Kuo: Apple Car Team Must Be Reorganized Soon to Meet 2025 Production Goal

[lazyrighteye]

Level-5 self driving cars are NEVER going to happen.

Please stop the insanity.

There are lots of other ways to make money, don't need to do it with cars.

Maybe start a restaurant or something? You already named yourself after food..

You know what they say about never... ?

I don't know... As more and more 'things' require OSes, it is completely in Apple's interest (where their OS is the center of their ecosystem and their bread-and-butter differentiator) to explore said 'things.'

Why?

For starters: math. Over the years, they have trained a GIGANTIC user base, that is growing exponentially, on a specific way to interact with their technology. Set specific expectations with those same users about what it means to interact with their technology. With such a massive, indoctrinated, user base comes a level of risk mitigation (though obviously not completely devoid of risk) few companies can afford. The sheer volume of this well-trained base lowers Apple's risk when considering/entering new markets. In many cases, when Apple decides to enter a market, they instantly bring a tangible audience to said market. Not every company can do this and it's one of their most significant advantages - the result of years of patient, diligent, unflinching work to establish their base (thanks Steve & team).

Secondly: vehicles will only grow increasingly dependent on OSes as time marches on. Apple knows this. Hell, I'm a doofus and I know this. Thus, why on earth would Apple not be interested in getting in on SUCH a massive OS category? Said another way, why would they leave such a massive OS opportunity open for someone else to swoop in and start shaping users expectations of technology? They have worked too hard, over decades to let an opportunity this massive, drift by. Plus, they have proven pretty good at such things.

Third: Apple is clearly focused on being in the place sand spaces there target most needs - on handhelds (check), on computers for work (depending the type of work, check), on computers for fun (check), on computers for creation (check), home entertainment (check in pogress) and traversing between all of these spaces in cars (currently figuring out how best to check this box).

Fourth: we have to get off this "Apple = desktop computers" mindset. There was a time that was more the case. But I would bet my precious new iPad mini 6 that even back in the day, Steve and company were thinking beyond "desktop computers." The minute we start thinking differently about where Appel could have relevance, the possibilities begin to open up.

Fifth: like most here, I don't work for Apple, I'm not privy to anything different than the rest of us. But I do have curiosity and an imagination. Bridging the gap between one of the leading OS developers on the planet and future devices that will likely require an OS at some point isn't really much of a stretch. Even for a guy like me.

 

[AbsurdlyJames]

I just don't get how this rumor makes sense as a business strategy. I consider myself an Apple fanboy but I wouldn't touch a car made by Apple for the next ten years for the same reason I wouldn't touch a car made by Kia or Hyundai until recently—they need to demonstrate enough reliability just as a car before I'd even consider it. And then to say they would jump out of the gate with full self-driving? No thanks, not touching that even with a very long pole.

The whole industry strategy with regards to self-driving seems completely messed up. There seems to be no progression from driver assistance to full self-driving, as though by magic some company will figure it out and establish enough reliability from the get go to have enough people put trust in it to actually use it. I suppose it works for Tesla because they have fanboys who believe (against all evidence) that the company can do no wrong, but for everyone else, how can anyone expect someone to even try self-driving features until there's a track record of established reliability?

If some company really wanted to get to full self-driving (as opposed to, say, grifting investors or consumers with promises (Hi Tesla!)), they'd have some sort of plan for establishing that reliability. For instance, I work in Yosemite Valley in the summer. They have shuttle buses going through the valley and, notoriously, they find it very difficult to hire enough drivers to make the shuttle service work. These are shuttles that go on a fixed circuit throughout the day, on a route where various aids to autonomous systems could be installed. A company serious about vehicle autonomy would focus on and deliver something like that first—not just because it's easier, but because reliability of the system is something you need to demonstrate before reasonable consumers will buy into the system. And from that point, there's a huge market for autonomous buses in urban areas, and then the trucking industry, until finally you get to the point where total self-driving becomes a reasonable option.

But I don't see any business actually looking in that direction, which is a shame because National Parks do need better shuttle service. That's why I think it's all a con of some sort. It's also why I am very skeptical of any rumor about Apple doing self-driving cars.

 

[Shirasaki]

First of all, apple cannot make a self driving car that is both connected and not controlled by a main computer of sort (for example, hacker can take control of your car and program it to crash now).

Second, how autonomous that car would be? Can it drive without developing special roads and additional infrastructure? Can it travel to unexplored areas without problem? How smart it can be to dodge pedestrians and other cars?

Third, servicing issue. Will apple force customers to service their own cars in their own store? Or allow third party to service non-computer parts? Let alone pricing.

There are so many more questions (including regulatory ones) waiting to be answered. I’m not confident to say apple has figured out most of them. Just scrap the 2025 deadline until everything is figured out, or are they going to repeat AirPower humiliation again? Beta testing a car en mass is equivalent of trashing human life and disrespecting other human drivers.

