[devwastaken]

Previous discussion was flagged - Self driving cannot work with the existing insurance and civil systems in place. By definition it can't be the driver that is solely responsible if the vehicle has automated driving. You can argue the driver was negligent in not stopping in time, not paying enough attention, etc. However civil isn't a blame it all on 1 person situation. Blame is distributed, as are the costs. So perhaps the driver is 50% responsible while autopilot is 40% with the other driver at 10%.

Especially with a name like "autopilot" the system is ripe for litigation. We would need new laws that specially protect automated driving if we wish to continue it. I imagine some cities and states may do it, but not others.

[x86x87]

Or here is a revolutionary idea: get the fuck off public roads with your beta poorly tested tech and come back wheb it's truly self driving and you accept full liability for whatever it does.

[RecycledEle]

One issue: To get a "good enough" self driving system, it needs millions of miles of testing on public roads.

[atoav]

Seems like you are pointing out one obvious challenge to any company that wants to develope such technology.

Maybe it is my european upbringing, but this is not my problem, but the problem of said company. I don't see why they should be allowed to test on public streets.

[enslavedrobot]

You should care because it will save the average western family ~$5k/year and reduce the 1.5million fatal accidents and >10million serious injuries that result from car crashes worldwide.

Just the insurance savings alone could give up to $800 billion back to rate payers in the US.

[atoav]

And how would that fare against something like decent public transport?

Let me guess, cars are king that's why we should sacrifice our kids to their cold hard steel bodies. /s

No seriously. This saving money doesn't mean they should be allowed to cut corners. If you truly want self driving you should be behind not using public roads as experimental tracks (with all the fallout that could come from that).

[enslavedrobot]

The saving money comes from saving lives, not the other way around.

Public transport and car sharing are two sides of the same coin. They approach the same solution from two different avenues. Public transport does not address the "last mile" problem of getting to the bus stop. Solving this issue would require more stops and smaller busses with complex adaptable routes. This is cost prohibitive if you have to have drivers. Car sharing is expensive and inconvenient because cars often end up parked in places were they are not needed and parking itself is expensive. Self driving cars (which will naturally evolve to mini busses) solve all these problems. No parking, no making grandma, and expectant mothers walk a mile to the bus stop. Everyone gets a cheap chauffeur service. And massively reduced road fatalities.

[JohnFen]

You say this as if it were fact, when it's really speculation of a possible future. Not unfounded speculation, but speculation nonetheless.

In the meantime, in the here and now, these things are obviously hazardous and shouldn't be on the public roads. Even if they would lead to a better future, no company has the right to endanger unconsenting others now to accomplish it.

[enslavedrobot]

If they are shown to be endangering people you have a point. If some drunk asshole plows into someone on autopilot I don't put that on Tesla.

No one even bats an eye when this happens on cruise control, and no the name autopilot doesn't make people think it's okay to be inattentive while driving. I have a Tesla and the number of nags warnings beeps admonishments you get is more than enough to let you know to pay attention.

[freejazz]

>You should care because it will save the average western family ~$5k/year and reduce the 1.5million fatal accidents and >10million serious injuries that result from car crashes worldwide.

When? What's the social cost to affording these companies public spaces to test their products?

[enslavedrobot]

Drivers are testing the product. I use FSD all the time. It's my choice.

If think that's making you unsafe too bad for you. I think it's vastly more unsafe to share the road with a million half blind geriatrics but nobody suing the AARP.

[olyjohn]

Self driving barely works in the US as it is. Go to a country where there are billions of people and shoddy infrastructure. Those places will take even longer to get self driving cars. And as it is now, those places are overcrowded with vehicles, and many of them have the majority of people on scooters and motorbikes. If all those people end up in cars, it'll be the end of those places, as congestion will be completely unbearable. There just isn't room for more cars in most places around the world. So I'm not sure how self driving cars will end all accidents world wide, other than clogging the roads up so badly, that nobody will want to go anywhere.

[x86x87]

So? Either accept liability and/or be way more conservative when making decisions or you don't belong on the PUBLIC roads. They are literally gambling with people's lives.

[ethanbond]

Waymo seems to be progressing much faster and at much lower risk to the public.

Tesla's approach specifically is just hubris/greed with the commons footing the bill.

[x86x87]

It's almost like they actually are thinking about safety.

[bdcravens]

Tesla could buy land, develop it with roads, and build a fleet for testing. They can get to multiple 9's of reliability before the need for public testing, over the course of a few years and tens of thousands of testing hours. Obviously, that would have a worse effect on the bottom line versus charging customers $10-15k for the "privilege" of being beta testers.

[tstrimple]

Or they could test it like Waymo is. I don’t ever see complaints about them on HN and it’s my understanding (but to lazy to verify right now) that they have one of the strongest track records on safety. Then again it shouldn’t take much to beat the safety of releasing it while lying about its capabilities and turning it loose on an unsuspecting public.

[x86x87]

Waymo's tech is way better and they do think about safety vs Melon where everyone seems to be a gimmick geared towards drawing attention.

[JohnFen]

If that's actually true and there's not another way, then tough. They shouldn't be on the roads.

[mminer237]

Most states already use a comparative negligence system exactly like you describe: https://usaccidentlawyer.com/blog/comparative-negligence-for...

[imtringued]

Musk can't even get "autopilot" to work in a closed system like his "Vegas Loop". Why would anyone expect it to work anywhere else?

[me_me_me]

and that's just the beginning. Heavy rain, snow, fog. There are tons corner cases and yet they couldn't have something working in a continuous loop

[chpatrick]

If it's your car and you turn on the autopilot button why shouldn't you be responsible?

[atoav]

Because you have been given the marketing material that it is safe.

Now let's continue this thought. Which company, do you think will produce the safer vehicles: The one liable for the faults of it's technology or the one not liable?

Now maybe the interesting question is why so many people shill (against their own interest) for the company known for the least responsible way of dealing with the matter.

[freejazz]

Because you don't control what the autopilot system does. Tesla does, and as the manufacturer of the product, is the party in place to minimize the risks it creates. Most especially in this sort of context where they have prior warnings as to the product's failings, and neglect to provide warnings or to fix the problem.

[chpatrick]

Of course you have control, you can not turn it on and drive safely like everyone else.

If someone sold you some kind of funky DIY autopilot system on ebay that turned the wheel for you and you decided to put it in your car and go in traffic, are you not responsible just because the seller assured you it's safe?