Hacker News Reader

Comment by fdye | Hacker News Reader

So I've always wondered if Apple actually building a 'car' was ever really the goal. It seems like a skunkworks where they try crazy things and then incorporate into Carplay. Future seems to be ever expanding Carplay support. I use it and probably would not buy another car that did not have it integrated. I imagine car manufacturers are somewhat tired of always building HUDs and in vehicle control systems. So gradually standardizing on a way to take all the displays/touchscreens in a vehicle and let them be run by Carplay seems like the future. Eventually, they can start handing more and more of the software side (not their specialty) off to people cellphones via some interface, particularly around media. As the ubiquiti of the M1/M2 type chip is found in all of our pockets (embedded ML silicon), the car companies will no longer have to actually embed it in a vehicle as an add on. Plugging in your phone with an M1/M2 type chip will unlock Siri or similar AI functionality in the vehicle. Sure it will drive without it, just like it does now, but it wont be 'cool' and have all the assist, nav, and media functionality everyone wants.

View on HN

I think this is a fair take, and I've had similar thoughts too. However, one thing computer industry people don't seem to get is how much the idea of Apple owning the "dials" on the instrument cluster as shown in the keynote this year will likely go down like a cup of cold sick at most European auto makers. In many cases the designs or colors used in the dial faces have decades of brand history behind them.

Regardless of what customers desire, auto-makers are in no rush to become just a dumb pipe for apple or google's driving software. So far not one automaker has of yet publicly announced support or plans to support the more extensive CarPlay Apple demoed. I think if it does ship, it will need to support far, far more customization on interface and dial-faces than was shown in the keynote for auto-makers not to feel totally sold out of the cabin.

It was slightly disappointing to me I've only really seen Nilay Patel at the Verge in the media point this out; anyone with experience of the car industry I think will come to a similar conclusion. People may forget, it took a very long time for some automakers to even trust adding the existing CarPlay, partly due to concerns regarding loss of control of cabin features. Old CarPlay is is a far less intrusive system than the one Apple are proposing now.

As a thought experiment - if I personally ran BMW or Mercedes, I would have real concerns about dilution of the brand by adopting a generic industry wide car UI, even if I also think Apple would probably do a great job. Maybe the carrot of Apple solving autonomy as part of the package might be enough to swing the deal, but even there, I think in time legacy automakers will work out who to aqui-hire to build their own systems instead of selling out completely.

> https://www.theverge.com/2022/6/7/23157963/apple-carplay-nex...

I'm willing to bet that Apple will lay down some cash to get car manufacturers on board if they think it will increase lock-in for iOS (and therefore the iPhone and other devices) by a certain margin.

Google I'm sure has its own incentives to do this.

Too early to tell if this will be the case or not.

Car makers are very well aware of the lock-in possibilities and execute them already. They were just very slow to wake up to the fast raise of, how to call it, "software defined car", and start pouring endless money now into this.

The car makers usually defined the functions in their car and how to split that to the different computers (plus quite a few of this is fixed to specific computers, anyway). They also defined the data going back and forth between these, at least in part. Everything else was defined and done by the suppliers.

This entire approach was upended by increasingly powerful embedded hardware and possibilities resulting from it, which rendered the entire software stacks in the cars so complex, the car makers had to seize control over this again. Someone has to define the larger scale software architecture in a vehicle etc.

Combine that with that juicy lock-in scenario (I mean the big idea is to have cars connected all the time, like your phone), and you can bet on the fact that the car makers are very well aware of the possibilities which come with that (think of all the data which comes with that). And they are not exactly willing to just yield to the likes of Google or Apple because of it, i.e. we are talking really big money here.

How all this will play out in the longer run, lets see. Car makers sit on a lot of cash, so this will be a fight which will be going for a while.

> auto-makers are in no rush to become just a dumb pipe for apple or google's driving software

The cellphone carriers and manufacturers wanted control too. They both used to dictate what you did with your phone. Apple wouldn’t be the Apple we know today if it hadn’t up-ended that market.

I can’t really see Apple producing something so good that it makes all other car manufacturers efforts look like crap, but then most people would have said the same about the phone industry.

CarPlay is already there as far as I'm concerned. Vastly better than the built-in UI of my car (VW), which is filled with ugly graphics, annoying, slow animations, and poor organization.

I agree in most cases, with the main exception of Tesla. While not perfect, the responsiveness/design/ease of use in their cars is generally consistent and well executed. I'd still take Tesla's OS over CarPlay, personally, given the choice.

I think my BMW i3's user interface is superior to CarPlay. Maps have a lower "ink to information ratio", it is snappy, and well-organized.

I dread the day when I end up getting a touch screen car to replace it.

My previous car had a touch screen, and I've used touch screen cars from a dozen vendors. They all suck badly, except Tesla, but they still suck compared to a jog wheel + physical buttons, or even an old fashioned "dumb" car.

One thing that is different regarding cars is that car industry executives have the benefit of hindsight here. Apple was not a trillion dollar business in 2007, in 2022 everyone pays attention.

Apple's business model is clear; they would hoover up any profits that might arise via new in car software business models. Apple's success in reducing cellular carriers to dumb pipes will be the cautionary tale every car CEO in the world is familiar with.

Remember that when the iPhone came out, wireless carriers thought Motorola and Nokia and friends had too much power and they were looking at whether they could get white label phone manufacturers to make phones for them directly.

1990’s and 2000’s mobile in the US was a refugee camp for executives from the Bell Telephone breakup. They were a bunch of dirt bags trying to snap up everything they could, including re-patenting information technology by adding, “but without wires” to existing patents.

