly just moral luck that decides the difference between that being something stupid you did in your 20s with no repercussions, and another equally irresponsible person being unlucky enough to kill a child.

So should that person spend 20 years in prison and be subject to wage slavery for the rest of their life as a result? Probably not.

Now, imagine the same outcome, except this time the CEO of Tesla or Mercedes is going to have to answer in court for it. "Mr. Musk, why isn't your company paying out these parents fairly?". Of course a self-driving car is unlikely to accidentally make that specific mistake, but statistically the outcomes might be comparable.

My point upthread is that we're almost certain to end up over-pricing the risk caused by self-driving vehicles, while directly and indirectly subsidizing the continued operation of motor vehicles by even more flawed human beings.

People are simply bad at reacting to diffuse v.s. concentrated risk, and are inherently conservative when it comes to evaluating the risk of something new, v.s. an even riskier activity they've gotten used to.

1. https://www.mirror.co.uk/news/world-news/watch-father-throw-...

That's neither here nor there. Liability shouldn't be separated from the driver, since that's incentive for them to drive better. However, with self-driving vehicles, the driver is the manufacturer, full stop.

> My point upthread is that we're almost certain to end up over-pricing the risk caused by self-driving vehicles, while directly and indirectly subsidizing the continued operation of motor vehicles by even more flawed human beings.

1. You're making the flawed assumption that self-driving vehicles will be better drivers than humans. Maybe they will be, maybe they won't. It's yet to be seen. We'll know once the manufactures can't cherry-pick driving conditions or rely on backup humans to smooth over their defects.

2. If the problem you outline is an issue, the solution of pushing liability onto people who are not actually driving is a bad one.

I don't think this argument has unanimous agreement.

If we force manufactures to be responsible, then they are going to find a way to reduce payouts. No country is going to destroy their automotive industry because of self-driving cars. And Too Big To Fail companies are never held responsible for their actions.

I'm in the camp of owner-responsibility. Insurance rates for self-driving cars will set a price for the safety of the system. It will happen rather quickly and will be largely accurate. Are you going to even use a self-driving system if it elevates your insurance premiums to $10k a month? Even a person who prefers convenience over personal safety is going to think twice about paying all that money to insure themselves.

No country would "destroy their automotive industry because of self-driving cars," by making manufactures liable for their products. Worst case is the shareholders would get wiped out in a bankruptcy, but the manufactures themselves would survive.

> I'm in the camp of owner-responsibility. Insurance rates for self-driving cars will set a price for the safety of the system. It will happen rather quickly and will be largely accurate. Are you going to even use a self-driving system if it elevates your insurance premiums to $10k a month? Even a person who prefers convenience over personal safety is going to think twice about paying all that money to insure themselves.

That's a ridiculous, Rube-Goldbergian idea. Lets simplify it: the manufacturer has the liability, buys the insurance, and includes the cost into the price of the self-driving feature. Instead of waiting for your insurance bill after you buy the car to find out if you can afford self-driving, it's there in the sticker price.

I've never run over anyone, no. [Edit: The next paragraph addresses the concept of moral risk raised by the parent. Maybe this was a bad opening line, as it seemed to be the only thing responded to.]

It's unclear from your link what the person did. But, in general, hitting and killing three people requires unsafe driving that I've never done. And, to preempt the false analogy, I've certainly broken traffic laws. But I don't think I've ever driven unsafely while doing so.

It's also worth noticing that that person never even apologized. That's horrible.

> Now, imagine the same outcome, except this time the CEO of Tesla or Mercedes is going to have to answer in court for it. "Mr. Musk, why isn't your company paying out these parents fairly?"

Why wouldn't he be able to pay the parents fairly? Either (a) he fundamentally mispriced self-driving cars (in which case he should lose money) (b) his self-driving cars were worse than he expected (in which case he should lose money) or (c) he correctly accounted for risk in pricing, in which case the payouts are already baked in as a cost of doing business.

In the event of (c), it's very similar to other insurance products. Everyone pays some extra to cover the inevitable rare but serious failures.

Yes, I oppose any set of laws that allows Tesla to walk away from a fatal car accident caused by their self-driving without at the very least an appropriate civil punishment.

    > I've never run over anyone, no.

The concept of "moral luck" is about people having drastically different moral outcomes for similar actions, mostly due to getting unlucky.

E.g. most people who own a cell phone and drive cars have probably broken the laws stating that you shouldn't operate your cellphone without using the appropriate hands-free equipment.

I certainly have at least. I know I shouldn't do it, and I can truthfully say it's not a thing I "do". I know about the dangers involved. But has it happened? Yes, and I'm not proud of it.

But nothing's bad has ever happened to me as a result, but there's some person somewhere where the first time they looked at their cellphone beeping the driver's seat they drove over someone and killed them.

Can we say that I'm a better person because that hasn't happened to me? I don't think so, I just happen to have good moral luck (so far).

I don't feel that the person took appropriate care in the way a reasonable person did. But the article was unclear. Why don't you react by citing the specifics instead of doubling down on the bigger general point, since I disagree that it was moral luck in this case, not with the concept.

But yes, I've avoided failures through luck. That's just a different failure state. I admit that moral luck it totally a thing (although I am using the other poster's phrasing, not familiar with that term.) And certainly I've avoided car accidents like that. But not hitting pedestrians.

Car accidents are one of the top causes of premature death in the US. Self driving cars aren't there yet, but when they are an improvement on human driving they should be encouraged, even if the absolute numbers are still not amazing.

A car that drives itself 99% of the time, but fails catastrophically the other 1% is doomed to failure. The human operator won't be engaged enough to take over that 1%. At least not without airline levels of squawks and beeps and wheel shakers and even that might not be enough - airlines don't have to worry about children chasing balls into busy streets, etc. And the pilots are highly trained - drivers are not.

Regardless of whether or not you personally would pull the lever, you have to admit it's not the same as saving 1,000 lives.

There's no such thing as a deterministic future, as you suggest.

We can't stop doing things because a theoretical future will be changed. This is absurd.

Is it fair to trade lives of people who choose not to drive a car at all for those who do? As of now NHTSA weighs weighs them identically which I think is a huge mistake.

That said its hugely dependent on the technology involved. I think a Waymo is safer than your average driver already for pedestrians. FSD Beta if unmonitored would be drastically worse than your average driver for pedestrian safety.

People have been saying that if driverless cars are safer than humans by any amount we should allow them. I'm saying they should be much safer than humans before we allow them.

The public will not tolerate car manufacturers killing 10's of thousands of people per year just because they are very slightly statistically safer than the average human driver.

Try finding a taxi in about 99% of the US more than a few miles out from big cities.

I live just outside Minneapolis and we were SOL until Lyft arrived. Lyft and Uber still aren't available in most small towns. This is not "solved".

> I live just outside Minneapolis and we were SOL until Lyft arrived. Lyft and Uber still aren't available in most small towns. This is not "solved".

You can call for a taxi, even "just outside Minneapolis."

Also, taxis aren't the only solution to drunk driving that's better than unsupervised self-driving cars. You also have things like designated drivers, etc. It would have been silly for me to enumerate every single one in my original comment.

It seems easy enough to do if we're serious about it.

It hasn't be solved at all, there are still too many deaths on road due to drunk/high drivers.

For example in Norway you can drink 1/4 of the amount that you can do in CA while driving, you get jail penalties at .05 and right from the start at .02 the fine is one month salary, meaning even if you are rich you will pay a meaningful amount. So in general Norwegians don’t drink and drive.

In my opinion, drunk driving is solved by city planning.