 

[sunman42]

Last time I checked, Apple was sitting on a pile of money that would cause Scrooge McDuck to weep. If they don't want to push a product until it's ready, they have the resources to keep trying. If they decide they can't make it work, they'll pull the plug when they get to that decision point. I suspect neither of those possible outcomes has anything to do with dates made up by industry analysts.

 

[huge_apple_fangirl]

Kuo is throwing out a lot of weird rumors lately… let’s see what German says first. If the project is disbanded I don’t see how they launch in 2025. Well actually, even if the project isn’t disbanded I don’t see how they launch in 2025.

 

[DevNull0]

This is Apple.... the biggest control freak company on earth.

That's a great joke. This is a company that outsources everything in their products. Samsung puts more into the iPhone and iPad than any other company (between the screen, flash, and DRAM).

Timmy's Apple is not.a control freak company, they throw together a design based on other company's products with outsourced assembly. And Timmy's core design philosophy is make it cheaper, nothing else matters. Apple certainly won't raise their costs just for the sake of more control over their products. Even with Apple Silicon, most of the design work is not done in-house.

 

[yossi]

It still funny to see people who believe this stuff though. Apple car…lol.

Full, driverless, no-steering wheel cars are the "flying cars" of our generation. At most, we will see them limited to closed environments like campuses or specially built cities. in 30 years, there will be ZERO cars completely autonomous on the streets of Manhattan

 

[Yrmmont]

My guess is that they will copy the Tesla 3. Maybe not 100% but enough under the hood. Why reinvent the wheel.

 

[huge_apple_fangirl]

That's a great joke. This is a company that outsources everything in their products. Samsung puts more into the iPhone and iPad than any other company (between the screen, flash, and DRAM).

The screens are customized. DRAM is commodity parts. By flash do you mean the camera flash module? Also a commodity part.

Timmy's Apple is not.a control freak company, they throw together a design based on other company's products with outsourced assembly.

Apple designs a lot in-house, and the proportion of parts designed custom by Apple is going up, not down. As for outsourcing, it's not the same as what, say, Dell does. Apple designs the manufacturing lines and often outright or effectively owns the plant machinery. Foxconn mainly provides labor and management, so Apple doesn't need to deal with that whole mess. That's why Apple has 35% margins on hardware, and Foxconn has 5%.

And Timmy's core design philosophy is make it cheaper, nothing else matters. Apple certainly won't raise their costs just for the sake of more control over their products.

That's not true at all. Apple Silicon for their desktop products almost certainly costs more than the Intel equivalents. They designed an entire custom interposer (UltraFusion) for Mac Studio and Mac Pro, products that together move maybe a million units a year max, with a very generous estimate. I bet their in-house GPU team costs more than what I'm sure was a very good deal from Imagination (actually, they're cross-licensing now, so Apple is paying Imagination and their own team for the sake of control). They developed Metal instead of relying on the open-source OpenGL. They're planning on a wireless SoC that will likely cost more than the commodity parts they get from Broadcom and Skyworks now. Intel's modem division was Apple's second-biggest purchase ever. Everything about Apple under Tim Cook is about controlling the roadmap at all costs.

Even with Apple Silicon, most of the design work is not done in-house.

Umm... what? Then who is doing it? Apple works closely with TSMC but they do the core design themselves, unlike most ARM chip vendors.

 

[HiVolt]

If Apple is going to do anything in the automotive sector, I hope it would be a partnership with a luxury and a mid range brand, to create an "Apple Edition" trim level.

Basically an interior treatment with Apple style, specific Apple UI with integrated Apple services & connectivity, and some minor exterior Apple Edition markings.

 

[5232152]

My gut tells me the idea of an Apple car is better than an actual Apple car, unless they littearly only delivers the entertainment syste, color scheme and that's it.

 

[svish]

Don’t think Apple Car will be out any time soon. 2025 seems early.

 

[CordovaLark]

Especially one involving a company which thought the butterfly keyboard was a wonderful idea and flogged it for three years...

And created the wonderful AI that is called Siri.

 

[erthquake]

apparently neither do you

Who's autonomy platform do you consider more capable than Tesla's?

 

[nsayer]

Whose goal is this? Apple certainly hasn't announced anything of the sort. This sounds like an artificial goal ginned up by a rumor journalist.

 

[nsayer]

Level-5 self driving cars are NEVER going to happen.

They absolutely are going to happen.

Whether they happen in my lifetime or in the absence of a ban on human drivers (the hard part about self-driving cars is dealing with the knuckle headed human drivers) is the question.

