9 ms·

>>> Good. This was an idea that never made any sense to me. It’s high overhead/low-margin and requires infrastructure, sales and after-sales support that just a

>>> Good. This was an idea that never made any sense to me. It’s high overhead/low-margin and requires infrastructure, sales and after-sales support that just aren’t aligned with how Apple does business. Apple, in a twist of fate, is in the same position that AT&T was when it started Bell Labs, and in the same place as Xerox was when it started Parc. The history of Parc and Apple is well known, the early history of unix being a thrown out the car window by AT&T at speed to avoid more anti trust issues is often forgotten. Apples products are built on legacies. "Resting on the shoulders of giants" is probably true in this case. Apple is now an AT&T, its now a Xerox, it is now the company that can just do R&D for the sake of doing it (as can Google and a lesser degree MS). Apple, unlike google, knows how to make a consumer product, and one of these moonshots could make it even bigger... I wonder if in 20 years something from apples car project changes the industry in a way we never expected. Time will only tell.

klik99 3y ago
Yeah, I absolutely believe that Apple can bring something to car industry - I remember Elon once said that Teslas hidden strength that no other car manufacturer caught on to was treating the car as technology akin to an iPhone with incremental updates and improvements pushed over the air. I've owned a Model X for a few years and I see that approach but don't think Tesla has perfected it. Apple has the potential to innovate there. Cars, even ones that look like "traditional" cars, have been computers on wheels for a long time - for example, when you press the brakes there isn't and hasn't been for a long time a direct connection to the physical brakes. I have no idea what that would look like, and as someone who thought the iPhone was a worse version of an iPod and that the iPad filled a non-existant need I don't think I can speculate. But it just feels like an area that lags behind in terms of UX, which Apple often excels at. Edit: I'm a dumb dumb and the brakes example was bad. I don't edit my mistakes in forums so peoples replies make sense. I still think the rest of the point is valid.
- throwaway11460 3y ago
  That seems wrong. My brakes work with both batteries disconnected, though I have to push very hard. Audi made in 2014.
  - ksherlock 3y ago
    With power brakes, there's a brake booster (usually vacuum powered) which applies most of the force. When the engine is off (or if you have a broken brake booster) braking is entirely on you.
    - throwaway11460 3y ago
      Indeed. The brake boost works one or two times right after engine power down, it's kinda funny - I like to push the brake before getting out just to feel it harden.
- davitocan 3y ago
  Small nitpick but almost every car today still has brakes that work via direct force vs by wire. https://en.m.wikipedia.org/wiki/Brake-by-wire https://en.m.wikipedia.org/wiki/Brake-by-wire
- gambiting 3y ago
  >>for example, when you press the brakes there isn't and hasn't been for a long time a direct connection to the physical brakes I'm sorry but you're getting it completely wrong. Brake by wire isn't legal anywhere in any market of the world and consequently there are zero cars implemented this way - every car currently on sale everywhere has to have a physical connection between the brake pedal and the actual brakes. Throttle by wire on the other hand - sure. Nearly all new cars only have an electronic throttle. Steer by wire is making progress, with Lexus making the first road legal car that has a steering wheel that's fully disconnected from the steering rack, with only electronic control.
  - klik99 3y ago
    Thank you, yes that was probably literally the worst example I could have given. But isn't there famously brake by wire in toyota EVs? And that physical connection is required as a fallback?
    - gambiting 3y ago
      Yes, they always have a physical fallback :-) and don't worry the systems like that do exist just not with brakes - it's an easy mistake to make.
  - iknowstuff 3y ago
    Lexus has a backup mechanical link it engages when a fault is detected. The Tesla Cybertruck is the first vehicle on the road with no such backup. They rely on redundancy of motors, power, and compute instead
    - gambiting 3y ago
      Does it have a backup link? According to this video: https://youtu.be/agMrewRJTow?si=DrjfGsrKp5_ZazB- https://youtu.be/agMrewRJTow?si=DrjfGsrKp5_ZazB- There isn't a mechanical connection of any kind - the car just relies on triple(!) Backup for every component, including three fully independent power sources.
  - 3y ago
- mfashby 3y ago
  > there isn't and hasn't been for a long time a direct connection to the physical brakes. I don't think this is true for most vehicles. Skoda Octavia 2020 for example still has hydraulic brakes. I was going to suggest that no mass-production vehicles use a brake-by-wire system but I would have been wrong, according to wikipedia: https://en.wikipedia.org/wiki/Brake-by-wire https://en.wikipedia.org/wiki/Brake-by-wire
- whycome 3y ago