The RAZR phone was one of the first times they’d blinked in a while, and Apple showed up in the afterglow.

Carriers did that to themselves.

I understand where the automakers are coming from and agree that it'll take a while for this to become mainstream because of this.

But I can help but laugh a little at this attitude from carmakers. Modern cars have almost all become the poster child for "bad ui" with their addition of touchscreens that almost universally look dated and run poorly.

> Modern cars have almost all become the poster child for "bad ui" with their addition of touchscreens that almost universally look dated and run poorly.

The dashboard dials have little left to be optimized. This is separate from the issue of media controls in the centre console.

Given occasional hiccups with CarPlay itself, intermittent dropouts or UI lag because of a flaky cable or an older phone running the last supported OS is intolerable to something as mission critical as the speedo.

The speedometer is one thing, but even more "mission critical" are the warning/error symbols, which have a low-to-medium level safety rating on them. Speedometer probably would have that, too, but monitoring "fancy" screen contents like this becomes a somewhat more involved task.

I can't help but compare to what happened with the wireless carriers with their crippled "smart" phones before steve jobs convinced AT&T to take the iPhone 1.

and that relegated ATT/TM/Sprint into pure utilities, with little intrinsic value

Maybe they can start working with a car manufacturer that doesn't have such an image to protect, like one of the Chinese ones that are expanding in the West.

If it's actually good, customers will then push manufacturers into it.

Agreed. The car manufacturers will lose control of the "car user interface" if they take on Apple's new CarPlay. My prediction tho is that some manufacturer will be desperate and take it and on then its a slow moving wave that crests over the industry as consumers ask for it.

Strongly disagree with this; we have many indications that Apple's efforts aren't just on CarPlay.

First, Apple's car project is clearly mostly focused on self-driving (tying into Apple's mapping efforts). To steelman your argument, it's possible that Apple's just trying to develop self-driving software that car companies can just adopt into their cars, but that would be absolutely unprecedented for Apple; they just don't do licensing deals like this. The fact that Apple's developing self-driving AI means that either they plan on rolling out a consumer car, or rolling out a robotaxi service (as a subscription service, which Apple is interested in). It's either licensing, or using it for their own cars; there's no scenario where self-driving is built-into CarPlay, since if you forgot your phone, your car wouldn't work; and it wouldn't work for Android users either. Self-driving will never be a CarPlay/Android Auto feature that "plugs into" a "dumb car". Too finicky.

Second, Apple's clearly investing heavily in car hardware and all related technologies (battery, etc). The most informed people on Apple are Mark Gurman and Ming-Chi Kuo, and they both agree that Apple's developing its own car hardware. See, for example, a 20-year Lamborghini vehicle R&D veteran hired to help design the car just two months ago.[0] That goes far, far beyond CarPlay. Kevin Lynch (who was in charge of Apple Watch) took over the project. There's tons of indications that this is about hardware, not just software.

Personally, I feel it would be better for Apple to design an electric bike than a car; but hey. Car likely comes first, then bike (just like: VR goggles first, then AR/XR glasses). Also far more ways to integrate content consumption (TV, Music) and productivity features in a car.

[0]: https://www.macrumors.com/2022/07/27/apple-car-lamborghini-e...

I would be inclined to agree that it’d unlikely apple provides self-driving software to other manufacturers, but wouldn’t iTunes on windows be somewhat similar? Create must-have software to drive adoption of some other hardware? (Phone?)

> Personally, I feel it would be better for Apple to design an electric bike than a car;

The iphone is already the self driving bicycle of the mind.

https://www.youtube.com/watch?v=KmuP8gsgWb8

> imagine car manufacturers are somewhat tired of always building HUDs and in vehicle control systems. So gradually standardizing on a way to take all the displays/touchscreens in a vehicle and let them be run by Carplay seems like the future.

What is then remaining as distinctive feature of a car vendor? The design?

With their custom entertainment system they can confront they driver with their own brand identity and differentiate how they integrate the different features.

But with EVs the engines aren't as different as fossil fuel engines, they don't have their gear shifting with that adjustment anymore. Lots of brand vlauebis lost and becomes obvious to the buyer that cars are 95% the same across brands.

Not really my wife’s BMW and my Jeep are both SUVs. They are pretty different and Jeep has started offering off road Hybrids. Eventually both companies will offer full EVs in existing lineups. The BMW is much better to drive but I wouldn’t take it off road. EVs will continue to fill the niche gaps their consumers want.

Who buys cars for their HUD? I buy cars for their performance, miles per gallon (or electric equivalent), brand value, etc.

It was a significant factor for me when I last bought a car.

A good interface helps performance in a way. If I don't have the right information and have to work at it I get slower reflexes.

Me. Just about any car is fine as far as performance goes. All I really care about are the creature comforts in the cabin and the infotainment system (or whatever they call it now) is a big part of that.

Well, I'd say there's a difference between HUD and infotainment. I might be influenced in a car decision by the actual HUD and its useful features (for example: the blind spot cameras on some new cars), but the infotainment screen between the driver and passenger is almost always just an annoying barrier between me and having the music/podcasts/whatever + maps from my phone playing in the car.

It was definitely a factor for us, especially as more cars move to all Software interfaces. A laggy or unintuitive system ranges from annoying everyday to dangerous.

Another anecdote, but this is exactly how I chose my last car. I went to CarMax, filtered by CarPlay integration, and then chose from what was available. In fact, the lack of CarPlay integration is one of the main reasons I didn't consider a Tesla at the time.

