81 comments

[ 2.3 ms ] story [ 170 ms ] thread
> Connected cars aren't just the future, they're the present.

Been the present for 10+ years

Taking on the persona of a boring consumer, who has had to go into the dealer to get several updates with their unconnected car, and sometimes charged for it ($200 to $700), I don't see this as bad, if it's done well.
OTOH i'm well pleased i can pay $200 for a new memory card for my in-car navigation and have 0 reliance on cellular service or some cloud provider that might not exist in a few years

the newest toyota's seem to lack functionality inside a parking garage

That doesn't strike me as a tradeoff based on a realistic assessment of the likely outcomes. For example, VW issued updates for the 2012 GTI only until 2018. Meanwhile, even if Google or Apple were to stop operating their maps services, you could switch from one to the other, or you could switch to OSM or whoever steps in.
(comment deleted)
As far as I can tell, literally nobody is doing it well.
In a way this is a blessing in disguise, because the moment a large car company thinks they can do software well, they'll want that team to be a profit center not a cost center, which means they'll start charging for subscriptions.

There's an argument that we're at Peak Mediocrity for car systems that one can own (at least at the levels that we think we own our cars now).

Orrrr perhaps they could just make their software update files publicly available on their website, so I can download it to a flash drive and stick it in the car myself? That's what the dealer is being paid $200-$700 to do, after all.
That makes it much harder to collect all sorts of useful statistics about the cars use, which is profitable, which is why all the manufacturers will eventually go with OTA.
It's telling that the hardware team could create a whole locking diff faster than programmers can type into a computer.
[dead]
Isn't the cybertruck 3-4 motor? In which case the "locking diff" is really a software feature since there wouldn't be anything between the axles to "lock?"
There is a two and three motor version so one or more diffs are needed.
I think all the units so far are 3 motor (2 motors in rear) and therefore the rear differential lock is probably (planned to be) implemented by software that keeps the two motors in sync instead of with a physical mechanism.
The rear would be a virtual or simulated diff then? Do we know this to be the case or can the driver only lock the mechanical real diff? I can "lock" the center diff in my car but there are two still two other diffs. Do they mean lock in the absolute sense as in a 4x4 or fake lock as in my AWD?
Are you allowed to recognize revenue for features that are coming soon?
(comment deleted)
In my country it could be argued it's just a deposit and thus not taxable income.
In the case of tesla they have a big bundle of cash sitting somewhere that they can't recognize until they ship full self driving/fsd, and they are probably never going to ship that imho
Could you describe what does it mean "can't recognize"? Where I live you can use deposits as you like...
Somehow due to accounting rules in the US, companies can take payment for software before it is done, and they can't count it as revenue somehow. So it sits there waiting for the feature to be done. I've seen it at big software companies in the us and at tesla (which is kind of a software company too). Found this random person on reddit who discusses it. Who knows if it is accurate but it matches what I've read elsewhere. https://www.reddit.com/r/investing/comments/10qpasz/tesla_co...
Seems like the same thing, basically. So this means they can use the money but don't have to tax it yet.
I'd venture to say that this locker function being delayed is a good thing. It'll force people to learn how to drive their truck off-road instead of hitting the Locker button, getting themselves stuck even worse, and then blaming Tesla.

[HEAVILY edited for editorial clarity]

Can you explain how any of this is related to a discussion on making expected features of a vehicle “coming soon”?
Parent says:

> I'd venture to say that this locker function being delayed is a good thing.

I edited my comment to lead with that, because that's the main point. Thanks for helping me see that.
Every offroad comparison I've seen of the Cybertruck uses the same driver for all vehicles being tested. If a driver (even an inexperienced one) can make it up a small hill with a Subaru, but not a 4x4, locking diff, tri-motor, Cybertruck what does that say?

You can't bolt on 4x4 engineering skills.

It tells me that the involved driver doesn't know how to get the most out of the Cybertruck... probably. A Subaru of any size is a much lighter vehicle that's going to handle completely differently. Getting out of that and into a Cybertruck would be like getting out of a Corolla and into a Suburban. Just completely different vehicles.

That being said, Subaru is renowned for being a highly capable off-roading car. Cybertruck is highly renowned for nothing so far :D

I used the Subie as a placeholder for at least a half a dozen different vehicles that I've seen besting the Cybertruck on X (of all places)

> Cybertruck is highly renowned for nothing so far

Exactly. Yet the Tesla fanboys crowned it offroad king from the jump!

Buy hipster things, have hipster issues...
Alright folks, pack it up. The word "hipster" has officially lost every ounce of its original meaning.
Hipster used to mean you bought things before everyone knew they were cool. Now it means you bought things after everyone thought it was cool (but by now they know better).
This truck is going to be a complete disaster in every conceivable way for Tesla. You've got a TAM on that thing of like.... maybe a few million people on earth? And even then, it's just going to be a silly short-lived fad for the ones who do. They more or less did the same idiotic thing as Ford with the Lighting. Tease a "$35k" starting MSRP for all the publicity, convince themselves they had a hit, and overcommit on manufacturing resources accordingly. Then, when the true out-the-door price becomes double that, act all surprised pikachu when no one shows up to buy.