  I wish more comments were like this. Misjudging the iPhone etc probably just comes down to a lack of information about broader needs. Its not always a bad take. I think Apple can provide an emphasis on modularity for the car manufacturers. What if you literally had a place to slip in an iPad as your display? What if cars legitimately came with VisionPro type headsets - if not for the drivers, then the passengers. What if apple encourages people fr.shifting away by car ownership by making it easier to jump into a car and have it show all your customizations (lighting, adjustments, heads up display arrangements). Another car company could make the shell and the power train. Apple would handle the "experience" and integrate it into its other offerings. Cloud Cars.
  - XorNot 3y ago
    This is just CarPlay or Android Auto though. > What if cars legitimately came with VisionPro type headsets Or what if the people who wanted that just bought their headset of choice? > What if apple encourages people fr.shifting away by car ownership by making it easier to jump into a car and have it show all your customizations (lighting, adjustments, heads up display arrangements). Question: when you get into someone else's car, is the first thing you notice the lighting, or is it how clean it is?
- _ea1k 3y ago
  The accelerator likely would have been a better example. I'm not sure that any car currently shipping has a direct connection between the accelerator and a throttle.
vineyardmike 3y ago
Embarrassingly, I was eves dropping on a (loud) woman in a coffee shop who worked on the Apple Car project. She did SLAM work. I was evesdropping because that was a core part of my thesis and I was curious. She told her dining partner she was moved to the Vision Pro team to build the inside-out tracking. Research for the sake of research sometimes pays off. Who knows if this product will recoup the billions spent on R&D but that’s how the game works.
- babyshake 3y ago
  There was a recent YC podcast about Vision Pro where SLAM and the overlap with autonomous vehicles was explained.
  - tambourine_man 3y ago
    Link? I didn’t know YC had a podcast and Google didn't help.
    - vineyardmike 3y ago
      https://www.youtube.com/@ycombinator/videos https://www.youtube.com/@ycombinator/videos
      - tambourine_man 3y ago
        Most recent episode? Tangentially, it's funny to me that people link to YouTube for podcasts these days.
        vineyardmike 3y ago
        It was at the time of the comment the most recent video on the YouTube channel - "Apple Vision Pro: Startup Platform of the future?"
  - whycome 3y ago
    Ohhh. I thought the commenter was using some sort of slang about the quality of their work. For others out of loop: https://en.m.wikipedia.org/wiki/Simultaneous_localization_and_mapping https://en.m.wikipedia.org/wiki/Simultaneous_localization_an...
pjmlp 3y ago
It even sells goods at Xerox prices!
dmix 3y ago
There's always going to be the debate between entrepreneurship vs "intrapreneurship". Whether large companies can successfully develop what amounts of startups in-house with separate R&D labs vs buying startups. aka The Innovators Dilemma https://en.wikipedia.org/wiki/The_Innovator's_Dilemma https://en.wikipedia.org/wiki/The_Innovator's_Dilemma The solution the follow up book to this one ^ was to build isolated teams that are flush with the resources/capital of the parents but aren't at the whims, or constant meddling, of the parent company's management class / stockholders / old ideas. For all the effort Google/Apple/Meta/Xerox/Bell/etc put into their internal moonshot divisions the whole concept mostly hasn't been very successful. But at the same time they also haven't been great at buying young up and coming startups either, often ruining them the second they arrive by the same impulses which demands R&D moonshot teams be isolated from the host.
- bluGill 3y ago
  Large companies often are significant investors in small R&D moonshots. Because they own a large part of the startup they can choose to buy it if the product turns out profitable, or they can sell. And since it is a separate investment when things go south they can stop investing and thus save money without appearing to let anyone go.
sydbarrett74 3y ago
When Ma Bell existed, Wall Street wasn't as myopic and short-termist. The stock market tolerated private-sector R&D because it mostly consisted of investors rather than speculators. These days, if something doesn't bear immediate fruit, the Masters of the Universe want no part of it.
- psygn89 3y ago
  And when it does bear fruit, many big tech companies come to copy or purchase it only to have it wither in some corner of their company.
brian_cunnie 3y ago
> unix being a thrown out the car window by AT&T at speed I beg to differ; when I was at AT&T in the late eighties we tried to enter the computer business, on the PC side with Olivetti, and on the UNIX side with the 3B2 minicomputers made with our own chips running System V UNIX. I even had a 3B2-200 in my house for a couple of years. Entering a new business is hard: although the engineering was solid, we didn't have a saleforce trained in selling computers, and we didn't have a rich ecosystem (Oracle! Ingres! Informix!). AT&T didn't throw UNIX out the window, but found that capitalizing on UNIX was hard.
