The future of car sales/resale

There are no laws saying specifically that self driving cars are the liability of the manufacture in an accident.

And there are no laws saying specifically that a failure in adaptive cruise control that causes an accident is the liability of the manufacturer, either. That’s because the laws governing liability for motor vehicle accidents are exceptionally generic - they’re all just products liability torts. The same laws govern cars and hair dryers and power tools and everything. If you make a product and the product ends up hurting someone because it malfunctions, you (as the manufacturer) are liable.

Legislatures don’t adopt new liability laws each time a new product comes to market. You don’t need a law specifically addressing liability for accidents caused by self-driving cars. There might end up being some, of course. But the absence of any specific laws in no way operates as a barrier to companies selling (or advertising) such vehicles.

This is not hard, and it’s already happening. You may be aware that an autonomous vehicle that was built and operated by Uber struck and killed a pedestrian in Arizona a few years back, when they were operating their pilot program. Uber ended up having to pay damages to the family of that pedestrian. There was no confusion - at all - about liability.

As I am saying nothing will be labeled a self driving car. They are here already.

Can you be more specific? Is there a specific car that you believe is self-driving? Or even a manufacturer that you can identify as making one?

Albaby

1 Like

And if it’s the car, and not the driver, that causes the accident then the manufacturer is liable.

Extending that to level 5 autonomy, manufacturers would be responsible for every accident, since there is no driver.

That suggests to me that either there will be close to no accidents with level 5, or there will be no level 5 vehicles.

—Peter

So we are back to what my nephew said in the first place, the laws are not in place for self driving cars.

You are the lawyer. Tell me if no one is driving the car and a major accident with three dead…who is to blame?

If you say the car company, then project that across the entire driving public using the same brand of manufacture. That is an unsustainable model.

It relieves the car owner of having to carry much insurance. It ruins the manufacturer.

Move away from the torts. Look at the economics as the companies like Tesla are doing.

Nothing will be called or labeled as self driving. There will be endless disclaimers that the driver must guide in the car at all times.

Extending that to level 5 autonomy, manufacturers would be responsible for every accident, since there is no driver.

That suggests to me that either there will be close to no accidents with level 5, or there will be no level 5 vehicles.

Why?

From an economic perspective, it makes sense to switch to Level 5 vehicles even if there are still lots of accidents - as long as there are fewer accidents than before. All the autonomy does is shift direct payment of the insurance burden from individuals to manufacturers.

You can think of it as simply switching who the “driver” is. Currently, the driver is a human who buys their own auto insurance. With AV’s, the “driver” is the AI, who has auto insurance provided by the manufacturer. That insurance is provided the same way that the manufacturer provides ‘insurance’ against claims caused by manufacturing or design defects - either self-funded reserves or their own claims insurance.

If the rate of accidents with the AI driver is lower than the human driver, the cost of insurance among all parties in the Level 5 scenario is lower than with human drivers. So there’s no economic barrier to switching, even assuming that there are still a large number of accidents even after the switch.

Note: despite, the above, we’re probably not going to see regulatory approval of Level 5 driving unless it’s very much safer than “average.” That’s because a very large proportion of accidents are caused by certain groups of individuals, specifically teens and intoxicated people. Before regulators approve Level 5, it’s going to have to be at least as safe as the “modal” driver - a sober, non-teen driving under generally normal weather conditions. The rate of accidents for this type of driver in those conditions is far, far lower than the overall rate (which is why Tesla’s results aren’t very helpful), so the Level 5 system will end up reducing accidents by a lot before it gets approval.

2 Likes

Extending that to level 5 autonomy, manufacturers would be responsible for every accident, since there is no driver.

That suggests to me that either there will be close to no accidents with level 5, or there will be no level 5 vehicles.

The manufacturers could just insure, right? That’s basically how it works now.