Add to that the complete torching of any Tesla brand value left by Elon over the last couple years, and things are looking pretty grim over there. Stock at it's lowest point since joining the S&P 3 years ago, and sales growth hitting a brick wall.

Very bearish take. A complete ROI is like half a million sales, and they still have two million preorders.
I don't think they'll ever sell many of time. It's just not that compelling of a vehicle, the other ev trucks (rivian and f150ev, plus some new ones coming reasonably soon like chevy silverado) look like more practical vehicles, if you actually need one. Tesla has that great drivetrain but when superchargers open up to the others, I really think there's no reason for the ct to exist. I guess it's great that it's a low volume 48v testbed.
They need to convert less than 1/3 of their preorders to make a successful truck. After this it's all profit.
If I worked 3 jobs as a software developer I'd probably be fired for "moonlighting" from all 3 of them. Elon's attention is ridiculously split between Tesla, SpaceX, and Twitter/X (oh and maybe X.ai, whatever that is). If I was a Tesla shareholder I'd be pissed.

But then again, this truck seems like it might have been Elon's pet project - so maybe it's better that he's less involved in Tesla? Who the hell knows. A lot of tradeoffs were made in favour of aesthetics and marketing at the cost of functionality with this one[1].

[1] https://www.businessinsider.com/tesla-cybertruck-manual-road...

Also Neuralink and Boring.

Tesla shareholders don't have much room to be mad about this, though. His attention's been split for a very long time, so either this was something that you knew would be the case when you bought Tesla stock, or you've made a vast amount of money off Tesla despite it.

> Then, when the true out-the-door price becomes double that, act all surprised pikachu when no one shows up to buy.

Double that plus an additional two year delay, while Tesla sells to the customers willing to buy the $80K+ and $100K+ models.

This isn’t an answer or even a disagreement. But don’t they have the #1 best selling car in the world?
(comment deleted)
>But don’t they have the #1 best selling car in the world?

No.

Fifth in the US: https://www.caranddriver.com/news/g43553191/bestselling-cars...

And barely top ten globally: https://medium.com/@wiack/what-is-the-most-sold-car-in-the-w...

Did you just link to a north america only and a two year old article to try and refute the statement that Model Y was the most popular vehicle model in the world last year?

"Topping the charts in Europe and China enabled the Model Y to lead the field in total world sales, despite it sitting in only fifth place in the United States, behind the RAV4, and the heavyweight Dodge, Chevy and Ford pickups that America loves so much."

https://www.hagerty.co.uk/articles/news-articles/the-worlds-...

https://insideevs.com/news/706169/tesla-model-y-best-selling...

They do, but this isn't it.

It's telling that they don't even put the Tesla logo on the Cybertruck.

I do not want my car reporting on me

I do not need "over the air" updates

Sell me the car. I will drive it away and make my own arrangements for maintenance.

I want an asset, not a subscription

Old man shouting at clouds?

Definitely. I don't think you're wrong to want that but I think people who wants teslas buy them for the tech and cloud features.
I think your last question would be answered in the affirmative if you are using a modern phone. They track everything already.
I can be willing to put up with it in my phone but not in my car. They're both qualitatively and quantitatively quite different.
That’s a perfectly fair answer. My problem is that the phone telemetry is so extensive I think it just doesn’t matter at this point. Since apparently no major company discloses its telemetry as much as it should I simply don’t know how wrong I am
Phone is far more intrusive anything’s else is negligible by comparison.
(comment deleted)
Tesla has done a fantastic job of updating the software, adding useful features. They still do (some basic) updates of the cars of 10 years ago. They managed to keep the UI fairly fast and responsive on old hardware, even as they have update the hardware for the console cpu etc many times. This is in contract to VW and many other legacy auto companies, that have struggled with software.

I have worried about an upgrade killing my system or bricking the car but more than 10 years late it's good.

Didn't they also redesign the UI, hiding useful features?
I'd say they made some good and some bad changes, but overall added useful features.
> an upgrade ... bricking the car

That this is even plausible should be cause for concern.

Modern software for cars is like that for phones. You are not in control of it, you can't roll back, you can't usually pick a particular software version to install. In a decently designed modern car, you have a drivetrain computer system separate from the infotainment (the big console with a map in a tesla, for instance). This is frustrating for me as a software engineer. You have to trust the software maintenance and we all know software can go bad.

In cars the general rule is legacy auto companies are terrible, although there are some companies doing okay.

Tesla has done a great job, with apparent good boundaries between the UI/infotainment and drivetrain. You can reboot the screens while driving even (which is worrying but they seem to be well divided). They pay people for security attacks on the cars. Tesla has faster update paths for the main UI, and seem to update the drivetrain less frequently, but they do add and remove some features over time. They have "recalls" that are just software updates. A few other companies have done a good job designing and executing on frequent software updates, Rivian is one that has done a good job. Notably though, Rivian did push a bad update this year that messed up people's cars, had to install a new update a few days later. Before that I would say rivian was equal to tesla in quality. Other than they they are but that is a big difference.