From basic things like a snappy interface and keeping my podcasts in sync, to suggesting the destination based on the recent maps search I did from my phone on the way to the car, the attention to detail makes it a better experience. I don't particularly get any joy out of driving. I just want to get to my destination safely and quickly, and CarPlay makes the process less taxing.

Today I learned that people pick a car based on how well it integrates with a cell phone.

I can see it especially if the various models are "similar" - I chose my last used car from Hertz from a field of two - based on how well the iPhone played music with the radio. Other than that the cars were functionally identical for me.

The last few times I’ve rented a car it’s had CarPlay and based on those experiences, the next time I buy a car, it must have CarPlay. The experience without it is just so poor (and don’t get me started on how ridiculously bad the experience of pairing bluetooth is in my parent’s car which requires a mix of voice commands and button presses in a completely undiscoverable way).

Once you get used to Waze on the big screen it’s really hard to go back.

yup! me too. Thats how i ended up buying a new car instead of a recent second-hand version of the same. Carplay was an absolute non-negotiable must have.

You can just get an older car and replace the radio.

With a large screen capable of CarPlay with enough real estate for navigation etc? Most everything is custom OEM sizing/fit/etc these days , gone are the days of interchangeable single-cd headunits

Custom OEM size/fit has been addressed to some degree for a quite a while now by both the "mainstream" car audio manufacturers [1] and, more extensively, by some lesser-known aftermarket brands [2].

YMMV based on just how old your car is and the type of OEM head unit, but there are kits (e.g. this one for a 2016+ Honda Civic [3]) that allow you to install "standard" head units with a fit/style similar to OEM.

[1]: https://www.crutchfield.com/g_462350/Apple-CarPlay-Compatibl...

[2]: https://www.joyingauto.com/head-unit.html

[3]: https://www.crutchfield.com/p_120957812B/Metra-95-7812B-Dash...

Yes, I was considering replacing my 2013 Honda Fit but just got a double-height Pioneer radio (AVH-2400NEX iirc) installed by an audio dealer.

It works perfectly, kept the existing car controls and I still have a car that mysteriously fits your entire apartment inside it while also not being subjected to a CVT.

Eh I've had decent experiences with BT in cars, there are some truly terrible implementations out there though.

My current vehicle doesn't have a functional BT connection, because some idiot put it behind a voice activation (and the mic broke).

I don't think this is the long term future, though - I've seen it work really well, if consumers actually start caring about how to connect to their cars it will work better.

BT will always "just work" - Android Auto and Carplay (never used) are enterprise monstrosities you must sign your soul for in blood that don't work on Thursdays, by comparison.

Yeah yeah I'm sure it does some app thing but I honestly don't car, what with cellphones becoming less reliable over time. I only want to play music in my car, BT works great (once it is setup).

My experience with CarPlay has been rock solid in four different cars from four different manufacturers. I was able to connect in about a minute in every case (the longest time for connection was the first rental car with CarPlay where I didn’t realize it was there until I plugged my phone into the USB jack to charge—this was in 2018 and it was wired CarPlay).

In contrast, the BlueTooth in the car that I own typically takes about half a block to a block’s driving before it actually connects to my phone, occasionally refuses to connect or will pair but not send audio. There’s some weird interaction between Overcast and the car’s Bluetooth that means that if I’m listening to a podcast when I turn off the car, I need to restart Overcast before it will send audio to any output (including the car when I get back in).

I don’t understand the “you must sign your soul for in blood” claim, unless you’re an open source absolutist (or maybe Android Auto is really bad, I don’t know), but your claims don’t match my experience.

CarPlay for rent car is really important because it don't need to learn car original system, and needs navigation than daily use.

People use the infotainment in cars almost as much as the actual driving interface. It's not all that weird!

Cars are a very mature technology, if you're not a car hobbyist, and the different brands are not very different.

The new-ish field of screen and cell phone probably has the most differentiation right now.

In the US, lack of CarPlay really only removes Tesla from consideration at this point. Everything else, from Ferrari to Kia, supports CarPlay. Is there another holdout I’m missing?

Weirdly I think rivian too. They think they can be better at car media software than apple - they’re wrong, and I hope they change their mind.

Millions upon millions of used vehicles don’t support CarPlay.

70+% of US car buyers per the Apple presentation where this was discussed.

It’s absolutely order qualifying for me.

Yeah, Tesla not having CarPlay the main reason I decided not to get one. Having the car be the continuation of my phone is just so convenient.

Same. My car has Carplay, while my wife's doesn't. Guess which one we discuss selling? A large part of that is the better integration of media. Its hard to go back to a propped up cellphone in a holder with bluetooth, once you've used Carplay.

It's in my top 3 criteria along with reliability and energy efficiency. Performance is probably dead last for me, assuming some reasonable floor (i.e. I can safely merge onto the highway).

It's something I'm going to interact with on a near daily basis, so it's important that it doesn't piss me off.

Not sure, but a car with good physical controls might jump to the top of the list.

The only reason I’m considering upgrading my 2011 Audi to a newer one is the technology. The car itself is great but it’s got crappy stop start features, crappy audio/infotainment tech, bad A/C etc.

It is relatively cheap to upgrade the infotainment systems on most cars. Apple and android compatible systems start as low as $350.

I did a self install on a 2010 ford Escape after my kids killed the radio (by inserting pennies into the cd slot…). Between that upgrade and new tires it was like driving a new car.