It's a fight and a half to convince most Americans that it's actually realistic for suburbs or even rural communities to have functioning public transit and walkable distances, even though towns and suburbs built that way exist in America (rarely, but they do exist).

Rural towns with functioning transit and walkable distances include a laundry list of college towns.

Suburbs with these features include a laundry list of inner ring and street car suburbs built before widespread automobile ownership.

Nothing grinds my gears more than hearing someone say "people need to stop moving into my town, the traffic is getting ridiculous." It's not the people that's the problem, it's the unsustainable forms of development that are just seen as normal.

On the other hand, I like it when my tickets are basically only table change for me. Egoistics aside, scaling tjose fines with income would be much fairer than a flat fee.

As it stands now, if you get a BS ticket from one of these towns[1], you're out a few hundred bucks. Scale it up to a month's wage. In a lot of these places, you can't even get a fair trial because the local court is in on the action.

https://www.thrillist.com/cars/nation/the-worst-speed-traps-...

But we should also take steps to fight policing for profit.

More precisely: it's got a solution better than a current-generation, unsupervised self-driving car.

This is just wishful thinking. This is a description of what people should do, not what people actually do, and I don't see what relevance it has to the discussion. Millions of people drive while over the limit of alcohol consumption [1]. It is totally plausible that self-driving cars could radically reduce that number and improve overall safety even if they are not perfectly safe and occasionally cause accidents.

[1] https://www.cbsnews.com/news/millions-of-americans-admit-to-...

And what exactly is the liability for anyone produces cars that go well above and beyond every legal speed limit? By your logic, any car made that can go above the highest legal limit in a country should be held 100% liable for any accident occurring in that condition.

A certain level of data should be collected and available upon every crash like with airplanes. If the manufacturer is found at fault due to a flaw in their design, then they should be liable.

> I assume any collision involving the self-driving systems would have an investigation to determine fault, much the same way they do already

agreed. specifically, I feel the manufacturer should design the system against some minimum level of maintenance and publish that. they would be liable for any mishaps that occur within design parameters. the driver/owner should be responsible for actually keeping the vehicle within that envelope. I don't feel that's an unreasonable burden.

Getting killed by a drunk driver (which is actually the real problem) is far from solved. It's something that can happen to anyone.

No, it has not. Taxis have been around as long as cars, and people are still dying.

Drowsy driving hasn't been solved, either.

It's not 'self-driving', it's driven by Mercedes's developers.

At the .08 BAC legal limit odds of crashing a car are 200% higher than a sober driver. The risks obviously increase with higher alcohol levels. A .16 ‘double the legal limit’ drunk is 1500% more likely to crash. See https://www.washingtonpost.com/news/wonk/wp/2015/02/09/how-j...

So yeah, ‘safer than a drunk driver’ means basically ‘only about twice as likely to crash as an average sober human’.

[1]https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in...

Also, the link to the NHTSA doesn't work, but I'd be interested to read the primary research. Concern is they're just taking accident rates for sober drivers and BAC 0.08 drivers, which isn't a fair comparison because they aren't time matched. As in, drunk driving is a lot more likely to happen very late at night/early in the morning when bars close, when traffic is extremely low and accident rates are also very low. If your chances are 3x the chances of a person driving any time at all, you're much more than 3x the chances of someone driving unimpaired at 2 in the morning.

Of course, even if they time match the rates, there's still the reality that a lot of sober people driving at 2 in the morning are still impaired. Tired driving is also quite dangerous.

I could have gone for a shock stat - 100000 people a year are injured in accidents involving drunk drivers. 10000 people are killed.

I could have taken the position that a typical drunk driver is probably actually 10x as likely to have an accident as a sober driver, based on those NHTSA numbers.

But I didn’t need to to make the point that a self driving car could be literally twice as likely as an average driver to crash and that would still be ‘better than a drunk driver’.

Granted, it becomes an issue when people who wouldn’t drive drunk decide to use self-driving functionality simply because is there. But, there will be a point where it crosses over into being a net improvement, and that point may very well be a fair bit before meeting parity with humans. Obviously it’s not quite that simple (not all failure modes are equal) but it’s not that far off, either…

People who may not have chosen to drive (e.g., tired, drunk, impaired, or otherwise incapable of driving) might choose to do so since they can "lean on" the car doing it for them, and if the car encounters a situation it can't deal with the driver is completely incapable of handling it.

I think the answer right now is no, we're still requiring the driver to be able to take over, and our thought experiment presupposes they're inebriated.

On the other hand, it's obviously better to have the computer take a shot (no pun intended) at getting home than letting the driver do it themselves, even today. At what point does that moral balance tip? It seems like we're not too far off from the answer being "obviously let the car drive, whatever the drunk person was going to do is probably riskier for both themselves and everyone else".

I'm way more interested in self-driving (and, more broadly, driving assistance) for emergencies like someone having a heart attack while at the wheel. For example, I feel a lot better now that, should I somehow die or lose consciousness while driving (but not due to being drunk!), my car with autonomous breaking will mostly keep going in whatever direction I had it but with a much better chance of stopping before causing someone else harm.

I think a lot of people are unfamiliar with rural America -- the nearest bar might be 10 miles away down a country road with nothing else around (which was the case in my hometown... well more of a "village" than a town), and your friend may live 10 miles on the other side of that bar, so he may not want to drive that far out of his way so you can get drunk and he can't. And taxis and buses aren't an option.

Relying on humans to make the correct and safe choice to not drive while drunk is a losing proposition, our DUI death rate shows that.

Even if a smart car can't be trusted to drive you, it should be able to notice that you're drunk and refuse to let you drive. My car notices when I'm losing focus and gives me an alert reminder, so it should be able to notice when I'm drunk.

How is the bar not sued out of existence? There is no plausible way to travel to and from the bar, other than a private automobile, and no taxi service within an hour's drive. There are no houses nearby with clientele that could walk to the bar. Why isn't that considered an attractive nuisance like a swimming pool?

So, hopefully not very often :) But it's one of those things where, since I could afford to change the car, I felt I had to go ahead and make that investment in safety.

I think they're wrong to do this, because the worst potential outcome of drunk driving is so devastating and catastrophic that it's personally optimal to eliminate that risk to zero. But the fact is that this isn't really how humans weigh risk.

Obviously anti-drunk-driving campaigns have been somewhat effective. People take this problem more seriously than they did 40 years ago. But I wonder if we aren't at a point of diminishing returns. Some people are going to drive drunk, no matter how many times you tell them to take an Uber.

Or he can just stop on his way home.

We both know what the right thing to do is, but one has significant costs the other doesn't.

Or imagine he does stop at the bar on his way home and tells himself he'll just have one beer, but he ends up having four and shouldn't drive. Now he has to get an Uber, take it home, get another Uber before work in the morning, and take it back to the bar to get his car. And then, finally, head to work.

Or he can just drive home.

Again, we both know what the right thing to do is, but one has significant costs the other doesn't.

Even if that were true, I have a hard time imagining that the cost of self driving packages won't continue to decrease until it's included in literally every new car. Just a matter of time (assuming someone is able to make a fully working version).

I don't live in an area that has Uber, so I'm not that familiar with it, only used it a few times when traveling :)

My impression of the the way drivers are compensated is that the hours, fuel, and wear on your vehicle make it kind of slim margins.

Other than when Uber deeply discounts to drive growth, it's always been cheaper to drive yourself than to pay Uber. The reason you pay Uber is for other reasons (e.g. convenience, cost of parking, being away from your vehicle, intoxication).

Now we're onto something!

