Live data from Hacker News

Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

cnbc.com

331–340 of 864 posts

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#331
post #298

Earlier quoted context omitted.

I'm not sure that the people who want one of those will want them if: * they are not the first (or in the first ~50 or so) to get there * they have a lot of money and a lot of income on Earth This is kind of a glory-seeking thing for people who haven't made a name for themselves otherwise. Rich people would want to have a return flight. That is not a recipe for making a ton of money.

think of it this way: you go there, work for a company, in a spare time, you find gold, raise capital to start mining it, got rich, build a huge house with swimming pools, and all the rich attributes, but on Mars. Mars will be colonized with great mixture of science and initiative people. Your kids will grew up in this environment, they'll receive best education and endless room for possibilities.

Sure buddy. You will be able to prospect for and mine gold in your spare time after your one-way trip to Mars, including building all of the equipment you need, or having it sent over by rocket. And you'll be able to own land there too, because homestead laws and stuff will definitely carry over.

If you are already rich on Earth, you don't need to take risks that insane to try. In other words, nobody who can afford a ticket to Mars will pay for one.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#332
post #178

Earlier quoted context omitted.

One can kill people the other not. Guess which one. >> "...The FSD Beta system may cause crashes by allowing the affected vehicles to: “Act unsafe around intersections, such as traveling straight through an intersection while in a turn-only lane, entering a stop sign-controlled intersection without coming to a complete stop, or proceeding into an intersection during a steady yellow traffic signal without due caution,…

> One can kill people the other not. No, that's not correct. Whether it can kill people or not is orthogonal to whether it's a true physical recall of the car or a software update.

I think it’s closer to the physical product recall: it’s a strong “everyone with our product needs to get it fixed” message which they’re doing to avoid liability and further damage to their reputation.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#333

Earlier quoted context omitted.

You'd be surprised how many people want a one-way ticket to Mars. I think Elon can make good money from that alone.

I'm not sure that the people who want one of those will want them if: * they are not the first (or in the first ~50 or so) to get there * they have a lot of money and a lot of income on Earth This is kind of a glory-seeking thing for people who haven't made a name for themselves otherwise. Rich people would want to have a return flight. That is not a recipe for making a ton of money.

Gosh we can only hope all the rich people shoot themselves to mars with no return trip though. What a dream that would be for the rest of us.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#334
post #84

Earlier quoted context omitted.

That's the thing about neural networks: any QA is going to be superficial due to their statistical black box nature.

Exactly. It's the same reason that no amount of unit tests can replace formal methods for safety-critical software, and we cannot apply formal methods to neural nets [yet].

That's not really true. Most safety critical software is tested without formal verification. They are just really really thorough and rigorous.

Formal verification is obviously better if you can do it. But it's still really really difficult, and plenty of software simply can't be formally verified. Even in hardware where the problem is a lot easier we've only recently got the technology to formally verify a lot of things, and plenty of things are still out of reach.

And even if you do formally verify some software it doesn't guarantee it is free of bugs.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#335

Earlier quoted context omitted.

I would suggest reading all the disclaimers and screens one has to review (or skip, at their own peril) in order to actually get access to FSD Beta. You would likely not believe a Class Action lawsuit is a cakewalk if you read those screens...

I'd love to see more instances of "this giant wall of text that no one actually reads absolves us of responsibility" tested in court.

Reminds me of this recent Live Nation suit that was thrown out because "buyers waived their right to sue: https://www.nme.com/news/music/live-nation-antitrust-lawsuit...

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#336

Earlier quoted context omitted.

Recall implies something being returned. This should be called a mandatory software patch, or something like that. The problem is Tesla owners repeatedly see "recall" to mean "software update", so this might lead to a lot of confusion if a physical recall is actually required in the future.

I agree. It’s like crying wolf, eventually you start ignoring it

Laypeople have an incorrect perception of what a recall actually means, especially when it comes to vehicles. The most important effect that comes along with an official vehicle recall is that the manufacturer has to fix the issue for you for free, or otherwise compensate you in some way for reduced functionality you may have paid for.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#337
post #264

Earlier quoted context omitted.

What? Black box testing has plenty of techniques: https://en.wikipedia.org/wiki/Black-box_testing Whether it's a neural network inside or not is completely irrelevant. That's why it's called "black box".

Practical neural networks operate in enormous parameter spaces that are impossible to meaningfully test for all possible adversarial inputs and degraded outputs. Your FSD could well recognize stop signs in your battery of tests but not when someone drew a squirrel on it with green sharpie.

Something a bit similar is clinical trial and it is accepted without problem.

You make a black box test on several thousands (sometimes only hundreds) patients, and if patients who received the drug perform better the patients who received the placebo, then the drug is usually accepted for commercialization.

Yet one isolated patient may be subject to several comorbidities, her environment could be weird, she could ingest other drugs (or coffee, OTC vitamins or even pomelo) without having declared it. In a recent past women were not part of clinical trials because being pregnant makes them very "non-standard'.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#339

Earlier quoted context omitted.

Are you planning to join a class action lawsuit so you can get your money back?

FSD Beta is a free[1] opt-in beta. You have to basically drive like a grandpa for a few months to even be eligible. They give you a driving score, and if you take all the fun out of driving a Tesla, then you might become eligible for FSD Beta. I spent months trying, and never got my driving score to the point of qualifying for FSD Beta. I think you need to have of a score of 98 or 99 (and I was in the 70s). [1] The B…

'only' ?

That's a significant amount of cash for features that I would likely never use.

Re: Tesla recalls 360k vehicles, says full self-driving beta may cause crashes

#340

Earlier quoted context omitted.

This is most likely a legislative issue with NHTSA, I don't think they have a mechanism by which they can enforce a software update since the concept didn't exist when recalls were first implemented.

This is the correct answer. The question is, how do we get better at regulating reality faster?

This is the correct answer. The question is, how do we get better at regulating reality faster?

I don't think "Use words that put a more positive spin for Tesla PR" is something regulators should be working faster on.

Post reply on HN