Cabin usability and ergonomics are a huge selling point. CarPlay & HUDs are just one piece of that, but it a critical piece of the overall usability experience.

Getting into a car that has a pre-CarPlay infotainment system feels like stepping back decades in time even though the car may only be a few years old. Slow, unresponsive, buggy, and just extremely unpleasant and unnatural to use.

IMO, cars that predate screen-based infotainment systems have aged far better than pre-CarPlay systems. OEMs are just disastrously bad and behind the times at building a quality infotainment system. Mirroring the smartphone that everyone already has is a fantastic solution.

It's a very good thing that CarPlay shifts so much of the responsibility out of the car and into the smartphone, something that is much more easily and frequently upgraded.

The exception to this is ~2009 era BMWs with professional Nav, which feels just as snappy and good as my 2022 replacement for it. My large survey of a decade of rental cars suggests that almost nothing else has a screen that isn’t blurry.

I've only test driven a few cars equipped with HUDs, but it's a super cool feature. I can see the value of a car that projects nav directions in addition to the more standard things like current speed. There's even phone apps you can use to simulate a HUD, if you keep your windshield clean.

I'm getting a lot of responses and it seems I can't edit my previous comment. To clarify, I mean who buys cars primarily for their built-in infotainment system that's not Android Auto or Apple CarPlay? If a car has either one, then I don't really care what it comes built-in with because I won't be using it. Beyond that, other factors are necessary too such as MPG as I mentioned.

Speaking as somebody who doesn't care at all about cars, the primary things that would drive my car purchase would be:

- Cost (per unit distance in fuel, maintenance, etc)

- Comfort (which includes integration with other software, ie Carplay)

I don't care about "performance"; I don't even know what that means, and frankly I don't care. The car needs to get from my house to some other point on the map and back cheaply and comfortably.

By performance I mean the comfort of driving on the road. I used to have an anemic car which sputtered every time it drove. The engine was fine, the mechanic said, just weak. So now I look for a comfortable drive. It doesn't have to be a racecar, but it should be powerful enough to have a smooth ride.

I didn’t buy my Audi for the HUD (or other UI) but I’m unlikely to buy another one after:

1. The Android based entertainment system regularly crashes. 2. The non-android based digital center cluster is designed so poorly as to have the date in two to three places and fuel levels hidden by default. 3. Safety systems that will needlessly engage the brakes because it believes that driving past the backed up left turn lane on a gently curving road is an imminent head on collision.

I’ve learned to work around these issues for the most part. Disabling all the wireless hardware reduces, but does not eliminate entertainment crashes. I memorized the steering wheel dial flicks to get the fuel gauge to display. I turn off the safety features.

Most of the car is very nice. The software stack is honestly something Audi should have outsourced to a competent development team elsewhere in the automotive industry.

If you mean the infotainment system or head unit, I think lots of people place a very high value on usability. Depending on much you rely on navigation or use apps for audio, high quality UI goes a long way.

Fortunately if you also want to prioritize performance, mileage, etc., CarPlay makes it a lot easier to tolerate a lackluster factory interface.

After buying a RAM, I carefully consider software quality.

Sometimes the GPS azimuth data gets out of sync, which is annoying. However, the dealbreaker is that the transmission has some software bug where it overheats if you drive for more than an hour or so in certain circumstances. (Turning it off and waiting five minutes reboots it enough to get it to do whatever it needs to do to cool itself off.)

Sometimes we get a dreaded error message like this on the dashboard: "Drive to destination without switching into reverse or turning off. The vehicle might not turn back on.", which means the drivetrain computer cannot talk to the transmission, and it is stuck in a random gear.

Also, there was a safety recall software update. Applying it broke the state machine transitions for the push button ignition switch. (If you see a Ram owner accidentally re-start their engine after parking, this is why.)

Other Fiats (or whatever they renamed themselves to) are apparently on the same drivetrain software platform and have similar problems. All of the above things happened in warranty and none are fixable, from what we can gather.

It also had a bug where sometimes the cruise control wouldn't disengage unless you stood on the brake pedal while forcing the shift knob past an interlock and into neutral. They fixed that with the software update that broke the ignition switch.

We're eagerly awaiting the day when we can upgrade to an EV pickup truck from another company.

The GMC we had before the Ram was even worse. It was even more likely to strand itself, and the ABS / traction control liked to unexpectedly engage. Also, we had many mechanical issues. Like RAM, we had difficulty getting GMC to honor the warranty (with < 100 miles on the odometer).

Anyway, if the radio is buggy, don't buy the car.

You say that as if it were a bad thing. I'd rather a good entertainment system than the shite that Toyota puts the bare minimum effort into.

Toyota are so glaringly bad these days. Their horrible BT implementation on our age-old Sienna really ticks me off on a daily basis.

> these days

> age-old Sienna

Interesting dichotomy there. I have a modern Toyota and BlueTooth works just fine.

My wife has a previous gen RAV4. I have a current gen Corolla Hybrid. The interface is VERY similar. However, she has a physical play/pause button on her RAV4 while my Corolla does not. This is INCREDIBLY frustrating because when you start the car, it MIGHT automatically resume your media. Sometimes this is OK, but other times, it is interrupting a conversation and losing your spot in the podcast you're listening to. A pause button is the perfect solution. However, in the Corolla, you have to wait about 30 seconds before you can tap the touch screen to switch to the audio touch controls, then wait a second or so before you can tap the touch screen pause button.

TLDR: Toyota removed the physical play/pause button and it's really annoying.