Self-driving cars which are merely as safe as driving with a 0.08BAC will still greatly reduce traffic accidents. Of course, no one will report it as such. Having something that's paying attention 100% of the time is going to be a huge boon for traffic safety. Spending 15 seconds picking a song on your phone is all it takes to rear end a car or blow through an intersection.

A 100% increase would be 2x as likely.

I gave the ‘better than a drunk driver’ claim the benefit of the doubt and allowed it to be merely 2x as likely to crash than a sober driver, since that is still ‘better’ than a driver who is exactly at the federal US legal limit (and who will crash 3x as often).

What are the odds of a sober driver crashing a car?

[1] https://www.allstate.com/resources/allstate/attachments/best... Page 4

[2] https://www.bts.gov/content/motor-vehicle-safety-data

Put in another way, I don't expect my taxi driver to be perfect, in fact I probably drive more safely than most of them. But if they get into an accident it's their fault, not mine.

For now I'm leaving aside things like "summon my car to me", or "tell the car to randomly drive around until it finds a parking spot" where no human would be present in the vehicle at all because I think those kinds of scenarios are still a very long long way off, where the vehicles can be expected to move with no supervision at all.

If instead we talk about driving assistance and I am supposed to keep an eye on the car, I am fine with being responsible for any accident.

Then Mercedes' move is exactly the sort of pragmatic, no-bullshit yardstick you want. 'Putting your money where your mouth is' is a widely-understood and respected way of showing you are not bullshitting.

Current liability laws have not forced car manufacturers to make perfectly safe vehicles (which, would, of course, kill the industry), and there's no reason to assume they will do so for self-driving cars.

And this addresses the "uncanny valley" of driverless cars, where the car does everything up until you're expected to leap into immediate action to prevent it from killing someone, even though for hours you may have been doing nothing.

The technology really does need to be all-or-nothing. The in-between solution that Tesla is trying to exploit isn't reasonable given how human minds operate.

I very much welcome this development and it is, in my opinion, necessary for level 3 autonomous driving. By definition, level 3 autonomous driving implies that the car must take responsibility. The Level 3 definition says that the driver may take his attention off the road and does not need to supervise the technology (if the conditions are right).

Regarding drunk driving, if the machines start killing people, good luck explaining to the victims and the scared public that it may have been worse if they were drunk drivers. The fact is that new technology usually requires notably higher performance to replace the old, this will be the case with self driving, so we may as well embrace that and go forward.

Manufacturers who produce safer self driving vehicles will have to spend less to insure their fleet, and be more competitively priced. This at least gets the incentives right

> Once you engage Drive Pilot, you are no longer legally liable for the car's operation until it disengages. You can look away, watch a movie, or zone out. If the car crashes while Drive Pilot is operating, that's Mercedes' problem, not yours.

If the car crashes, it's your problem primarily because you're the one in the car that could be injured. Maybe they'll pay for repairs or your hospital bills or even a loaner, but its primarily your problem. We have to remember the limitations of these systems and accept that they're level 2 and require full attention

What Mercedes does here is not just empty advertising, it’s offering actual lvl 3 autonomous driving complete with the car being liable in those situations.

Which in practice means Mercedes can actually deliver what Tesla so far has only been advertising with; Taking your attention actually away from the street to watch a movie or check your emails.

This is not a “problem”, it’s the literal definition of lvl 3 autonomous driving; Car takes liability in certain situations to such a degree that the driver is legally pretty much just a passenger.

If I were the kind of person that can afford and buy an EQS[1] I would rather hire a driver so I can do it all the time, not just at 40km/h in a slow moving highway jam.

[1] https://www.mbusa.com/en/future-vehicles/eqs-sedan

Do you have experience with this tech or are you taking the word of an marketing campaign? Taking "liability" doesn't mean much if you get in a serious accident and have medical issues for the rest of your life. If you get in an unsafe cab does it give you a lot of assurance that technically the driver is responsible?

For a more apples to apples comparison: Tesla vs Mercedes. Tesla broadcasts L5 confidence in a poorly implemented L2 system, and accepts zero liability for its faults. Mercedes broadcasts L3 confidence in their system (which I'm not familiar enough to actually characterize), and accepts full liability for its faults. I'd prefer to share the road with a self-driving Mercedes than a Tesla, because mistakes are inevitable and insurance payouts are not. If I'm disabled for life, I'd rather get paid for damages than not.

Tesla, until maybe this year, was never cashflow positive and the remaining lifetime of the company was always a significant risk. MB is an institution here in Germany. For many many reasons they will be one of the car manufactures that will disappear last. This is a company that cannot do such a risk without massive personal retribution risk to the executives, especially not after the Diesel scandal.

Of course that doesn't remove the "but I'm in the hospital/disabled/..." part of being personally in an accident. But that can happen _anyways_, irrespective of you being responsible even if you drive yourself. Compared to any other offering, this is a significant change of the rules of the game.

People do not care what “level” their driverless control is. After a certain point of “it seems to do everything”, they assume it does everything. If they think they can get away with letting it take control, they will.

Requiring automotive manufacturers to deal with this is manifestly responsible. It is demanding an engineering control over an admin one in the hierarchy of risk controls, and that ecpectation is in line with how every non-software engineering discipline of engineering approaches the world.

Also: Drive Pilot is a level 3 system, not a level 2 system. That's the entire point.

Short of MB putting an executive to sit alongside every passenger in their car (which you'll note that Uber, the airlines etc don't do either), what exactly are they supposed to do above and beyond taking liability for accidents while the system is in operation?

Level 3 is the lowest level where “the car drives itself”

https://www.motorweek.org/images/SAE_Chart.png

"as it happens for example for vehicles with smaller power trains."

Is this still true? I know they did some HP based stuff in the 70s-80s, but thought they did away with that since then.

Vehicle value impacts the premium, since the policy may cover things like theft and vehicle damages. However the biggest risk is damages against persons. Truck insurance for example is very expensive not because the value of the truck but because an accident is more likely to involve deaths and permanent damages. It's not unusual accidents for several million €. That said there are lot of considerations for selling policies. The insurer may make a discount for a risky sports car if it thinks it will recover the cost cross selling you other products. Or for commercial pressure, accidents are a problem for the future. But yes, you are right, it's not a certainty.

I'm actually afraid of the opposite: where self-driving becomes so associated with failures and danger in the mind of the public that the technology never take off.

And it doesn't take that much effort to make the public afraid of something, just look how terrified of shark people are.

Accepting liability really just comes down to who's insurance company pays when it fails.

If the manufacturer can demonstrate to the insurance company that their self driving system fails less frequently (or lower overall cost of claims payouts) than human drivers then the insurance company will happily charge a lower premium for cover. Even if it's not perfect. The closer the system can be demonstrated to be to perfect the lower the insurance premium is going to fall.

There is certainly a regulatory moat being set up here as it will prevent new startups from entering the space unless they can find an insurance company willing to cover their risk profile, which in turn means demonstrating a suitably high level of safety to keep the premiums affordable. But I think in this matter I want a level of regulation that says if you're making something that's used on public roads, and risks injuring or killing people you'll accept liability when it goes wrong. It's not much different from the system of mandatory insurance for drivers in most countries. If you're going to drive on the roads, you must meet a minimum insurance level whereby you can accept liability for any failure that causes damage to other property or people.

Personally I wouldn't want to be near automatically driving vehicles for now. Maybe because I worked in computer vision and even with drastic advancement of the field, the possibilities are still quite primitive, even with "AI" support. Sure, I could get hurt by a drunk driver. But they are also not allowed on the roads for that matter.

I think a solution is to restrict the zones which an autopilot is allowed to be active at first. Ultimately the driver must be responsible. He made the decision to activate the auto pilot. Manufacturers should be forced to be honest about the capabilities though.