samstave 3y ago
you're (mostly not) wrong. If anything of !35 years of cyberpunk input - and output building shit that spies on you: Apple is Ono-Sendai and your iPhone is your deck. - the real battle is going to be how much agency does a Human Being have over all PII - all their data? Your jack-in is your screen. Your data is your ID. But, who are YOU?
StableAlkyne 3y ago
It's an interesting parallel because part of the unspoken agreement between AT&T and the feds was that they had their monopoly, but had to use the funds for some kind of public good. Bell Labs was part of the public good - and why they employed multiple physicists and materials scientists whose employment involved basically researching whatever they wanted as long as it had some link back. Shannon was hired after his master thesis (which basically created the field of information theory) and, among other things, had a side project involving the application of computers to chess. They constructed the New Jersey lab specifically to encourage watercooler conversations and deliberately had greenhorns to work with the most senior researchers like Shannon. Bell Labs solved engineering problems needed by AT&T and Westinghouse, but they had the financial security to spend money on incredibly theoretical projects like transistors, operating systems (unix), and programming (C). Those pie in the sky projects would both benefit AT&T through automation, and covered the public good requirements of their monopoly. AT&T (owner of Westinghouse and Bell Labs) then proceeded to take their monopoly and patent factory, and started buying up competitors and new small companies. Eating their golden goose in this way is what caused the government to break them up. Bell Labs was independent for a few years doing... Stuff. Spent their remaining prestige on falsification scandals because of the publish or perish culture this new profit motive created. They were bought by Nokia a few years ago (now called Nokia Bell Labs) and now only employ a couple theoretical physicists last I read. The lab that put into practice the foundations of modern tech (Unix and C are in almost every non-consumer-facing device) just does some Nokia product development nowadays. What a loss.
- Animats 3y ago
  > It's an interesting parallel because part of the unspoken agreement between AT&T and the feds was that they had their monopoly, but had to use the funds for some kind of public good. It wasn't an unspoken agreement. It was an explict one, the Kingsbury Commitment.[1] [1] https://en.wikipedia.org/wiki/Kingsbury_Commitment https://en.wikipedia.org/wiki/Kingsbury_Commitment
Keyframe 3y ago
company that can just do R&D for the sake of doing it (as can Google and a lesser degree MS) I'm curious what makes you say that about MS? Out of three, I'd actually put MS first and Apple last if for nothing else than Microsoft Research which is like a separate entity in its own strong right (judging from the outside). So many things useless to Microsoft business from them, yet so many many things..
starluz 3y ago
If jobs was still around I’d be as curious as you. Under cook I have no faith
_ea1k 3y ago
The problem is that the companies that can do that are rarely also the company's that are willing to do the hard work of productionizing those moonshots. Lots of people have nice salaries and jobs that depend on pumping out ideas and PoCs, but few have a clear dependence on getting it to market. But the really big stuff all requires working like your future depends on it. They can't do that, so they usually just end up with laundry lists of fancy PoCs. Even when they work, the rollouts can be glacial at best (eg, Waymo).
complianceowl 3y ago
> Time will only tell. Tim will only tell. ...I'll walk myself out now.
filmgirlcw 3y ago
> I wonder if in 20 years something from apples car project changes the industry in a way we never expected. Time will only tell. I hope so. That would be a great result, even if the consumer product didn’t land. That said, I would argue that sometimes you do just have to call something and move on. You’re right that Apple is one of the few places that can do research for research sake, but this wasn’t a research project. By all accounts this had a real goal of making and manufacturing a car. It appears this project was in the works for more than a decade, had numerous stops and starts, leadership and focus changes and if your goal isn’t just research for the sake of research (which I do actually think is demonstrably different from Bell Labs or Xerox Parc which are more akin to things like Microsoft Research and Google X and the like), you need to ship at some point. I think giving ten years to something like this is definitely a gift and I hope we see fruits from some of the work that went into it other places. But at the same time, there are other moonshots you can try if one doesn’t work out and there is arguably a cost if you keep focused on one idea that isn’t going to pan out for too long. Ten years for a project like this seems like a fair amount of time to try and a fair time to pull the plug.