As I understand it, a potential hang-up with self-driving liability is the trolley problem. That is, a situation could arise when the AI needs to decide who dies in an accident. For example, does the AI drive into a lamp post killing the driver, or swerve and take out a pedestrian instead? What if the pedestrian is elderly vs. a teenager? In those types of situations, it is possible the AI functioned exactly as designed, yet someone still died.

I am not a lawyer, but it seems to me there needs to be a legal framework or at least a set of industry best practices for how the AI is supposed to function in those types of situations.

Tell me if no one is driving the car and a major accident with three dead…who is to blame?

The manufacturer. (For simplicity, we’ll assume that there’s only one manufacturer that makes the AI, the car, the sensor equipment, etc.)

If you say the car company, then project that across the entire driving public using the same brand of manufacture. That is an unsustainable model.

Why? Manufacturers are already liable for 100% of the accidents that are caused by defects in their cars. If there’s a major accident with three dead that results from a defect in the brakes, or the cruise control system, or some other component of the vehicle the manufacturer is liable for it today. Yet that’s perfectly sustainable. Why do you think it wouldn’t be sustainable in a self-driving scenario?

There will be endless disclaimers that the driver must guide in the car at all times.

Again, I think you’re mistaken in thinking that these are merely disclaimers. Tesla’s Autopilot and FSD are not self-driving. Tesla isn’t just telling drivers that they have to pay attention and guide the car at all times in order to minimize liability. They are telling drivers that because these are only Level 2 systems - they are not capable of driving themselves without the driver being focused, attentive, and ready to take intervening actions when necessary.

No one is selling a self-driving car today. That’s not just marketing. The cars aren’t self-driving yet. They are only driver-assistance systems. They can do some things for the driver, but they cannot do enough for the driver to make it safe for the driver to stop driving.

Albaby

4 Likes

The rate of accidents for this type of driver in those conditions is far, far lower than the overall rate, … so the Level 5 system will end up reducing accidents by a lot before it gets approval.

Which is basically what I meant.

If level 5 AI can’t drive better than a teen or a drunk, it won’t make it out of the lab. The product liability cost would likely be too high.

If it can drive as good as a typical good driver, then we’ll get it and there will be significantly fewer crashes.

—Peter <== who recently had a teenaged Tesla driver smash into his parked car, causing several thousand dollars of damage to his car. Also causing him to not be all that impressed with Tesla’s driver aids.

“Note: despite, the above, we’re probably not going to see regulatory approval of Level 5 driving unless it’s very much safer than “average.” That’s because a very large proportion of accidents are caused by certain groups of individuals, specifically teens and intoxicated people. Before regulators approve Level 5, it’s going to have to be at least as safe as the “modal” driver - a sober, non-teen driving under generally normal weather conditions. The rate of accidents for this type of driver in those conditions is far, far lower than the overall rate (which is why Tesla’s results aren’t very helpful), so the Level 5 system will end up reducing accidents by a lot before it gets approval.”

Matter how ‘safe’ your car is, as long as there are drivers not using Level 5 driving…there will continue to be accidents, as people pull right out in front of you, leaving no room to stop, or they tail end you. Where there is bad weather, the idiot behind sliding into you will still slide into you on icy roads, and despite your Level 5, you’ll still have problems with icy spots and icy roads, or deer deciding to cross the road at 50 mph right over your hood and windshield, or in front of your self driving car as it bounds out of the woods 15 feet from the road edge.

Now it would be great for drunk people, if you can keep them from turning OFF the system to get them home alive and accident free.

Luckily in 50 years of driving, I’ve had no two vehicle accidents. Got hit by deer twice - once $1000 damage as deer on first day of hunting season in afternoon bounded out of woods, over my hood, almost leg through windshield in MO. Got hit by baby deer after I slammed on brakes at 5 mph or less. Broke plastic grill - easy to fix. Hit and run driver took off side mirror on car one night late in downtown Dallas as he sped past me using left turn lane and continued on. Got his plate but was stolen vehicle from out of state. No other damage.
Driven well over 700,000 miles…

then again, my sister in suburban MD had 3 year old car totaled as idiot tail ended her at 40-50 mph. Niece had 5 year old car when ladder fell off truck ahead of her at 50 mph on Beltway in DC and she couldn’t avoid it. Little Geo Metro without sufficient ground clearance. In neither case would ‘safe driving car’ have helped. Have half dozen friends whose cars have been done in by deer. But deer don’t sue. People still get injured though in the cars.