Mercedes is just going to place the assumed liability with a reinsurance company anyway. It seems like a very efficient solution to ensure accountability SOMEWHERE without vast amounts of litigation every time there is an incident. It's honestly hard to see how self-driving vehicles will ever meaningfully make it to the consumer market without an agreement like this.

https://old.reddit.com/r/germany/comments/tejhxv/all_my_amer...

"All my American friends drive after drinking and I don't know what to do"

https://www.itf-oecd.org/alcohol-related-road-casualties-off...

A friendly reminder from a fellow European that the entire continent isn't part of Germany :)

Not wanting to crash into random things when rolling at 120 km/h is not "risk aversion". It's just common sense.

> risk aversion will lead us to demand perfect

And no, perfection is not a reasonable bar.

It's the only bar. Considering the regular person overconfidence in their ability they will either want something that doesn't crash ever or if it crashes it has to bring about other advantages such as speed.

Eg. Same rate of self driving crashes with an avg. speed of 100mph as human 40mph.

Like putting your Mercedes in the hands of Lewis Hamilton. People won't have any problem doing that, because they understand that Hamilton is a much better driver than them. If it's close people won't relinquish control, at least when sober.

If it's close I'd rather bet on myself. Many people feel the same way.

people discount the psychological effects of letting go of the wheel, the advantage has to be crystal clear

It obviously isn't as there are many people who have jumped through many hoops to get access to self-driving software that is much worse than Lewis Hamilton (i.e. Tesla FSD beta). Clearly different people have different standards for what they consider safe enough to let it drive for them.

It seems like you are redefining "perfection" to be whatever your personal bar is for adopting a self driving car (which would make your argument a tautology if you weren't also overgeneralizing your personal bar to everyone.)

I would also point out that the bar for personal adoption is not necessarily the one that we should use to determine which self-driving solutions are allowed to operate at what level on public roads. While this is a complicated topic, I don't think "Lewis Hamilton" is the only reasonable place to set those bars.

I am talking about fully giving control to the car and preventing humans from interacting, after clear statistical evidence of autopilot being better than the avg. driver. Given that by definition the majority of drivers are avg. then society would be better off summoning Skynet and putting it at the wheel. It's a clear cut reasoning that works in theory.

Better than avg. doesn't mean that accidents won't happen though. Accidents will still happen and as soon as they start to happen the whole thing will be reversed. That's because people don't consider themselves the "avg. driver" and would rather bet on themselves being better than the avg. driver than the computer being better than that .

Even 95 yr old people consider themselves the local Lewis Hamilton.

You raised the example of FSD purchases, I cite you the counterexample of vaccines. Vaccines are much much much better than the immune system against COVID, if it were to be close you'd have much less adoption because skepticism is not linear.

When in doubt , people bet on themselves or the status quo. In this case the status quo is literally betting on themselves so you have a double effect there too

People are already letting cars drive for them. They are doing it even when they are not allowed to take over for the car (i.e. Waymo in AZ). Clearly some people have lower bars so your claim that the above average safety is required for anyone to give up control of the vehicle they are riding in to a computer is false.

Reading between the lines of what you are saying, is seems that what you are actually trying to argue is more along the lines of "achieving (significantly?) above-average safety for self-driving systems is crucial to achieving mass adoption of self-driving technology." That's a much hard to disprove statement (and one I suspect it is at least partly true given varying levels of "cruciality").

It won't be the car manufacturer, but their insurance underwriter who has the final say (unless the car manufacturer self-insures, which, let's face it, is a non-starter).

I'm not sure if you've ever dealt with business insurance, let alone underwriters.

Underwriters tend to be, by definition, fairly boring risk-averse types.

If you catch them on a good day, they might be willing to make a few relatively small adjustments to their standard wording in order to relax the scope a bit.

However this assumes you are lucky enough to be dealing with one underwriter. If you're having to pool at Lloyds, then the chances of you being able to get significant wording relaxation drops dramatically because everyone in the pool has to agree, not just one underwriter.

You'll never find an underwriter willing to cover 100% liability for anything under the sun. That's just not how it works. Especially with new and little-understood risks such as self-driving.

You will own nothing and like it.

I'm fine with a company saying "We don't think we can do X safely, therefore we don't devote as much to research doing X, therefore we don't want to be penalized in the market because people who do X unsafely can obfuscate the risks and sell it to the market." And then attempt to create standards to do X safely. That seems reasonable to me.

The right solution has always been public transportation. It's just, you won't hear that from car companies. The right solution to self driving cars is not to have any cars.

There are 1.35 million deaths per year attributed to car crashes. I don’t know about you, but I would rather companies play loose and fast rather than take forever to figure out the perfect solution. If it takes five years longer, that’s more than 7 million deaths.

Even if self driving cars are responsible for 1000 deaths per year, which they won’t be, it would still be well worth playing fast and loose.

That being said, the number you stated is world wide? If so, it's not the one to use. And just ignoring the overall trend, in death per km / mile driven, is just disengenious.

Also, traffic related injuries and deaths have been rising in recent years due to distracted driving

Two, driving conditions have to factor in to make it comparable, including traffic deaths in Africa and India (global) is kind of pointless

Three, I'd like a source that per driven mile injuries and deaths are rising, all numbers I know of show the opposite trend (including pedestrians and cyclists).

Edit: if something happens you have 10 seconds to react . I don’t know if in that 10 seconds those responsibility counts.

I don't think legislators will ever accept the safer criminal activity argument.

German cars are designed for cities American cars are designed for LA.

Those are some big caveats that mean that you won't be able to use this in most situations. It is basically only good for stop and go highway traffic which is a situation that other driver assist features handle pretty well.

All other companies don't go that far and make absolutely no promises. Sometimes their marketing wink-wink-nudge-nudges you and implies that they do the same but in reality they don't.

If this is successful you will soon see customers and regulators requiring the same for all competitors.

I don't think it's that big of a deal but it's clearly well done from a PR/marketing standpoint.

Insurance/warranty is just an expense from a companies POV. You control it via a combination of

a) building a well working product

b) limiting use of that product

c) putting a high enough premium on it

__

a) is what all customers want. b) is what Mercedes is heavily leaning into right now, because looking at the restrictions it's pretty clear that they don't think they have accomplished a) to any satisfying degree.

I am sure they will be setting c) to where their insurance math says it has to be to be financially viable.

This is a really big deal. If you are required to instantly take over, you need to permanently pay attention to the current road situation, at which point the autopilot is really just a fancy cruise control. People still stop paying attention, of course, but that's actually a massive risk.

A longer takeover window actually allows you to do something useful, such as read or look at your phone, without taking this risk, since you will have time to adjust to the situation if necessary.

How will the car detect a construction zone it can’t see yet with enough time to hand over to an inattentive driver?

I look forward to seeing this system in operation. I have significant doubts about the feasibility of its operational claims.

a) The system would have to be allowed to disengage automatically when conditions change unfavourably, in which case you would still have to be alert, all of the time for when that happens

b) It would not be allowed to do that automatically and you are liable from the moment autopilot drives into an area that is exempt from its legal responsibility as laid out by the insurance coverage limitations

For example, take a look at the exemption of "construction sites": Either the car disengages and says "from here on our it's your job, not ours anymore" or it does not, and then in the case of an accident you are not covered by their limited legal responsibility. What the autopilot can definitely not do, is making the construction site disappear or guaranteeing that the car will never hit one after having been engaged.

c) The system needs to detect worsening conditions early either prompt you to take over with enough time to spare (or fail gracefully).

