What I don't understand (and maybe I'm just naive) is how it's legal to test a so-called "beta feature" -- which, if it fails, clearly can be dangerous (lucky no one was hurt) -- on customers in a live environment, vs using its own test drivers at the very least? I can't imagine this being legal with any other form of transportation (or just large equipment; they don't even sell lightbulbs that aren't CE certified).
Yeah, you'll notice that aviation industry doesn't have an equivalent.
Which is not to say that none of their new features fail catastrophically -- as we were reminded recently by the 300 MAX -- but they don't consider this a valid defense.
the auto industry doesn't have this equivalent either, its just tesla. They'll eventually just claim everything is beta and take no responsibility for anything not working.
Of course Tesla tests things itself with its own drivers. What they can't do is replicate every single one of the unusual corner cases that people encounter in the real world.
Sure, but that’s true of all companies. Imagine a drug company put a drug on the market that might kill your or leave you crippled if you don’t match the results they had on their 3 test patients, and them telling you “That’s on you, the drug is a beta”.
That's a bad example, because drugs typically have all kinds of warnings about negative effects. They certainly test on more than 3 patients, and they won't tell you the drug is a beta, but I'm sure they will still claim they are not responsible, because they warned you about very small chance of failure.
All of us are making assumptions about how much testing Tesla does internally before deploying a feature to customer cars. I think it's way more than the equivalent of "3 test patients", but from the outside of course there's no way to know.
From the linked tweet, there's no way to evaluate how unusual the failure case was. And of course we don't know how likely it was to have been a pedal misapplication. We do know that that's not uncommon, though.
The original comment I replied to assumed that there was no internal testing, and you're assuming there was very little internal testing. Both of you are jumping to that conclusion with very little support.
Not as serious as death; but their argument suggests that the owners are subsidizing Tesla's R&D monetarily when their systems fail. I assume there is some type of legal agreement the owner had to click through to gain access to the feature. That is a pretty crappy situation for Tesla owners; hopefully, there is a way to opt out of agreements once opted in, given new information.
IMO this is pure FUD, no actual correspondence with Tesla to back it up.
It would be silly for Tesla to argue that it's driver's fault because "beta".
Every other car manufacturer will say it's driver's responsibility when using autopark, full stop.
Do people have any personal accountability anymore? "The company will protect me" is such a foolish thought process, especially when you're playing with beta features that have risks.
This to me isn't all that different than the guy who crashed his Tesla while watching a movie behind the wheel. Culpability. Assumption of risk. But-for causation. All the usual legal theories apply.
>Do people have any personal accountability anymore?
Yes.
The problem is that people shouldn't have to be accountable for failure-modes that they in no way have the training to predict, understand, or recover from.
The FDA explains the specific risks of a drug. They don't just say "This drug can cause issues.".
Hand-waving all corporate fault by admitting that your corporation is perfectly okay with beta software being on the public road as long as the car owner assumes responsibility is not a solution to the liability question.
A responsible beta program would 1) not be on the public highway in any means 2) require the participants to have a high familiarity with the product and the technical know-how to be effective members of the testing community; which should include training on not only error and incident reporting, but failure recovery and failure scope.
It's perfectly reasonable that a user would operate a mechanism incorrectly when their only advice is "Good luck, it's experimental!".
There are a few reporters who reached out to the poster via twitter -- I'd expect follow-up stories over the next couple days (with them at least asking Telsa for comment; no guarantee they'll respond, of course, but would really like to see what Tesla says).
23 comments
[ 3.3 ms ] story [ 57.3 ms ] threadWhich is not to say that none of their new features fail catastrophically -- as we were reminded recently by the 300 MAX -- but they don't consider this a valid defense.
Taking out of context by cutting off your sentence, but pretty much any drug has a listed side effect like this.
From the linked tweet, there's no way to evaluate how unusual the failure case was. And of course we don't know how likely it was to have been a pedal misapplication. We do know that that's not uncommon, though.
The original comment I replied to assumed that there was no internal testing, and you're assuming there was very little internal testing. Both of you are jumping to that conclusion with very little support.
It would be silly for Tesla to argue that it's driver's fault because "beta". Every other car manufacturer will say it's driver's responsibility when using autopark, full stop.
This to me isn't all that different than the guy who crashed his Tesla while watching a movie behind the wheel. Culpability. Assumption of risk. But-for causation. All the usual legal theories apply.
Yes.
The problem is that people shouldn't have to be accountable for failure-modes that they in no way have the training to predict, understand, or recover from.
The FDA explains the specific risks of a drug. They don't just say "This drug can cause issues.".
Hand-waving all corporate fault by admitting that your corporation is perfectly okay with beta software being on the public road as long as the car owner assumes responsibility is not a solution to the liability question.
A responsible beta program would 1) not be on the public highway in any means 2) require the participants to have a high familiarity with the product and the technical know-how to be effective members of the testing community; which should include training on not only error and incident reporting, but failure recovery and failure scope.
It's perfectly reasonable that a user would operate a mechanism incorrectly when their only advice is "Good luck, it's experimental!".
Beta software that has safety implications on a car seems strange.