I’m getting old enough where it would be nice to have self driving car for evening/night driving. I don’t do much at night these days out far from home. Especially in rain/fog weather!

The problem might be to keep those teens from TURNING Off the systems and still driving like idiots.

t.

1 Like

The manufacturers could just insure, right? That’s basically how it works now.

Of course. But if their cars cause too many accidents, the insurance will be prohibitively expensive and the manufacturer will have to choose whether to improve their car or scrap it.

Or take the Shinyland alternative of paying the CEO massive bonuses for skipping the insurance and saving the company lots of money, then spinning the self driving business off into a standalone company, again earning the CEO a big bonus, only for the spun off business to ultimately fall into bankruptcy from all of the product liability lawsuits. That allows the CEO to buy back the technology out of bankruptcy for pennies on the dollar, again earning a big bonus with promises of fixing the technology.

Oh. Sorry. I slipped into cynic mode there for a bit.

—Peter

2 Likes

—Peter <== who recently had a teenaged Tesla driver smash into his parked car, causing several thousand dollars of damage to his car. Also causing him to not be all that impressed with Tesla’s driver aids.

Was the driver using Autopilot at the time or just driving through a parking lot?

Sorry, but is “him” the driver or you?

Mike

Try not to get to enamored of new technology. Remember when there was going to be no more paper because of email?

Remember when they said the ipod was going to replace Walkmans. Or that music streaming was going to completely disrupt how people buy music. Hey, wait minute…

Goofy, I get the feeling you still use carbon paper and rent VHS movies. You have to admit the transition to EVs is happening a lot faster than people envisioned just 10 years ago. Ford has already built a national network of 75,000 charging stations. And that’s Ford, best known for gas-guzzling trucks.

1 Like

Was the driver using Autopilot at the time

Unlikely. It was on a suburban residential street in front of my house. My car was parked on the street.

Sorry, but is “him” the driver or you?

I am the one unimpressed by the driver aids. The accident appeared to be a typical pedal mix up, mashing the accelerator instead of the brake. Apparently auto braking isn’t all that it’s cracked up to be, as she continued to accelerate after hitting my car, pushing it about 5 feet after the initial impact. That low speed torque from the electric motor is real.

—Peter

No one is selling a self-driving car today. That’s not just marketing. The cars aren’t self-driving yet.

It is getting closer though. A driverless delivery vehicle has been approved for use on city streets in much of the Silicon Valley.

The deployment permit grants Nuro permission to use a fleet of light-duty driverless vehicles for a commercial delivery service on surface streets within designated parts of Santa Clara and San Mateo counties, which includes the cities of Atherton, East Palo Alto, Los Altos Hills, Los Altos, Menlo Park, Mountain View, Palo Alto, Sunnyvale and Woodside. The vehicles have a maximum speed of 25 mph and are only approved to operate in fair weather conditions on streets with a speed limit of no more than 35 mph. https://www.dmv.ca.gov/portal/news-and-media/dmv-approves-nu…

Approval was also given for part of Houston TX. Uber Eats will initiate driverless delivery beginning this fall.
https://www.theverge.com/2022/9/8/23339830/uber-eats-nuro-ro…

The vehicles have a maximum speed of 25 mph and are only approved to operate in fair weather conditions on streets with a speed limit of no more than 35 mph.

If your drivers are anything like the the ones in our city, there’s going to be a lot of passing and/or pissed-off drivers stuck behind them.

It will be interesting to note any changes in accident rates. And complaints to Nuro and city officials.