That's the big thing that Mercedes guarantees here: You'll have enough time to take over even if you're doing something else; if the system does fails to give you a warning in time and you crash, Mercedes takes the responsibility. In all other systems, once the autopilot prompts you to take over, you are responsible. With this system, once the autopilot prompts you to take over, Mercedes is still responsible for the next ten seconds, which should be more than enough time to take over in an emergency.

Regulators also love b) and now that there is precedence we should see the next step by some of the a) companies.

If their product is better they should be able to easily match or exceed Mercedes (and I think they will. Either voluntarily or by law).

To me this is a really exciting step along the way to autonomous vehicles.

However claiming they have beaten Tesla seems like a bit of a stretch given the circumstances.

Tesla probably has the lead on average but can't (or won't) guarantee that their system is safe in a specific set of best case conditions.

It's a different approach if you ask me.

I am still waiting for the promised full autonomous cross-country trip from East to West. I think it was promised for 2019.

While Mercedes had a full autonomous trip including small town traffic and several round-abouts.

Mercedes just applies higher standards on what they deploy on average Joe.

https://youtu.be/o7oZ-AQszEI / https://old.reddit.com/r/teslamotors/comments/s7vea9/fsd_bei...

Although his earlier claims in 2014 / 2015 were met, as autopilot works reasonably well on roads and highways. The claim for LA->NY was 2016[1]

1: https://www.businessinsider.com/elon-musk-autonomous-tesla-d...

The automotive industry is more like Apple. They don't talk and show off until its done. Like: Never over promise and under deliver.

Do people not see how much of a nightmare it would be if liability is constantly switching back and forth depending on the circumstances the self-driving system is being used under? The only practical approach is to either have the driver always be responsibile when the system is on or the manufacturer always be responsible when the system is on. The choice to use the latter approach means the system has to be limited to very narrow circumstances. That doesn't mean the system is necessarily safer than the competitors in those circumstances, it simply means that the companies are approaching the same problem from differenct angles.

There's a reason for that and it's because of a lack of confidence in the technology.

If this is true why don't all of companies do it?

From Slide 23: If you don't take over when requested, after an automated stop, car will unlock the doors call to your emergency response center.

"An Automated Driving System for the Highway - Daimler" [PDF]

https://group.mercedes-benz.com/documents/innovation/other/2...

"A Joint Approach to Automated Driving Systems - Daimler" [PDF]

https://group.mercedes-benz.com/documents/innovation/other/v...

Edit:

"The technical requirements"

"When the DRIVE PILOT is activated, the system continuously evaluates the route, traffic signs, and occurring traffic incidents. As a layperson it’s hard to imagine how sophisticated the hardware and software of the S-Class is in order to be ready for Level 3. Even the “normal” latest-generation Driving Assistance Package has the following:

• A stereo multi-purpose camera behind the windshield.

• A long-range radar in the radiator grille.

• Four multi-mode radars (one each on the right-hand and left-hand sides at the front and rear bumpers).

The optional parking package additionally includes:

• A 360°-Camera consisting of four cameras in the right-hand and left-hand exterior mirrors as well as in the radiator grille and at the trunk.

• Twelve ultrasonic sensors (six each at the front and rear bumpers).

"For the DRIVE PILOT, many additional components are needed besides the sensors of the Driving Assistance Package. The long-range radar in the radiator grille is combined with a LiDAR (light detection and ranging) system. Whereas radar uses radio waves, LiDAR employs pulses of infrared light in order to optically determine an object’s speed and distance and to create a highly precise map of the vehicle’s surroundings. This combines the strengths of both technologies: LiDAR sensors operate with higher precision, while radar is advantageous in bad weather, for example."

"The rear window is equipped with a rear multi-purpose camera that scans the area behind the vehicle. In combination with additional microphones, this device can, among other things, detect the flashing lights and special signals of emergency vehicles. The cameras in the driver’s display and MBUX Interior Assist are always directed at the driver so that they can determine if he or she falls asleep, turns around for too long, leaves the driver’s seat, or is unable to retake control of driving for other reasons."

https://group.mercedes-benz.com/magazine/technology-innovati...

The company is now officially called Mercedes-Benz as of recently, BTW.

The Daimler brand is now with Daimler Trucks, following a split.

I'm also sure they will not take any responsibility if someone rear-ends you when the car stops in confusion on the highway.

Yes they do. That's the point of their announcement and the reason why they only allow it under very specific and favourable conditions

First, a software vendor accepting responsibility for the software's actions? Wow.

Second, they're confident of being able to predict accidents ten seconds in advance? That's up to 160m away and I think that's great, even if they limit the circumstances sharply and allow many false positives.

I don't find it remarkable that the software has many reasons to disengage. I find it remarkable that potential accidents >10s into the future are on the list of reasons, even in limited circumstances.

The first software I bought came with a warranty that covered nothing: It explicitly said that the software wasn't guaranteed to perform "any particular function". As I read that text, that vendor had the right to sell me three empty floppy disks. You've seen similar texts, right?

And here we have Mercedes guaranteeing considerable foresight in limited circumstances. No matter how limited the circumstances are, that's a giant leap.

It's all a trade-off. My accident-rate thus far is one serious accident in over 700000km of driving. My fender-bender rate is three FBs in over 700000km of driving.

My understanding of the Tesla system[1] is that it requires roughly one intervention every ~5000km of driving in order to avoid an accident. For me this is an unacceptably high risk, because not intervening in 4999km will definitely (100% certainty) mean that I will be in a poor position to react when the intervention is necessary.

Now, you might claim that the driver has to be alert while not in control for 4999km to avoid the accident on the 5000km mark, but if drivers were that good at being alert while not engaged with the act of driving, then the self-driving system is redundant anyway.

[1] I read the stats a long time ago, so maybe they've changed.

It reminds me of this other excellent discussion about taking risks the other day: https://news.ycombinator.com/item?id=30264760

Tesla's system still requires me to pay attention to what's happening to exactly the same degree as normal because I might need to intervene (and in fact makes it harder to be ready to do so). All it does is take away the (to me) trivial aspects of pressing a pedal and turning a steering wheel.

Whereas this system introduces a set of circumstances in which I don't need to drive at all. No need to pay attention at all. And it's the most tedious form of driving there is - stop start traffic on a motorway.

As someone who has used autopilot in the circumstances you describe, I'd be shocked if you need to pay more attention during autopilot than the Mercedes system. Autopilot generally requires intervention in danger situations, which are incredibly rare in stop/start traffic.

What you're saying is like "the engineers that built this bridge never drive over it, but that doesn't mean it's shoddy" - technically correct, but almost certainly wrong in practice.

From a practical standpoint, liability needs to be all or nothing. You can't have a driver worrying about whether they are going 40mph or 41mph. You can't immedaitely give the driver liability if it starts to drizzle or the sun sets.

Mercedes is taking the approach that they are always responsibile. Other manufacturers are taking the approach that the driver is always responsible. The end result is that the Mercedes system is much more conservative in how it can be used. This says nothing about the quality of their technology in comparison to their competitors. It simply says they are focusing on the easiest problems first while their competitors are taking a more holistic approach trying to design a system that has more broad usability.

So, the current goals must be to achieve something similar in certain well defined limited conditions, and with reliable automatic checking that you are still within those conditions - hopefully conditions that one is actually likely to encounter. Until we have that, letting self driving cars on public roads is a menace.

Current self driving cars are at best at the level of a driver going through their first driving lessons, and one with very bad eyesight at that. Having a human act as the driving instructor, theoretically prepared to step in whenever the AI makes a silly mistake, is not enough to make these cars as safe as the average (non-drunk, non-sleep-deprived) human driver.