klik99 3y ago

Yeah, I absolutely believe that Apple can bring something to car industry - I remember Elon once said that Teslas hidden strength that no other car manufacturer caught on to was treating the car as technology akin to an iPhone with incremental updates and improvements pushed over the air. I've owned a Model X for a few years and I see that approach but don't think Tesla has perfected it. Apple has the potential to innovate there. Cars, even ones that look like "traditional" cars, have been computers on wheels for a long time - for example, when you press the brakes there isn't and hasn't been for a long time a direct connection to the physical brakes. I have no idea what that would look like, and as someone who thought the iPhone was a worse version of an iPod and that the iPad filled a non-existant need I don't think I can speculate. But it just feels like an area that lags behind in terms of UX, which Apple often excels at. Edit: I'm a dumb dumb and the brakes example was bad. I don't edit my mistakes in forums so peoples replies make sense. I still think the rest of the point is valid.

throwaway11460 3y ago
That seems wrong. My brakes work with both batteries disconnected, though I have to push very hard. Audi made in 2014.
- ksherlock 3y ago
  With power brakes, there's a brake booster (usually vacuum powered) which applies most of the force. When the engine is off (or if you have a broken brake booster) braking is entirely on you.
  - throwaway11460 3y ago
    Indeed. The brake boost works one or two times right after engine power down, it's kinda funny - I like to push the brake before getting out just to feel it harden.
davitocan 3y ago
Small nitpick but almost every car today still has brakes that work via direct force vs by wire. https://en.m.wikipedia.org/wiki/Brake-by-wire https://en.m.wikipedia.org/wiki/Brake-by-wire
gambiting 3y ago
>>for example, when you press the brakes there isn't and hasn't been for a long time a direct connection to the physical brakes I'm sorry but you're getting it completely wrong. Brake by wire isn't legal anywhere in any market of the world and consequently there are zero cars implemented this way - every car currently on sale everywhere has to have a physical connection between the brake pedal and the actual brakes. Throttle by wire on the other hand - sure. Nearly all new cars only have an electronic throttle. Steer by wire is making progress, with Lexus making the first road legal car that has a steering wheel that's fully disconnected from the steering rack, with only electronic control.
- klik99 3y ago
  Thank you, yes that was probably literally the worst example I could have given. But isn't there famously brake by wire in toyota EVs? And that physical connection is required as a fallback?
  - gambiting 3y ago
    Yes, they always have a physical fallback :-) and don't worry the systems like that do exist just not with brakes - it's an easy mistake to make.
- iknowstuff 3y ago
  Lexus has a backup mechanical link it engages when a fault is detected. The Tesla Cybertruck is the first vehicle on the road with no such backup. They rely on redundancy of motors, power, and compute instead
  - gambiting 3y ago
    Does it have a backup link? According to this video: https://youtu.be/agMrewRJTow?si=DrjfGsrKp5_ZazB- https://youtu.be/agMrewRJTow?si=DrjfGsrKp5_ZazB- There isn't a mechanical connection of any kind - the car just relies on triple(!) Backup for every component, including three fully independent power sources.
- 3y ago
mfashby 3y ago
> there isn't and hasn't been for a long time a direct connection to the physical brakes. I don't think this is true for most vehicles. Skoda Octavia 2020 for example still has hydraulic brakes. I was going to suggest that no mass-production vehicles use a brake-by-wire system but I would have been wrong, according to wikipedia: https://en.wikipedia.org/wiki/Brake-by-wire https://en.wikipedia.org/wiki/Brake-by-wire
whycome 3y ago
I wish more comments were like this. Misjudging the iPhone etc probably just comes down to a lack of information about broader needs. Its not always a bad take. I think Apple can provide an emphasis on modularity for the car manufacturers. What if you literally had a place to slip in an iPad as your display? What if cars legitimately came with VisionPro type headsets - if not for the drivers, then the passengers. What if apple encourages people fr.shifting away by car ownership by making it easier to jump into a car and have it show all your customizations (lighting, adjustments, heads up display arrangements). Another car company could make the shell and the power train. Apple would handle the "experience" and integrate it into its other offerings. Cloud Cars.
- XorNot 3y ago
  This is just CarPlay or Android Auto though. > What if cars legitimately came with VisionPro type headsets Or what if the people who wanted that just bought their headset of choice? > What if apple encourages people fr.shifting away by car ownership by making it easier to jump into a car and have it show all your customizations (lighting, adjustments, heads up display arrangements). Question: when you get into someone else's car, is the first thing you notice the lighting, or is it how clean it is?
_ea1k 3y ago
The accelerator likely would have been a better example. I'm not sure that any car currently shipping has a direct connection between the accelerator and a throttle.