1 Like

Remember when they said the ipod was going to replace Walkmans. Or that music streaming was going to completely disrupt how people buy music. Hey, wait minute…

I’m a big EV proponent and while I believe EVs are better than ICEs, they aren’t that much better. You can have a perfectly fine driving experience in an ICE.

But there are some limitations to EVs too. One is that they are expensive, but more importantly if you can’t charge where you normally park, like either at home or at work, they almost certainly won’t be practical.

Who can’t charge where they normally park? Lots of people. And some people legitimately need more range than a typical EV provides. That will change, but it will be incremental.

3 Likes

skye: “But there are some limitations to EVs too. One is that they are expensive, but more importantly if you can’t charge where you normally park, like either at home or at work, they almost certainly won’t be practical.”

The cost at a 15-20 minute charger is typically $0.48/KWH. To recharge a 35 or 50 KWH battery for a small/intermediate car - will cost you $20 for the 200-250 miles and that is only to 80% typically.

Right now, gas in TX is $2.99 at most local stations, less if you pay cash at some, and likely now down to $2.80/gal at Costco. So 200 miles of driving at 30 mpg is maybe 7 gal or $21. No big savings there for sure. If you have a truck with 50KW or more battery, $30 or $40 to juice it up to 80% at a fast charger.

Buy a hybrid that gets 35-40, or 55 for a Prius, and you beat the EV that has to charge at a ‘filling station’.


skye:"“Who can’t charge where they normally park? Lots of people. And some people legitimately need more range than a typical EV provides. That will change, but it will be incremental.”


Or need more towing capacity -

GM promised a $30,000 EV this fall or into next year. Chevy Equinox - 300 mile range - With Li prices up, might wind up $35K but if they can do it - watch those $55 and $60K EVs fall off a cliff in sales.

40% of Americans live in apartments - a lot of them in vertical spaces (multi story buildings) and those will take a while to be EV friendly - and I doubt the charging will be ‘free’. someone is going to have to pay off the initial cost of wiring up everything. 220V, 40amp circuits to hundreds of parking spaces. Keeping them repaired. Not wiping them out with snowplows. Etc.

t

Ford has already built a national network of 75,000 charging stations.

Is that true? I thought the Blue Oval charging stations were like a group of existing charging networks, like ChargePoint that have agreed on a billing method. These are primarily L2 chargers, which are great, but not for long trips unless at a hotel where you park at night.

Mike

I am the one unimpressed by the driver aids. The accident appeared to be a typical pedal mix up, mashing the accelerator instead of the brake. Apparently auto braking isn’t all that it’s cracked up to be, as she continued to accelerate after hitting my car, pushing it about 5 feet after the initial impact. That low speed torque from the electric motor is real.

Are there any cars that won’t go if you press on the accelerator?
The Tesla will warn you with an error message and beeps if you press on both pedals at once.
Since the car does have ultrasonic sensors all around that chirp when you are within ~2 feet of something I don’t see why they couldn’t stop the car or at least force the speed to 1 or 2 miles an hour when something is detected right in front…maybe this could prevent the pushing an extra 5 feet. :slight_smile:
The only downside (unless I’m mistaken) is that it might take you an extra few seconds to park sometimes.

Mike

1 Like

That is, a situation could arise when the AI needs to decide who dies in an accident.

Syke,

That is the play ground idea or theory.

The truth is the hardware can not stop all accidents in time. Whether it is the braking or the speed etc…the computer can not act in time…which goes to the radar etc…but my nephew does not want to directly own that. Al wont end one of this and my nephew wont own the other end of this. The car companies do not want to own it. The drivers are not being offered it. But it is here and they are driving on it.

They are only driver-assistance systems. They can do some things for the driver, but they cannot do enough for the driver to make it safe for the driver to stop driving.

Al,

This is another aspect of the folly. My nephew is discussing what is possible now. The car companies are avoiding it is all you are telling me by keeping it to driver assistance systems.