What Mercedes seems to be doing is responsibly pushing the state of the art further. Having a car that is safer than a human driver without depending on your constant vigillance is a huge step forward. Obviously, this only works in certain conditions, but the car itself detects when those conditions are no longer met, and gives you quite ample warning to assume back control.

* Elon's shameless lies about having your Tesla act as a self driving taxi and generate a profit for you while you work in the coming years have well been put to rest.

They are saying that they have enough confidence in the system that they will pay for anything that goes wrong. This means that they have run the calcs to work out how much its going to cost.

Tesla have basically gone: "fuck me its bad, lets just legal boilerplate ourselves out of the consequences. Oh and charge people to QA our shit"

Mercedes will accept legal responsibility only while DP is engaged so driver still has to pay attention and react(so called manual take-over).

DP max speed is 60km/h or 17m/s. System is likely to detect and issue under 100m away so driver has less than 5s to take over and sort it out.

Going below 60km/h on the Autobahn... This seems to only be useful for the narrow use case of very bad traffic.

It still seems like we need to adapt the roads to autonomous vehicles and not the other way around. In the UK, and there's no indication of it whatsoever at the moment, I hope we never end up with the US's crazy car-centric town streets where as a pedestrian you can only cross at crossings, and if you get hit by 2 tons of steel, it's your fault.

Don't all highways in Germany have minimum speed of 70 km/h? So basically there's no road where Drive Pilot can be used?

No. The only legal limit is that the vehicle has to be type-licensed for going more than 60 km/h, which an S-class clearly is. Actual minimum speed requirements are rare.

Nevertheless this is for now clearly primarily useful as an assistant for being stuck in bad traffic, and you can get ticketed for going slow enough to hinder traffic flow without a good reason, and if you can argue a good reason depends. An upgrade to 80 so you can stick it behind a truck in the right lane would make it a lot more practical.

... but because Tesla has the reality distortion field in full effect you never hear these complaints about their solution.

Driving up sidewalks, killing people by slamming into stationary objects, people trying to sleep in your glorified ADAS guided car, all the cost of being able to claim you work in more situations.

I mean if you commute regularly in a major metropolitan area that could easily be 90% of your driving.

In Tesla you have to have hands on wheel all the time.

Nobody likes driving in that, a free highway can actually be fun, but as soon as the traffic gets dense it becomes the opposite and quite high stress.

Level 2 systems can often work in more conditions because they have a lower requirement for reliability. They are permitted to make mistakes without realizing it, because the human driver is required to monitor and override the system.

Tesla is innovative, but they get ahead of themselves, and Musk is a glorified internet troll who bullies and arguably lies.

The Big 3 are respectable in that they don’t really overhype or misrepresent, but Mercedes will probably out-innovate them in self-driving like they’ve out-innovated them (and pretty much everyone else) in just about every other aspect that matters over the past century.

Disclaimer: I drive a Toyota Sienna and a Subaru Outback. But I’ve seen a few Mercedes up close (and sometimes driven them) over the years. The only really crappy one in my experience was the ~2006 G500.

https://jalopnik.com/mercedes-may-have-been-caught-using-def...

https://en.wikipedia.org/wiki/Diesel_emissions_scandal#Merce...

When Mercedes will make this feature and promise available to each A-class sold worldwide we will be able to compare abilities. This day will come some time in the future, but the current move is not leading me to believe that Mercedes is a safer bet then Tesla.

Try the CLA 250. Their mid and high ends are pretty nice, but they tried to target <40k without sacrificing too much of their margins.

So how does this compare to Teslas FSD (not Autopilot, which really is only an assistant)? The answer is: not much at all. The Mercedes system has limited the conditions to a point, where it basically just has to do collision avoidance with traffic with predictable behavior. It is limited to selected roads. You don't need much "AI" here and for this task, LIDAR excels in the specified weather conditions.

The Tesla FSD system on the other side aims to be a generic driving AI. It has no limitations about on which roads it operates, it uses maps only for navigation to the destination. But the driving operation is done solely by the evaluation of camera footage. Which already works in many more environmental conditions than the Mercedes system.

These systems approach the goal from entirely two directions. The Tesla system aims to be a complete solution, but certainly is not able to do that unobserved in the current state. It is doing reasonably well at that, but the literally billion dollar question is, can it ever get so reliable, that the driver can completely hand over to car ever? On the other side Mercedes has full autonomy, but they achieved that in a very limited and controlled environment - it remains to be seen how much they can extend that.

Using the current machine learning algorithms? Not likely. Imagine releasing a new airplane and telling the pilot - "this airplane flies well in 99% of scenarios, but 1% of the time it refuses to respond to pilot commands and kills everyone on board. We don't know why it happens but we'll keep training the model". That's the state we're currently at with "AI"; as someone put it, it's alchemy. If it works, we don't know why. If it doesn't, we don't know why - but we tune the parameters until it seems to work. We cannot prove it will work for every possible input. ... Not exactly a scientific approach by any means. Such self-driving cars might be accepted in USA but I imagine they'd quickly get banned in Europe after a few random fatalities.

https://www.tesla.com/VehicleSafetyReport

1) Autosteer and active safety features are not "autopilot", its adaptive cruise control.

2) if they think something bad is about to happen, it beeps and hands over control, thus its the drivers fault, as the system wasn't engaged at the time of the crash

3) its not their self driving tech.

2) Tesla count 5 seconds after a disengagement and count it as a fault of autopilot if a crash occurs.

3) It's part of their self driving tech, but it's different to FSD - yes.

They may have the worlds best driver assist features, but they have yet to produce a product that drives itself in any capacity.

https://www.motorweek.org/images/SAE_Chart.png

The SAE is looking at a different performance hurdle between levels 2 and 3: reliability.

If you ask a normal person on the street what "autopilot" means in a car, you'll get a range of answers. But the consistent opinion is that autopilot means that you don't need to concentrate. Most will say that its automatic driving, and will drive for you.

But autosteer requires concentration. Its just sparkling lane assist.

> it's different to FSD - yes.

Which is the point. Its marketing fluff. Autopilot isn't really that, its adaptive cruise control with lane assist. Thats the thing that rankles. All of this stupidity, injury, and noise comes from a marketing decision. One designed to cover up that the CEO over promised and wildly under delivered _yet again_.

"NHTSA's Flawed Autopilot Safety Study Unmasked (2019) - The safety regulator's claim that Autopilot reduces crashes by 40% was based on flawed data and analysis, which it attempted to keep secret.": https://www.thedrive.com/tech/26455/nhtsas-flawed-autopilot-...

"Tesla Autopilot Safety Stats Said Imbued With Statistical Fallacies, Interpret Cautiously": https://www.forbes.com/sites/lanceeliot/2019/06/09/tesla-aut...

"In 2017, the feds said Tesla Autopilot cut crashes 40%—that was bogus": https://arstechnica.com/cars/2019/02/in-2017-the-feds-said-t...

1. NHTSA has reiterated that its data came from Tesla, and has not been verified by an independent party (as it noted in a footnote in the report).

2. Second, it says its investigators did not consider whether the driver was using Autopilot at the time of each crash. (Reminder: Drivers are only supposed to use Autopilot in very specific contexts.)

3. And third, airbag deployments are an inexact proxy for crashes.

... which all sound like really flimsy reasons to conclude "doesn't stack up." A more honest summary based on those would be something like "hasn't been verified by a 3rd party yet."

It's incredible how openly the media makes things up about Tesla, likely just to generate page impressions.

https://twitter.com/Tweetermeyer/status/1488673180403191808

The primary issue is that the data doesn't adjust for road classification in any way. Highway driving, where autopilot is used, has considerably fewer crashes per mile than city driving. Tesla compared Autopilot's rates against all driving rather than just highway driving which would be the relevant metric.