I forgot to mention, it's even worse! There is still a perfectly usable, unused button available to be a play/pause button! There are two knobs, each press-able. The one on the right does nothing when you press it. It SHOULD be the physical play/pause button and I have no idea why it's not.

Same energy as complaining about not being able to provide customer with "dazzling box cover art design" because digital downloads made shelf boxes irrelevant.

Cars have handling, space, looks... just about everything else other than the dash/touchscreen is meaningful competition space.

Yeah, we sold cars for the better part of a century without digital entertainment systems, so we can go back to that not being a differential.

Particularly when it seems like none of these bespoke entertainment systems raise the bar, and instead the standard seems to be “not annoying and in the way.”

I was thinking along those lines, but the majority of phones are Android and I can't see a car manufacturer tying major features of a vehicle to users of Apple phones.

Do they get Google to implement equivalent Android features just as CarPlay and Android Auto are complementary now? In some ways that makes sense (you get the personal assistant you prefer instead of one provided by the car company) but it means differentiation must be limited because the car can't change too much depending on what phone is plugged into it.

I have a 1 year old Honda Pilot - it works with Apple CarPlay as well as the equivalent from Android.

It seems easy enough for car manufacturers to support both.

HUDs are the small part of the story. far more important thing is what sensors you have access to and what can you control? Car manufactures are notoriously behind and bad on properly expanding and standardizing both sensors and controls. So, it is impossible to deliver consistent experience and capabilities. Even in 2022, most manufacturer haven't still figured out how to do firmware update without using USB drives. You can't even use phone to unlock cars. These are not hard things, but car manufacturers just can't get around to do it given their ancient and inefficient supply chains and factories which are only optimized for price wars, not leap-frogging experience. CarPlay is great for music and may be some cute graphics of basic gauges but the real value ultimately lies in some level of self-driving, assisted features and holistic integration which cannot be enabled without having complete control of hardware.

> Car manufactures are notoriously behind and bad on properly expanding and standardizing both sensors and controls.

Does Tesla make their own sensors and are they "ahead of the curve"?

> Even in 2022, most manufacturer haven't still figured out how to do firmware update without using USB drives.

This... makes sense to me? Physical access and automotive service go hand-in-hand. Over-the-air presents update hell and security issues.

Why does it not present hell in any other use case, such as computers, phones and games consoles, or actual cars on sale today? Why aren't Tesla or Rivian owners collectively going crazy - there are over 1 million Teslas receiving updates approximately every two weeks? These arguments do not appear to be borne by any evidence in reality.

It's not like physical car repair places have some amazing flawless track record either... Physical access and the servicing of electrical devices do not need to always go hand in hand in 2022.

It was several years before I was completely comfortable upgrading my iPhone without backing it up first, and Apple had decades of system software experience and complete control over the hardware.

Auto manufacturers are much, much worse than Apple at software, and I suspect they have less confidence all of their embedded electronics work correctly under all circumstances, although that’s just speculation on my part. From what I’ve heard about canbus, it’s a bit of a mess.

Anyway, I can completely understand why traditional auto makers are less comfortable allowing a car to be updated anywhere at any time, and why they might be less competent at securing the update mechanism.

> Why does it not present hell in any other use case, such as computers, phones

It absolutely causes problems on computers and phones. Windows updates are famous for interrupting/losing work at this point. MacOS upgrades are less aggressive but far more likely to break formerly functional software (and stomp custom configs). Phone security is always an issue if not an outright disaster in dozens of ways. Look up the Qualcomm / MediaTek audio decoder issues from just this year.

Somebody can spy on you and steal sensitive data w/ access to your computer or phone. What can they do with a car?

Have you read the Wired article from 2015 about hackers remotely killing a jeep on the highway?

Some levels of computer security rely on avoiding connecting to public networks. Or any network.

> Why aren't Tesla or Rivian owners collectively going crazy

Just last year there were over the air updates on Teslas where owners reported vehicles stopping without warning. They fixed that promptly with another emergency OTA update, but not exactly a shining example.

Just last month someone I know told me that the performance profile of his Tesla was weird since a recent update.

These are the realities. There's a ton of evidence.

I'm not arguing things don't fail. Entirely non-connected cars fail for all kinds of hardware/software related reasons too.

One simply can't ignore the fact that today, including outside my own home, millions of cars update software just fine every few weeks, without issue. The number of phones or computers managing it is in the billions.

I dunno, a billion dollars a year (for 8 years) for CarPlay skunkworks seems like a bit much even for Apple..

When you have that much cash on hand, is it really?

I mean how much of a return can carplay generate on an initial investment of 8 billion and on-going OPEX even if selling to all the manufacturers (which it isn't).

I think the point is that it was unlikely their intention to be building solely for carplay and an actual car was a bonus. However if they wind down the hardware play of an actual car this could be something that soften those sunk costs.

Imagine a world where the car manufacturers no longer deal with software development for the vehicle, but instead you have a licensing agreement with Apple where CarPlay is the interface. Similar to Volvo and Google and Android Automotive. A lot of car manufacturers are standardizing around Android in general (see Acura, Volvo, BMW). If Apple could get in with an A-series chip and a custom software stack, then that would probably be a decent amount of money through some hardware purchasing and software licensing. Look at how Cariad (the VAG stack) blew up schedules for upcoming Porsche and Audi models and early versions of it were laggy, missing features, OTA updates are slow/non-existant, etc. If Apple can get their foot in the door, it opens up new service opportunities for things like Apple Music alongside any licensing, which may be the bigger goal.