vineyardmike 3y ago

Embarrassingly, I was eves dropping on a (loud) woman in a coffee shop who worked on the Apple Car project. She did SLAM work. I was evesdropping because that was a core part of my thesis and I was curious. She told her dining partner she was moved to the Vision Pro team to build the inside-out tracking. Research for the sake of research sometimes pays off. Who knows if this product will recoup the billions spent on R&D but that’s how the game works.

babyshake 3y ago
There was a recent YC podcast about Vision Pro where SLAM and the overlap with autonomous vehicles was explained.
- tambourine_man 3y ago
  Link? I didn’t know YC had a podcast and Google didn't help.
  - vineyardmike 3y ago
    https://www.youtube.com/@ycombinator/videos https://www.youtube.com/@ycombinator/videos
    - tambourine_man 3y ago
      Most recent episode? Tangentially, it's funny to me that people link to YouTube for podcasts these days.
      - vineyardmike 3y ago
        It was at the time of the comment the most recent video on the YouTube channel - "Apple Vision Pro: Startup Platform of the future?"
- whycome 3y ago
  Ohhh. I thought the commenter was using some sort of slang about the quality of their work. For others out of loop: https://en.m.wikipedia.org/wiki/Simultaneous_localization_and_mapping https://en.m.wikipedia.org/wiki/Simultaneous_localization_an...

pjmlp 3y ago

It even sells goods at Xerox prices!

dmix 3y ago

There's always going to be the debate between entrepreneurship vs "intrapreneurship". Whether large companies can successfully develop what amounts of startups in-house with separate R&D labs vs buying startups. aka The Innovators Dilemma https://en.wikipedia.org/wiki/The_Innovator's_Dilemma https://en.wikipedia.org/wiki/The_Innovator's_Dilemma The solution the follow up book to this one ^ was to build isolated teams that are flush with the resources/capital of the parents but aren't at the whims, or constant meddling, of the parent company's management class / stockholders / old ideas. For all the effort Google/Apple/Meta/Xerox/Bell/etc put into their internal moonshot divisions the whole concept mostly hasn't been very successful. But at the same time they also haven't been great at buying young up and coming startups either, often ruining them the second they arrive by the same impulses which demands R&D moonshot teams be isolated from the host.

bluGill 3y ago
Large companies often are significant investors in small R&D moonshots. Because they own a large part of the startup they can choose to buy it if the product turns out profitable, or they can sell. And since it is a separate investment when things go south they can stop investing and thus save money without appearing to let anyone go.

sydbarrett74 3y ago

When Ma Bell existed, Wall Street wasn't as myopic and short-termist. The stock market tolerated private-sector R&D because it mostly consisted of investors rather than speculators. These days, if something doesn't bear immediate fruit, the Masters of the Universe want no part of it.

psygn89 3y ago
And when it does bear fruit, many big tech companies come to copy or purchase it only to have it wither in some corner of their company.

brian_cunnie 3y ago

> unix being a thrown out the car window by AT&T at speed I beg to differ; when I was at AT&T in the late eighties we tried to enter the computer business, on the PC side with Olivetti, and on the UNIX side with the 3B2 minicomputers made with our own chips running System V UNIX. I even had a 3B2-200 in my house for a couple of years. Entering a new business is hard: although the engineering was solid, we didn't have a saleforce trained in selling computers, and we didn't have a rich ecosystem (Oracle! Ingres! Informix!). AT&T didn't throw UNIX out the window, but found that capitalizing on UNIX was hard.

samstave 3y ago

you're (mostly not) wrong. If anything of !35 years of cyberpunk input - and output building shit that spies on you: Apple is Ono-Sendai and your iPhone is your deck. - the real battle is going to be how much agency does a Human Being have over all PII - all their data? Your jack-in is your screen. Your data is your ID. But, who are YOU?