That adjustment alone almost completely eliminates any safety advantage of Autopilot before you get into any of the other adjustments like age.

It's like this:

Tesla: Look how great we are, our apple is redder than their orange.

Critic: It doesn't make sense to compare the colors of those things in that way. Here's why ...

You: So being a redder apple should now be disregarded?

Could you explain the flaw in Tesla data? How is crashes/mile not a good proxy for safety?

But this is not FSD , is a combo of human + some driver assistance, come back when Elon discloses how many times the human was forced to intervene to prevent a crash.

I am wondering why Tesla fans think this stats are proving that so called FSD is safe , don't you understand stats? did PR tricked you,? this proves that soem assistant like one keeping safe distance between cars is safer, make all the data public and let us compute the real stats please.

If you only engage Autopilot in the easiest situations to drive in, it might look safer artificially because of that.

Comparing a Tesla to all vehicles means old cars with less safety features enter the picture, skewing the statistics again.

Well, it exists. Courtesy of Wikipedia:

> In December 2015, Musk predicted "complete autonomy" by 2018. At the end of 2016, Tesla expected to demonstrate full autonomy by the end of 2017, and in April 2017, Musk predicted that in around two years, drivers would be able to sleep in their vehicle while it drives itself. In 2018 Tesla revised the date to demonstrate full autonomy to be by the end of 2019. In February 2019, Musk stated that Tesla's FSD capability would be "feature complete" by the end of 2019. In January 2020, Musk claimed the FSD software would be "feature complete" by the end of 2020, adding that feature complete "doesn't mean that features are working well". In August 2020, Musk stated that 200 software engineers, 100 hardware engineers and 500 "labelers" were working on Autopilot and FSD. In early 2021, Musk stated that Tesla would provide SAE Level 5 autonomy by the end of 2021 and that Tesla plans to release a monthly subscription package for FSD in 2021. An email conversation between Tesla and the California Department of Motor Vehicles retrieved via a Freedom of Information Act request by PlainSite contradicts Musk's forward-looking statement.

https://en.wikipedia.org/wiki/Tesla_Autopilot#Predictions_an...

(Also, incidentally, Musk's decision to use cameras instead of LIDAR – for depth perception etc – is insane and is widely regarded as such. He's doubled down on saying LIDAR is rubbish, since the entire industry has bet on it, and presumably Tesla is too far into the woods to reverse.)

Let me know when humans get lidar implants because they can't perceive depth via their eyes.

MB also uses cameras extensively. LIDAR and maps are used in addition.

I wonder whether these approaches will one day meet in the middle.

I just felt the urge to clarify that this is written with the intention to mislead readers, and that 40 mph are also 60 km/h in other parts of the world.

I think it's just bad editing.

It’s still my problem if I’m inside the car whilst it crashes.

As long-time grey-beard MB owner, and, honestly, it's not just them, the sheer amount of noise and distractions in modern cockpits is absurd now.

By this I mean, reversing sensors, slow-forward speed sensors, massive LCD screen (which may or may not be touchscreen) positioned at windscreen level, the right-hand "-ometer" that's configurable, a dial, a trackpad, a steering wheel with at least four buttons, and two 'joystick' things. And 3 instrument clusters on stalks. Heating buttons, fan, windscreen control, radio button, all down the bottom of the centre console where I have to take my eyes off the road.

It's a shit show.

My point? My current MB has never been so distracting to drive where I have to take my eyes off the road for a second to hit the demister button. Versus a car from the 1990s - no centre-LCD so the heater/demister buttons were in your peripheral vision. No touchscreen - regular buttons with tactile feed back that even had "braille like" raised paint so you could figure out where your hand was without taking your eye off the road. Don't even get me started on trying to use the radio or equaliser.

I hope, I really do, that autonomous driving works and reduces crashes, but I'm astonished there hasn't been a noticeable uptick in crashes based on just what I've written.

From the article, this is yet another button on the steering wheel. Probably because they've run out of dashboard space for all the other buttons. Auto-stop, car alarm off, assistance, sport mode, are at least 4 buttons I never use.

Where's the Jony Ive for car cockpits?

That's the single touchscreen world of the Model 3 and Y :)

“Brick wall” is doing a lot of work there - note that it’s not the same as hitting a stationary car at 50mph, that car would move. We rarely encounter actual immovable objects on or near the road, which is why a 50+50 head-on is more violent than a 50+0 collision into something softer.

https://youtu.be/C_r5UJrxcck

https://electrek.co/2020/11/11/tesla-most-attractive-company...

Elon is the chief designer and chief engineer at SpaceX. If you watch the video referenced in this article and are not convinced that Elon does engineering, then I don’t know what to tell you.

https://observer.com/2021/09/elon-musk-spacex-title-design-e...

If you consider that it is the driver who is liable today for road traffic incidents (including failing to adapt to environmental conditions)... then why wouldn't the "driver" still be liable when the driver changes?

https://www.synopsys.com/automotive/autonomous-driving-level...

In this it's plain that Level 3 autonomy is the first level where it's considered that the car is the driver. This is what Mercedes are proposing, to accept that the sum of software and hardware that they provide at Level 3, assuming it's used in the situations and conditions described, is the driver and that they are liable.

Those caveats seem reflected in what they're proposing (if the article is accurate), and it seems reasonable.

This would also solve issues like Tesla naming their level 2 systems 'Autopilot' and 'Full Self Driving'. I love my Model 3, but those names are deeply misleading at best and downright fraudulent at worst. The FTC really needs to step in and regulate what manufacturers can claim based on liability.

Only legally speaking. If the car drives me into an incoming truck, it still is very much my problem.

So currently it is only enabled for very clear situations, driving under 40 mp/h(60 kmh) in trafic jams on a motorway (where there usually is a physical separation to the other lane, except in roadbuilding situations, but I would guess, that the autopilot will be disabled then).

So it should be quite safe. Otherwise they would not accept the risk. German car makers are very conservative.

Edit: correct speedlimit

I hope/think they considered a pointy metal, like it sometimes is on back of trailors in front of you, stabbing through the windshield for example?

Maybe they can agree to pay any civil liability you'd have, but taking a nap after you turn a car on may open you to criminal liability that Mercedes can't intervene in.

If a knife salesman tells me his knives can't hurt people, and signs a contract saying he'll assume all liability, I am nevertheless liable when I stab someone with it.

The judge will say you damn well should've known you couldn't go to sleep in a car you're operating, and are in fact culpable for your negligence.

"But Mercedes said!"

"If Mercedes told you to jump (or drive) off a bridge..."

I imagine airplane manufacturers have long been on the hook if their autopilot engages Kill All Humans mode, despite the best efforts of Boeing's lawyers to blame some of the humans their software killed.

Airplane autopilots are so simple they they can be implemented with entirely analog circuits. They are simple feedback control loops that operate off a single scalar variable in each axis.

There is ample time for the pilot to correct any “mistakes” the autopilot makes.

There is no expectation that the autopilot operates independently of the pilot’s supervision and judgement.

In terms of decision-making and autonomy and liability - an airplane autopilot is really analogous to cruise control in a car, nothing more. It operates the flight controls to maintain simple parameters chosen by the pilot, and that’s all.

Comparisons of the systems is irrelevant to the very simple point I made - putting your money where your mouth is garners trust.

As for your claim about autopilot being simple, and humans having ample time to correct its mistakes, well, two 737 MAXs and one Air France flight are very straightforward counterarguments.