Given the thread on Volvo and polestar bemoaning all the issues with the new Android Auto versions, I’m not so eager to continue in that trend. Let the car companies do what they do best. Build an automotive grade car that does the car thing really well and leave the entertainment to the CarPlay. Getting hud and center driver console display driven by CarPlay would be awesome though.

Looks like that might be coming in 2023 if the CarPlay site is any hint: https://www.apple.com/ios/carplay/

>Next generation of CarPlay

This next generation of CarPlay is the ultimate iPhone experience for the car. It provides content for all the driver’s screens including the instrument cluster, ensuring a cohesive design experience that is the very best of your car and your iPhone. Vehicle functions like radio and temperature controls are handled right from CarPlay. And personalization options ranging from widgets to selecting curated gauge cluster designs make it unique to the driver.

Apparently CarPlay itself has had the ability to drive multiple displays for ages. It is the car vendors that haven’t utilised it.

and how they would differentiate between themselves?

what would be left? colour of the seats?

no new features until Apple decided it's a time

If it convinced people to buy high margin iPhones a lot. Not to mention with the push into services, what’s the lifetime value of an iPhone customer.

The five pillars of revenue are:

1. Selling a thing.

2. Selling add-ons and upgrades to a thing.

3. Selling other things to people who bought one of your things.

4. Preventing people who are subscribing to your thing from switching to somebody else's thing.

5. Preventing your competition from commoditizing things and driving prices down.

(There aren't really just five, there is no one "theory of everything," it's just a literary device.)

CarPlay is obviously #1 "Selling a thing," manufacturers license it. But it could also be #4 "Preventing people who are subscribing to your thing from switching to somebody else's thing."

If automobiles all have their own proprietary interfaces or worse, Android Auto, people who drive cars may end up buying phones that integrate nicely with their cars and ditching their iPhones.

If my conjecture is correct, Apple is investing in CarPlay to protect the most profitable product the world has ever seen.

>But [CarPlay] could also be #4 "Preventing people who are subscribing to your thing from switching to somebody else's thing."

The more reasons an existing user has to value something they already use, the harder it becomes for the user to switch. With few points of stickiness, it's easy for a user to check if an alternative has, say, similarities to features A and B. But if it's ABCDEFGHIJK, the likelihood of that user leaving drops dramatically. It's a common point that Apple does well iterating on its products, but that iteration goes beyond performance and capabilities into areas of stickiness.

Apple refuses to fund development of their own apps and platforms, yet sinks a billion into Carplay?

Something, something, "put the armour where the holes aren't."

If you are selling a lot of phones and tablets and laptops with your apps and platforms as they are, you don't want to invest billions in making them better just to make your existing customers stop complaining. They bought your product anyways, clearly whatever they're complaining about wasn't a deal-breaker.

You only want to invest in your own apps and platforms in areas where improvements would drive meaningful business outcomes. So you need to look at features that would cause someone who would otherwise buy a Samsung phone to buy an iPhone. Or you need to find people who didn't but an iPhone or iPad or AppleTV or OS X laptop and figure out if there is something that can be added to the product to get people to switch.

In the case of a dominant player like Apple, those are hard to find in the mainline product. Big new sources of revenue are most likely to come from entirely new product areas (thus their investment in phones, watches, tablets, and set-top boxes) or from going down-market and making commodity products.

They're allergic to cutting margins to the bone, so we're left with trying to find new markets, and that is why they invest so much in an automobile skunkworks while being much more careful about investing in the products that are already successful.

Let's take a look at it. I'm sure Apple is getting paid in some way for CarPlay integration. Now if the next generation has higher level of integration that car manufacturers use it instead of their own built software, there's certainly money flowing to Apple.

So, let's say that Honda starts integrating it into their vehicles. In 2020, 1.3mm vehicles. Of course, initially, not every car Honda sells will have CarPlay installed, but a few years out, let's assume every vehicle does.

If Apple got 2% of the selling price of the vehicle. Across all Honda models, the starting MSRP averages out to about $32,091. So, on average, Apple would get $641 per vehicle sold. Over the course of a year (if every vehicle sold with it), just from Honda, Apple would get $833mm. Add in a few more larger manufacturers and it's not a bad investment.

https://www.macworld.com/article/233855/carplay-faq.html

> Whether it costs money to get CarPlay support in your car is up to its manufacturer. While Apple doesn’t charge automakers a fee for the necessary software to integrate CarPlay, there are some costs associated with meeting the necessary hardware requirements.

The only reason an automaker does not include CarPlay is to try and keep their product from becoming a commodity. It was disappointing to see Toyota be stubborn and refuse CarPlay for so many years.

That, and greed. Looking at how many features car manufacturers keep behind absolutely meaningless arbitrary levels and packages, it's just ... argh

Almost every new car sold in the developed world has a display in the dashboard, but very few actually have a Heads-Up Display (HUD) projected in the driver's line of sight. It's mostly just premium models from GM and BMW that come with this feature. It's a shame that more cars don't have a HUD as it really does help to reduce driver distractions.

I think all the Genesis models have them

Volvo has a HUD projected on the windshield.

This is by far the best feature of our BMW F10 (manufactured in 2014 IIRC). I hate the distraction of constantly checking my speed in town, where there is a mix of dense traffic, many obstacles, outright stupid cyclists ignoring all traffic rules and aggressive motorcycles drivers and tons of radars that fine hard.

