I’ve never spoken one-on-one with Sergey Brin or Elon Musk so I don’t know what they really think about autonomous vehicles. I can only give you my opinion.
My experience has been that people with technical knowledge and understanding think that autonomous vehicles are being tremendously overhyped and are something for the distant future if ever while non-technical people think they’re just around the corner or even already here. That leads me to conclude that Mssrs. Brin and Musk know that and don’t really care whether people are killed by prematurely autonomous vehicles. They just want to get in on the gold rush.
I repeat my proposal: strict liability. People who are injured by prematurely autonomous vehicles or the families of those killed by them should be able to recover from their developers regardless of intent or the ability to prove specific responsibility. All that should matter is the injury. Injured by an autonomous vehicle? You should be able to recover from Google (or Tesla or Uber).
Auto manufacturers are sued now all the time, what’s the difference?
BTW, I don’t know anyone who would ride in one right now, have to ease into it. (Like cruise control), (Like anti lock brakes),(airbags). Some people don’t even trust seatbelts.
Their liability is limited. Plaintiffs need to prove reckless disregard.
I honestly think I’d be dead today if there were no driver in the semi trailer behind me last fall. It was an odd situation, all westbound traffic was at a standstill on interstate 80 north of Lincoln, me at the back, the truck came over a small hill, (only kind we have), saw the situation and locked up his brakes. In my rearview, I saw smoke rolling off all eighteen tires, knew he’d never stop in time. He knew it too, let off his brakes so he could steer, and went to the right shoulder, hit his brakes again and went a hundred yards before he got it stopped.
Now, what would a driverless truck have done?
Stay on the brake and kill everybody is what I think.
Sellers and manufacturers are strictly liable for injuries caused by defects in their product, including parts. An example of a defect is where an alternative design could have prevented the injury. I see a lot of discussion about how the software uses algorithms that better make trade-offs than the human driver. This is not within my area of experience at all, but that seems to be a framework that would fascilitate finding a defect: if the safety dial can be shifted from 5 to 6, then an alternate safety design is available. Human drivers make mistakes in the moment (negligence), while products liability deals with designed systems and the decision making process is not really comparable.
Also, plaintiffs tend to sue in courts where the product was transported and are known to be plaintiff friendly.
I tend to think the opposite is going to happen; the technology will only be truly marketable if liability issues are federalized. Not only because I think product liability cases will find a path to success, but there will be a need to standardize roads and road usage laws (jaywalking is one of the most localized laws there are).
@Gray, the software would have perceived its limitations and switched control over to the sleeping driver.
Well, Elon deleted all his firm’s Facebook accounts – so there’s that.
I think it will be terribly difficult, maybe impossible, to prove defect in artificial intelligence.
I’m sure jurors from the poorest parts of the country will get to the heart of the matter.
At the very least, the system of laws and local custom are going to be way too messy. And it won’t be the fatalities, the real money is in permanent, non-fatal injuries.
Yes, PD, that’s why I think that although the actual technology suggests that the actual feasibility of fully autonomous vehicles is somewhat farther in the future than many seem to think the liability issues will probably prevent them from being used other than in very narrow niches.
However, while we’re on the subject why aren’t railroads fully autonomous? We’ve had the technology for that for decades.