[belter]

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,” according to the notice on the website of the National Highway Traffic Safety Administration..."

[xdavidliu]

> 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.

[indigodaddy]

I think that it’s more about using the most appropriate known terminology in order to try to get the most people to do the needful. “recall” sounds more urgent/dire than “software update”, and will likely encourage many more people to take action vs using “software update” or some less familiar terminology. The word “recall” in terms of autos has built up a lot of history/prior art in people’s minds as something to really pay attention to. I have no idea, but perhaps that is why they are going with this known terminology.

[slg]

The whole point of over-the-air updates is that the owner doesn't need to do anything. For example, both Tesla and Toyota have had bugs in their ABS software that required recalls. The owners of the Toyotas had to physically bring their cars in to get the software update which slows down the adoption drastically. The Teslas received the update automatically and asked for the best time to install the update the next time the owner got in the car.

There are really two issues here. The FSD and the OTA updates. Let's not throw out the baby with the bathwater and blame OTA updates just because Tesla's FSD software is bad. The OTA updates do provide an avenue to make cars much safer by reducing the friction for these type of safety fixes.

[adamjcook]

> The OTA updates do provide an avenue to make cars much safer by reducing the friction for these type of safety fixes.

True, but let us also acknowledge the immense systems safety downsides of OTA updates given the lack of effective automotive regulation in the US (and to varying degrees globally).

OTA updates can also be utilized to hide safety-critical system defects that did exist on a fleet for a time.

Also, the availability of OTA update machinery might cause internal validation processes to be watered down (for cost and time-to-market reasons) because there is an understanding that defects can always be fixed relatively seamlessly after the vehicle has been delivered.

These are serious issues and are entirely flying under the radar.

And this is why US automotive regulators need to start robustly scrutinizing internal processes at automakers, instead of arbitrary endpoints.

The US automotive regulatory system largely revolves around an "Honor Code" with automakers - and that is clearly problematic when dealing with opaque, "software-defined" vehicles that leave no physical evidence of a prior defect that may have caused death or injury in some impacted vehicles before an OTA update was pushed to the fleet.

EDIT: Fixed some minor spelling/word selection errors.

[slg]

This is a totally fair response since I didn't say that directly in my comment, but I 100% agree. OTA updates are a valuable safety tool. They also have a chance to be abused. We can rein them in through regulation without getting rid of them entirely because they do have the potential to save a lot of lives.

[adamjcook]

I agree.

[caf]

It'd probably be just as effective to require that every version of the car software that is made available to the fleet must also be provided to the NHTSA. There's no sweeping shoddy versions under the carpet then.

[buffington]

> The word “recall” in terms of autos has built up a lot of history/prior art in people’s minds as something to really pay attention to

Tesla didn't choose the word "recall." The legal process known as "recall" chose the word. It's not like people at Tesla debated over whether or not to call it a "recall" instead of a "software update."

If Tesla had it their way, they'd have quietly slipped it into any other regular software update alongside updates to the stupid farting app, if they cared to fix it at all.

When a company issues a recall, it's because there's pressure from regulators, or investors, or both, and/or a risk of class action lawsuits and fines. Using the word "recall" isn't a preference or even a synonym. It's a legal move meant to protect them.

If Tesla gets sued over a flaw, "we issued a software update" isn't legally defensible. "We cooperated with official government bodies to conduct a recall," does because a recall describes an official process that requires manufacturers do very specific things in specific ways as prescribed by law. In exchange, manufacturers are legally protected (usually) from lawsuits related to that flaw.

[adamjcook]

It is the terminology that exists in US automotive regulations (what little there effectively are).

A "recall" is just a public record that a safety-related defect existed, the products impacted and what the manufacturer performed in terms of a corrective action.

Additionally, I believe that the possibility exists that Tesla must update the vehicle software at a service center due to configuration issues. Only a small number of vehicles may require that type of corrective action, but the possibility exists.

Historically, there exist product recalls (especially outside of the automotive domain) where the product in question does not have to be returned (replacement parts are shipped to the impacted customers, for example).

[mcguire]

No, really, this is true. It has nothing to do with how the defect is fixed.

https://www.nhtsa.gov/sites/nhtsa.gov/files/documents/14218-...

https://www.law.cornell.edu/cfr/text/49/573.6

(Except for tires.)

[adamjcook]

Hmm. Perhaps I should have read the parent's comment more carefully. I think that I might have misinterpreted it.

You (and the parent comment) are correct.

My comment was not intended to argue that a recall prescribed a particular corrective action.

[freejazz]

> "true physical recall"

ah, a made up term in order to justify your point. how convenient.

[Daz1]

No we're talking about making terms actually fit their definition, which is generally helpful.

[freejazz]

If it actually fit the definition then why would you need to add "true" and "physical"?

[myko]

> whether it's a true physical recall

I hope by participating in this thread you're aware by now but just to be clear there is no "physical" recall necessary. The recall is about documentation, customer awareness, and fixing the problem. "Physical recall" is meaningless and unimportant, it's not what "recall" means at all.

[chc]

And I hope you see that you've demonstrated why "recall" is a poorly chosen word for that, since the word's normal definitions have nothing to do with "documentation," "customer awareness" or "fixing the problem."

[acdha]

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.

[cptaj]

You're right. A physical recall doesn't necessarily imply death.

This should be labeled "holy shit need to fix this now, people could die"