I've noticed with my previous car (also BMW, E46 that I will remember fondly for the rest of my life as amazing car but not due to this) that if I looked at speed, my peripheral sight of whats happening on the actual road was almost 0. I may notice break lights of car close to me if its newer car, or an atomic blast, but anything else simply no. Just something very bad waiting to happen.

Its such a great safety feature even with minimal info (current speed, current speed limit, and if car navigation is on some basic directions, but maybe thats the key, don't clutter it with too much info) that I would make it mandatory if I were EU and subsidize it.

I've often thought that we should use that massively huge A pillar for some simple gauges, such as speed

Gauges on the A-pillar would interfere with airbag deployment.

I wonder if you could do e-ink displays with a seam that can tear open

Yeah, even just an A-pillar that changed color if you went over the speed limit.

> It's mostly just premium models from GM and BMW that come with this feature.

Some Mazda SUVs (e.g. CX-3) had a HUD starting back in 2018 or earlier.

Land Rovers also have a nice HUD.

Agree Apple is using it as a play ground for CarPlay and never going to ship an actual car. Maybe at project inception that was a possible delivery. At this point with legacy makers actually making competent EV offerings.. no.

Car market has many challenges that cause production and sales to scale slowly. Even if they outsource manufacturing, like Fisker has down with Magna (which no one has done at the many many millions scale Apple would need to).. what about service centers?

Next cars face various tax, tariff and regulatory challenges far beyond any market Apple participates in today. They couldn't simply make the cars in Shenzhen and ship them across easily. Almost certainly some non-trivial amount of assembly would need to happen domestically which means a US workforce and potential union concerns. This is due both to tariffs and to maximize applicability of the new EV tax credit law.

Further, I don't think Apple likely brings any special sauce to ADAS/Autonomy, and its a hard challenge many of the current players underestimate & overpromise.

Oh and repairability and granularity of components/modules. EVs need to be serviceable in ways that phones/tablets/Macs are not. Car needs to be designed for this, and parts supply networks need to exist. This is not an Apple forte.

All this to enter a traditionally low margin, boom&bust industry? No thanks.

Making the iPhone more sticky seems like a solid play. Use the phone in place of a key. Keep maintenance records on the phone. Tap a few buttons on the app and give the dealer a virtual key to the car so you can get it serviced. States are already putting IDs in to apps and the wallet. etc.. Maybe use the phone for various road passes and such, that seems like an easy and natural extension.

Around the time the self-driving craze sort of took off, the noise in the echo chamber was that Apple was terribly far behind in 2 main categories: services and AI/ML. Now, they're charging ahead with services and they've got custom ML hardware on every single device they sell. A car seems like too big of a project with too much hype to use as a forcing function for all that stuff.

Seems like there are some strong health and safety plays as well. If you have a watch on, it already can do fall detection. Car crash detection is a logical next step, if all the passengers had watches on, they could start sending real-time telemetry to the paramedics, maybe encourage them to prioritize a passenger that was in greater distress. With the cameras and such, there is absolutely enough processing power in your iPhone to look for drivers falling asleep and with some other data they could probably make a pretty good guess if you're intoxicated.

My concern was that Apple would make a car and it would be a McLaren or something, it would be coveted, look amazing, and be just about completely unattainable. Now if I could buy a Toyota or a Hyundai and just plug my phone in and it became the brain of the car? I'd talk myself in to taking a new car for a test drive to try it.

> Making the iPhone more sticky seems like a solid play. Use the phone in place of a key. Keep maintenance records on the phone.

Apple already does this,[1] though not many car manufacturers support it (BMW and a few models from Kia).

Teslas let you use any iPhone or Android phone as the key. You can also use your phone to turn on valet mode (which limits acceleration and max speed) and track and control the car (turn on heat/AC, lock/unlock, view through the cameras, etc). Lastly, you can book service or request roadside assistance from the phone app. It's all very handy, and I'm surprised more car companies haven't copied these features yet.

1. https://support.apple.com/en-us/HT211234

Can't say about building or selling an entire car, but they did have a massive self driving division which went nowhere.

Is their self driving division still active?

More than ever if I go by how often I see their cars on the road. Three yesterday.

What do they look like?

Some photos here, although this is from 2017

https://appleinsider.com/articles/17/04/27/apples-self-drivi...

SUVs with car size iPhones on top. I see them almost every day here in Cupertino.

Like this: https://macreports.com/wp-content/uploads/2021/08/apple-car-...

I think new ones have more doodads around the fenders.

Seeing how boring Carplay has been so far I’d have my doubts about skunkworks. That said it makes sense they might be working on exclusive operating system of sorts. The way car systems (100s of suppliers and their implementation) work nowadays is insane.

I don’t know what you mean by “boring” but CarPlay is fantastic. It works really well and covers most things I’d like to do with a central console.

"Boring" is good compared to most of the systems the car manufacturers came up with. A some of them still manage to screw up the integration of CarPlay with their own infotainment systems which often don't seem to want to get out of the way.

The display of outside temperature is a curious oversight. That's the one thing I find myself regularly toggling back to the native console for.

iOS plasters that information all over the place when running on the device itself, so it is not like it is out of character.

CarPlay does not work well at all on my 2018 Golf. There's a multitude of obvious, awful bugs. (eg. audio directions not occurring, but nonetheless ducking spotify audio for them)

How much of that is the fault of shitty CarPlay hardware in most cars? We see how well things like AirPlay works when Apple controls both ends of the hardware.