 

[wonderings]

Why does anyone want an Apple car? Not like we are in short supply of manufactures. If Apple was, and I don't believe the article at all, but if they were to actually make a car they would be announcing it now to start selling to release in 2025. A company with no history in the automotive world does not just release a car. You would need a massive network of dealers and service centres as you can be sure they would not make it self service and certainly no one but an Apple tech would be able to open the thing. I am completely happy without Apple in the auto business self congratulation themselves even more for how they believe they are making the world a better place.

 

[jole]

Level-5 self driving cars are NEVER going to happen.

Cruise and Waymo both have L4 cars already driving passengers around (with no-one in the drivers seat).

Going from L4 (limited area of operations) to L5 (driving anywhere a human could) is not that big of a leap considering their respective autonomous technologies. It should not take too many years when we start debating if a human could drive in some particularly bad snowstorm or flood an autonomous car has refused to drive in.

 

[huge_apple_fangirl]

And created the wonderful AI that is called Siri.

Apple didn't create Siri, they acquired it. And AI... let's not get ahead of ourselves here.

 

[6787872]

Who's autonomy platform do you consider more capable than Tesla's?

having nothing is better than the barely functional crap tesla has. but if i had to pick, waymo.

 

[jole]

I consider myself an Apple fanboy but I wouldn't touch a car made by Apple for the next ten years for the same reason I wouldn't touch a car made by Kia or Hyundai until recently—they need to demonstrate enough reliability just as a car before I'd even consider it.

In my experience, Apple has been exceptionally responsible in fixing problems their products have had - even when that have cost them a fortune (thinking of the butterfly keyboards here). Their reputation is: expensive products, but customer is not left alone when there is a problem.

If anyone, Apple would be my bet for a 1st gen car product. They have their impeccable customer satisfaction reputation to safeguard and enormous resources to do that. If they don't, people would quickly lose trust to all Apple product releases.

With the above in mind, I can imagine that a decision to release an enormously complex ~ $100k product is such a high risk that they must set the quality control bar very high internally. And even with that, they should probably restrict themselves to a very high end for many years to limit the scale of the problems. It will be easier to deal with 50,000 $100k cars sold than 1,000,000 $50k cars sold. At worst, the former is "just" a $5B problem while the latter $50B problem could sink the ship.

 

[jole]

Who's autonomy platform do you consider more capable than Tesla's?

Tesla is far behind the competition in getting to L5. They have not even started to test cars without a driver. Waymo and Cruise are not only testing, but already commercially operating without drivers.

Tesla is probably ahead of everyone else in building autonomous driving for cars limited to cameras only. This is their target due to having many cars on the road without other sensors and added cost of lidars. That said, Waymo and Cruise have not shared how well they could already operate with just cameras.

 

[erthquake]

having nothing is better than the barely functional crap tesla has. but if i had to pick, waymo.

Waymo only works in a few locations. Tesla FSD beta works anywhere in the US, hasn't killed or injured anyone like Waymo has. FSD Beta errs on the side of safety vs smooth performance.

 

[jole]

Tesla FSD beta works anywhere in the US

If "works" is defined as car fully drives by itself (like in the name), FSD beta does not work anywhere. It has a bold vision of someday doing that.

If "works" is defined as you can take your hands of the wheel for 30 seconds and will probably not die, FSD beta works. Also 1970 Toyota Corolla passes the same definition.

hasn't killed or injured anyone like Waymo has

This is news to me. Reference, please.

 

[erthquake]

Tesla is far behind the competition in getting to L5. They have not even started to test cars without a driver. Waymo and Cruise are not only testing, but already commercially operating without drivers.

Tesla is probably ahead of everyone else in building autonomous driving for cars limited to cameras only. This is their target due to having many cars on the road without other sensors and added cost of lidars. That said, Waymo and Cruise have not shared how well they could already operate with just cameras.

Cruise and Waymo only work in a few locations where they've collected hyper-HD maps. It's not a scalable solution because collecting and QC'ing that map data is expensive, time-consuming, and fragile as the cars are much less likely to be able to adapt to changing street conditions like construction zones. So while you may be technically correct in saying Waymo is ahead of Tesla in Phoenix, it's not true anywhere else.

Moreover, Cruise and Waymo are not able to collect enough miles of data to experience edge cases because they have orders of magnitude fewer cars on the road than Tesla. Then you have to consider who has or will have more compute available to train the neural nets that control driving policy. Waymo? Maybe, depending on how much resources Google decides to dedicate to the problem. Cruise? Certainly not. Tesla when Dojo comes online late this year? Almost certainly.

What matters most is total amount of training data and the diversity of that data. What matters even more is training collected with a driver in control of the car. Data from driverless test cars isn't as valuable as the data Tesla collects. Adversarial neural networks train best with lots of driver input correcting the car and letting it know when it did something wrong immediately followed by the driver showing what the correct action should be. Waymo and Cruise are lightyears behind Tesla in total data, its diversity, and human driver input. And they will never catch up because Tesla is putting hundreds to thousands of new data collecting cars on the road every month.