My biggest regret about getting it wrong is that it gives pedants something to fixate on and split hairs about, in lieu of engaging with the actual point.

For reference: https://www.398th.org/Research/B17_AFCE.html

From the article:

"...Handing over driving responsibility completely requires extremely particular circumstances. Right now, Drive Pilot can only engage at speeds under 40 mph (60 km/h in Germany) on limited-access divided highways with no stoplights, roundabouts, or other traffic control systems, and no construction zones. Eligible roads must be mapped by Mercedes for Drive Pilot use (similar to GM SuperCruise); the automaker has already mapped every such highway in Germany, and most of those in Nevada and California. The system will only operate during daytime, in reasonably clear weather, without overhead obstructions. Inclement weather, construction zones, tunnels, and emergency vehicles will all trigger a handover warning. And no, you can't close your eyes or go to sleep while it operates..."

No, technically whether they or liable or not reverts to normal principles of product liability, rather than their special acceptance of liability.

But manufacturers liability for harms caused by their products, while varying in detail from jurisdiction to jurisdiction, tend to be pretty broad in most of the West (incl. the US, despite it's otherwise weak consumer laws.)

As the default speed limit for such a highway would presumably be 65 mph (and traffic usually much faster), anywhere with such lowered speeds would presumably have special circumstances that would probably make it unsuitable.

The guaranteed 10 second handoff is impressive, however.

From the article:

"[...] Unlike all currently available driver-assist systems, Drive Pilot is designed to give drivers a 10-second warning before switching off; engineers had to make sure that, in every situation, the system would safely and faultlessly hand over control. [...]"

I think they discovered that it was the pilots who were accidentally leaning on the that switch when getting up for a coffee break.

Terrible decision to put it right next to an armrest...

Until it disengages. Not until you disengage it. So close, yet so far to go.

Unfortunately since this isn't a full self driving I don't know if they can do much better. If it hits the end of the road it needs to hand control back.

It's good for manufacturers to err on on the side of limited capabilities to keep things safer, but we don't want them to err too far in that direction. And if they're providing ongoing insurance with no revenue, that's an incentive to be overly restrictive.

Ok, I'm not insisting this is bad. But now what's the point of buying a car into your ownership instead of just calling it like a taxi when and where you need it?

It sounds like a good self-driving car is going to be really expensive, too, using technologies companies like Tesla are afraid to implement (e.g. LIDAR). Renting by the minute probably makes a lot more sense financially, too.

Bold move.

So you are in control of the car. Not the trajectory, but you control its velocity. And you are responsible according to tesla.

So you are not in full control of the car then. If a car decides to go in reverse when instead it should go forward and proceeds to hit a pedestrian that was tying his shoelaces behind it. Can you explain how that was an action of a person pressing a button that has no control over the trajectory of a vehicle and why we should consider them somehow responsible for that action (of backing up) instead of the car manufacturer?

> And you are responsible according to tesla.

That is irrelevant for this discussion.

You became liable as soon as you weren't paying enough a attention that someone could apparently sneak up next the car to tie their shoelaces.

What is the point of the feature if I have to come to the car and inspect that there are no blockages around it? Should I then walk to the place from which I intended to use the feature and use it or should I maybe just enter the car seeing that I am expected to circle the car before using the feature?

The rest of society is under no obligation to scale its expectations of your liability in proportion to you getting a perceived return on investment out of your expensive toy.

I.e. maybe you should always walk up to and around the car before using the feature in a public parking lot, and only make use of it to e.g. have the car back out of your own garage, provided that you have a clear view of your driveway being unobstructed.

...the software uses microphones and cameras to detect emergency lights and sirens far enough in advance to issue the full 10-second warning before manual takeover.

That doesn't help when the lights and sirens are switched on a few dozen cars away and reach you in 6 seconds.

Therefore they need to be very confident in that ten second rule, and naturally we look for cases where it might not work out well.

Nothing wrong with that as it is, but while it can be seen as a signal that "Mercedes is super sure their cars won't kill anyone" it should be seen more as a signal that "Mercedes is pretty sure they'll make a profit even when subtracting the price of killing people."

Just imaging actual emergency braking systems: you see an accidents and you decide that the best option is going against another vehicle or even a wall for instance to avoid crushing on a school group on a trip. Actual system do not and can't know that, they'll try to stop your emergency maneuver and due to the little time you have even if formally deactivate you can't deactivate them. Who will be responsible and how to prove what really happen?

That's the issue of automation. I can't trust back box logs from a proprietary black box designed, built and installed by the same company who sell the car. Who can? Even public authorities how much practical resources can have to really investigate?

Ah, so maybe this would be great for getting around town.

> ... on limited-access divided highways with no stoplights, roundabouts, or other traffic control systems...

Oh, so not where the majority of driving happens.

> ... and no construction zones.

And none of Pennsylvania.

Mercedes must really think it has a better system than its competitors.

Either that or these expenses are going to be considered part of its marketing spend?

Another thought: isn't it still usually the responsibility of the at-fault party's insurance company to pay out for crashes anyway? Or, in no-fault states, each insurance company covers their own driver and passengers, so it's not really important who cases a crash.

I wonder if this announcement is legally meaningless in most cases...but I'm not expert in this area.

I think in 3-5 years everyone else will have caught up to tesla and we'll have a more level playing field, mainly in terms of the technology and AI. And "marketing" features like this help accelerate the closing of the gap.

And the person you hit. Unbelievable that we value these metal boxes on wheels more than human bodies.

Maybe a similar case is how business jets that used to always require two pilots, even tiny ones like the Learjet 23. Flying one of them alone might not only be against the regulations, but even be criminally negligent if you crash and kill someone. But newer business jets have more automation and are designed for single pilot operation, so operating it alone would not be negligent.

At some point a judge will decide.

That said, every material step forward is a great thing. If we started out in the autonomous world that we're going to end up in, the idea of humans getting behind the wheel of a vehicle would seem incredibly crazy.

Fascinating questions.

Volkswagen is the first traditional automaker showing large investments in software with CARIAD [1], however, according to employee reviews, it‘s very poorly managed (by people who have never worked in software).

[1] https://cariad.technology/

[0] https://developer.nvidia.com/drive

Does the system legally act as an "insurance policy" that will indemnify the driver against any damages claims from a third party? Does it have a monetary cap? (Car insurance policies vary, but might only cover $250k, which actually won't cover many multi-car accidents).

This is a step in the right direction - but they although they may take responsibility for any financial penalty or civil liability - but can't control what a criminal court decides.

If my car plows into a tree at 40mph and I get turned to mush that still sounds like a "me problem" to me.

The difference is that Mercedes will have to pay your relatives while Tesla will say "after 2 hours on autopilot the driver wasn't able to take over in a split second"

Neither company really cares about your life and wellbeing, but one is also risking their own money and brand image. I usually trust companies to be good at judging financial consequences

Unfortunately, no matter how much legislation you put in front of it, that's not how physics works.

If the car causes a crash, the car manufacturer has to be liable. End of story.

As long as the self driving system performs well, it should be a win for everyone!

And if it does, we'll just get a few enterprising folks delivering "hitman as a service" in no time at all (driving hacked Mercedes vehicles into people claiming no responsibility).

Where can one locate these specific highways that offer guaranteed legal responsibility? I seemed to have missed it in the article.

> “On certain highways, below 40 mph”

Is it stock market price, followers on Twitter, or another criteria?

My claim is that if you narrow it down to actual autonomous driving; lidar + camera + a big chip will beat camera + a very big chip for the foreseeable future.

I don’t know that anyone is shopping for cars based on legal liability of a crash. You are still in the car. You still are injured or die.