Well sure, it’s a good thing and the government made them do it this time. But if I don’t want this new “feature” I have no choice, unlike normal vehicle recalls. I really don’t want my car to be like my phone where the manufacturer sends updates that sometimes help and sometimes…make it worse. I really like having control over things I buy.
Attitudinally I’m with you. But it’s an increasingly quaint notion.
In many areas of our lives, we don’t buy “things” anymore. We purchase the right to participate in an ecosystem of stuff controlled by business and influenced by the public. Of which we each are but one member. We get physical possession of something out of the deal, but that something is useless without the ecosystem it’s part of. And is heavily influenced by what happens out there, even if they never directly tweak your device.
SDMB moving from vBulletin to Discourse installed nothing on your PC or phone. But utterly changed the experience of being here. You bought into going on that ride whether you liked it or not.
Resistance is (largely) futile.
But that’s the thing, and some automotive writers have been saying for a while – conceptually, a Tesla is more of a network device that’s on wheels and street-legal. Just like your “phone” is actually a pocket network device that happens to contain a function to make calls to the legacy telephone network. Your early-2000s Nokia did not undergo periodic updates every few weeks and you could keep using it for years as a phone/pager. Just wait until they get to the part where they begin saying “Models of this vehicle from before [year] will no longer be supported after this July”.
Oh, I get it. And I hate it. I’m not even 40 years old, and already I’m an old man yelling “Get off my lawn”.
Of course, much of the hardware in a Tesla is the same as on any other car. They have a CAN bus, they mostly use the same suppliers, they have a similar basic architecture with a split between infotainment and drivetrain computers, etc. Just about all cars these days have a cellular modem. The difference is that Tesla fully internalized this fact. If you do have all these computers, and a network connection, and all that… then it should act like a fully networked computer. Including remote updates and all that entails. For some reason, other automakers haven’t quite gotten there, even though their hardware is capable of it.
My Bronco gets firmware updates OTA, but I suspect that’s only for the “Infotainment” system. I doubt Ford can update the PCM OTA.
Perhaps not allowing over-the-air updates is deliberate to avoid malicious hacking?
Well, it might be deliberate in the sense that they aren’t confident that they can update car systems securely with their current software stack, and so don’t. But there’s no obstacle to secure updates except software. As far as I know, no one has managed to load unsigned firmware on a Tesla.
Also, if the software is sufficiently bad, the boundary between infotainment and drivetrain systems can be broken (and has been). So that’s not a perfect method anyway.
It’s a bold strategy, Cotton. Let’s see if it pays off for them.
sounds like they take a chapter or two out of the “catholic church managing scandals” playbook …
(send the problem two states away)
“Hitting” might be some unfortunate phrasing.
I think you’re thinking of Cruise and/or Tesla. Waymo has been much more cautious than its competitors.
… AND … you are completely right …
I’ll put a foot in my - sometimes - self driving mouth
BTW, they are everywhere downtown!
We went to a play a few weeks ago, and must have seen 5-6 on that trip alone. It’s still a shock to see them driving with out a person behind the wheel.
There’s a bill in the California legislature to grant local control to their own streets relative to autonomous vehicles. Or self-driving cars I guess I should call them.
I’ll take the robocars over the drivers on my side street any day of the week.
Hell yes. And I suspect that insurance companies will quickly join our mindset, accelerating the shift to robocars. Humans, on average, are crap drivers.
But to paraphrase Winston Churchill, all the alternatives are so much worse.
ChatGPT etc may be able to make up bullshit text that fools most of the people most of the time.
But controlling a vehicle in real time, and recognising any unexpected problem like a child running in to the road after a ball?
I don’t think we are very close to real artificial general intelligence yet. Will we get there by just throwing more and more processing power and memory at it, or is a paradigm breakthrough needed?
We live in interesting times!
That’s a matter of checking the data. According to a recent report from Waymo, they are significantly safer than human drivers under equivalent conditions.
They’re trying to be really good drivers under strict conditions, then expanding those conditions, rather than being lousy drivers who go everywhere and eventually learn how to be safe, like teenagers.