VW is kind of in the middle. Wants to have good software and updates, but it has been a very painful sequence. They had updates that took too long, they were afraid they'd run out the 12v battery during big updates, possibly bricking a car, so you had to bring them to the dealer. Lately I think I read they had a successful over the air update. Oh yeah, tesla and rivian do all their updates over the air. Park you car for an hour, pick "go". Don't do this on a trip ;0)

Then all other legacy cars are terrible. Maybe the worst ever is the GM Lyriq which had a stop sale order it was so bad. Basically unusually, reboot the ui frequently and it keeps getting stuck on people. Hardly believable that they had those problems, apparently no related to a software update bug.

There's android auto and apple carplay. A lot of people like them, I think they are very decent. Legacy auto has almost given up on writing their own UIs, they are mostly bad, but some companies are coming back and trying to stop using them so they can control the experince, and maybe charge people for new features eventually. Tesla and Rivian never charged for their main software. Tesla continues to sell FSD for outrageous prices and it is not done, drives like a scared and maybe drunk beginning teenage driver.

> Tesla has done a fantastic job of updating the software, adding useful features. They still do (some basic) updates of the cars of 10 years ago. They managed to keep the UI fairly fast and responsive on old hardware, even as they have update the hardware for the console cpu etc many times. This is in contract to VW and many other legacy auto companies, that have struggled with software.

I don't know about VW, but my Honda has decent (not great, but decent) software for the head unit (Android, basically). Because my car has no modem whatsoever, it's never been updated, which I vastly prefer to having to deal with "upgrades" of any kind.

IMHO, car software should be developed like a pre-internet videogame: finish it, test the hell out of it before release, then bugfixes or minor enhancements only, distributed through a totally consumer-controlled channel.

A car is a tool, and like a hammer, it's UX shouldn't change after you buy it.

> test the hell out of it before release

Paradoxically, being able to fix mistakes later has had an adverse effect on software quality.

> Old man shouting at clouds?

Reasonable person reacting to insanity?

I believe there's a small but growing market for "dumb" tech. Unfortunately there aren't really many options for cars that I'm aware of. I suspect a quality-but-stupid display company is incubating somewhere, just waiting to nail manufacturing.

> I believe there's a small but growing market for "dumb" tech. Unfortunately there aren't really many options for cars that I'm aware of. I suspect a quality-but-stupid display company is incubating somewhere, just waiting to nail manufacturing.

I'm pessimistic. The product would be too niche.

The problem is twofold:

1. Experts who have the users best interests at heart have bee disempowerd by business-culture and technological change (opening up new avenues of customer exploitation that were previously impossible).

2. A lot of these issues are technical and hard to understand by the average consumer. Furthermore consumers have been trained over the last few decades to put up with a lot of crap. Even if a individual consumer doesn't like something, it will be a vague feeling or a gotcha that happens long after purchase. And what choice are they given by the market? Usually none (e.g. all the copycatting).

Just as an example of how terrible the market can be: I understand the quality of many household appliances has fallen through the floor over the last few decades: they break more quickly, more expensively, and are harder to repair. To enable that, many of the major players own so many brands that even a totally unsatisfied consumer is likely to be a repeat customer, even of the exact same model they hate (e.g. have a terrible experience with a Whirlpool washer, then swear off the brand and end up buying a Maytag, which is literally a restyling of the same exact model).

There might be a market for ripping that crap out, though.

The car has to continue working without ambient data, because it won't always have ambient data. So just metaphorically (or literally) rip out all its antennae.

Even if, IF a company started selling a basic dumb reliable car AND it became popular and people started buying it THEN it would only be but a brief moment before the MBAs came it to ruin it and enshittify that car too. I see no stable market place for good basic reliable (and easily repairable) automobiles. Any attempt to do so either: won't happen to begin with, be unsuccessful because it won't induce perceived status like a giant F150, or will eventually turn into the thing it was originally designed to subvert.
Just wait until self driving is mastered and you not only have to subscribe to the car but additionally your attention is freed up for other things like ads being projected from every interior surface. :) What a wonderful future!
or you know just get any other car lol
Might be worth it discussing the OTA and subscriptions models taking over the industry, but I don't have compassion for people who bought this. You deserve everything coming your way. Have fun with it I guess..
Tesla, America’s favorite vaporware manufacturer since 2006.
This isnt an issue. The target demographic for a cybertruck will never need locking diff's and will cry glory from the heavens once theyre permitted to own them.

The vehicle is a pavement princess designed to help its owner cosplay a working mans aesthetic. The cybertruck is to transportation as the cheesegrater Mac is to the programmer.

In what way would the Cybertruck need to change in order to become a proper offroad vehicle?
Be 2000+lbs lighter? That puts it in the same class as a 4runner, which is already heavy for an offroader. The crosstrek that climbed the hill the cybertruck couldn't is about 1/2 the weight.
That's going to be difficult with current battery chemistry. Maybe a Cybertruck/zeppelin hybrid?
(comment deleted)