Probably about 90%? Apple is writing software in an attempt to replace native software and whenever their software doesn't perform as well as the native software performs it's on them.

I do understand that Apple needs to integrate with hardware from hundreds of different manufacturers and I feel their pain - but nobody is twisting their arm and forcing them into the market.

I'm sure 10% of the time or so the car maker is just being completely incompetent in using incompatible hardware or switching things out at the last minute without warning - but yea, mostly on Apple.

> audio directions not occurring, but nonetheless ducking spotify audio for them

FYI - there are two different audio “sources” at play: “music” and “announcements”. Your “announcements” volume is set to 0.

Not that I have any idea how to change that on your car, but on my VWAG car I can adjust the “announcements” volume by turning the volume dial while it is speaking the directions (or not, in your case)

You are right, maybe because it’s boring it is good, but IMO its not fantastic. Main use case is using google maps in a pretty inconvenient way…

Like, you iOS doesn’t support my language maybe then just let me read message on that massive screen.

And if my kids want to watch peppa pig - let them? Blocking video from central console is such immature safety feature.

> functionality everyone wants

Depressing if manufacturers and users want more chip/software dependencies after the last two years of supply chain fun.

For better or for worse (IMO mostly better), users care more about features than the tech that goes into making those features work

Getting cars onto portable, standardised software makes the chips much more of a commodity and should reduce supply chain issues a lot - you go from needing to source xyz specific microcontroller to just needing to source "a chip that can run iOS/android version x".

If it would be an environment like desktop PCs it would be awesome. If it would look like the Apple landscape it would suck.

I wouldn't buy such a car to be honest. I thought the future would be more accessible cars, not some locked in trash where you have to pay dividends to even more suppliers. Although the car industry is heavily guilty here as well.

For me a car is a tool. Aside from playing music, a technology available since forever, I don't expect much. Navigation is a solved issue for the most part. Self-driving cars are far away. An actual HUD is really nice, but many manufacturers already include them. If there is ever a standard, Apple would not be a good custodian.

Of course it is. Apple wants 30% of everything their car delivers you to. You're Apple's customer and they're bringing their customers to your business so clearly they deserve 30%. Plus even more lock-in, if you want to switch to Android or something else you'll have to buy a new car.

I always assumed it was an employee retention project. Jony Ive seemed bored of re-designing yet another rounded rect with a black mirror and was ready to leave, to attract top AI people you need a self-driving project, etc. But I guess at this point it has gone too far for just that.

It seems Mercedes are not going that direction at all with with the EQ range. They've gone all in, the MMI is deeply embedded into the car, speech recognition runs locally, it can control performance mode, seat massagers, huge swathes of car specific functionality, via speech, that I think Apple would have an uphill to catch up with, universally, across all vendors.

I doubt the upper tier of manufacturers would be happy with mere UI skins either.

Cars that are more like appliances, shared cars, with user customization in your phone, maybe that could work?

I would guess that for like twenty years Apple had sort of plan stored away to launch a television. Do you remember Mac TV? Pippin? They probably considered it as a possible integration with Apple TV. That would mean for decades there was someone planning the product: How long would it take to launch? How much would it cost? What might the design be? What would be the business plan?

They probably have a few plans like that.

Imagine not being able to control basic aspects of your car like heating and stuff because you forgot or choose not to use a smartphone!

Is there a car on the market where this is the case? I'm not sure why that would be anything to worry about. I don't think any manufacturer except potentially Apple would require this for safety and competitive reasons.

As they've taken over more of the car's UI, I've wondered if the main goal is to bring a Comma-like feature to CarPlay. They could sell a lot more Pro-level iPhones if Apple offered self-driving functionality through CarPlay (but only for the top-tier Pro models).

> Future seems to be ever expanding Carplay support. I use it and probably would not buy another car that did not have it integrated.

What specifically about Carplay makes it so important, versus the non-Carplay equivalents that other modern cars have?

It doesn't suck as much. OEM car UIs are terrible.

And CarPlay gets regular updates and support, where usually what you buy from the car manufacturer had better make you ecstatic, because it’s not going to change, and you’ll be lucky to even get bug fixes.

My wife's car isn't that old (2017 model) but using the navigation is a painful experience. Inputting an address takes multiple minutes and can only be done by turning a knob one letter at a time. First, state. Then city. (There is no defaulting to current location, so this must be done every time). Then street name and finally numerical address. If any of these items aren't in the database, you get to go back and start over or try for the hopefully nearest address that is in the system. It's a mess but sometimes necessary.

It also no longer gets traffic updates since the modem is 3G with no upgrade path currently. It does not support AA or Carplay. I hate it but she loves the way it drives, so we'll be keeping it for a while.

Yeah, exactly. Those built-in GPSes are just atrocious. I only use mine for the radio and have a car mount for my phone for all the rest of it.

I also wouldn't buy another car without CarPlay.

But why won't Apple possibly build a device like comma.ai to basically fit into the car's OBD port and also bring in some self driving abilities and eventually full self driving, just by a completely new device that pairs to the iPhone. They can possibly strong arm autos to make the plug and play real smooth. It's not that far fetched. They might actually be able to do something using mixed reality, given how either pure vision based or lidar based self driving approaches are still not reaching the levels they were supposed to years ago.

This could be under the usual AI umbrella like other companies, but the fact that there is an AI DR to Tim Cook makes me think self driving is big on the agenda. I"m not sure which other AI effort is so valuable to Apple to warrant a SVP. https://www.apple.com/in/leadership/john-giannandrea/

Reply on news.ycombinator.com