StableAlkyne 3y ago

It's an interesting parallel because part of the unspoken agreement between AT&T and the feds was that they had their monopoly, but had to use the funds for some kind of public good. Bell Labs was part of the public good - and why they employed multiple physicists and materials scientists whose employment involved basically researching whatever they wanted as long as it had some link back. Shannon was hired after his master thesis (which basically created the field of information theory) and, among other things, had a side project involving the application of computers to chess. They constructed the New Jersey lab specifically to encourage watercooler conversations and deliberately had greenhorns to work with the most senior researchers like Shannon. Bell Labs solved engineering problems needed by AT&T and Westinghouse, but they had the financial security to spend money on incredibly theoretical projects like transistors, operating systems (unix), and programming (C). Those pie in the sky projects would both benefit AT&T through automation, and covered the public good requirements of their monopoly. AT&T (owner of Westinghouse and Bell Labs) then proceeded to take their monopoly and patent factory, and started buying up competitors and new small companies. Eating their golden goose in this way is what caused the government to break them up. Bell Labs was independent for a few years doing... Stuff. Spent their remaining prestige on falsification scandals because of the publish or perish culture this new profit motive created. They were bought by Nokia a few years ago (now called Nokia Bell Labs) and now only employ a couple theoretical physicists last I read. The lab that put into practice the foundations of modern tech (Unix and C are in almost every non-consumer-facing device) just does some Nokia product development nowadays. What a loss.

Animats 3y ago
> It's an interesting parallel because part of the unspoken agreement between AT&T and the feds was that they had their monopoly, but had to use the funds for some kind of public good. It wasn't an unspoken agreement. It was an explict one, the Kingsbury Commitment.[1] [1] https://en.wikipedia.org/wiki/Kingsbury_Commitment https://en.wikipedia.org/wiki/Kingsbury_Commitment

Keyframe 3y ago

company that can just do R&D for the sake of doing it (as can Google and a lesser degree MS) I'm curious what makes you say that about MS? Out of three, I'd actually put MS first and Apple last if for nothing else than Microsoft Research which is like a separate entity in its own strong right (judging from the outside). So many things useless to Microsoft business from them, yet so many many things..

starluz 3y ago

If jobs was still around I’d be as curious as you. Under cook I have no faith

_ea1k 3y ago

The problem is that the companies that can do that are rarely also the company's that are willing to do the hard work of productionizing those moonshots. Lots of people have nice salaries and jobs that depend on pumping out ideas and PoCs, but few have a clear dependence on getting it to market. But the really big stuff all requires working like your future depends on it. They can't do that, so they usually just end up with laundry lists of fancy PoCs. Even when they work, the rollouts can be glacial at best (eg, Waymo).

complianceowl 3y ago

> Time will only tell. Tim will only tell. ...I'll walk myself out now.

filmgirlcw 3y ago

> I wonder if in 20 years something from apples car project changes the industry in a way we never expected. Time will only tell. I hope so. That would be a great result, even if the consumer product didn’t land. That said, I would argue that sometimes you do just have to call something and move on. You’re right that Apple is one of the few places that can do research for research sake, but this wasn’t a research project. By all accounts this had a real goal of making and manufacturing a car. It appears this project was in the works for more than a decade, had numerous stops and starts, leadership and focus changes and if your goal isn’t just research for the sake of research (which I do actually think is demonstrably different from Bell Labs or Xerox Parc which are more akin to things like Microsoft Research and Google X and the like), you need to ship at some point. I think giving ten years to something like this is definitely a gift and I hope we see fruits from some of the work that went into it other places. But at the same time, there are other moonshots you can try if one doesn’t work out and there is arguably a cost if you keep focused on one idea that isn’t going to pan out for too long. Ten years for a project like this seems like a fair amount of time to try and a fair time to pull the plug.