Dozens Of Apple Car Employees Laid Off?

[HardOCP News]

Until Apple goes on the record saying it does indeed have a car project in development, I am taking all these stories with a basketball sized grain of salt. You would think that by now there would at least be one person to come forward with proof that they worked on the project.

Apple's effort to build a car may have hit a snag, as dozens of employees have been laid off from its secret Project Titan, according to a report from The New York Times. The publications' sources say Apple is rethinking its car strategy, with the unconfirmed lay-offs being part of a "reboot", The Financial Times also reported several key staff have left the project. In the two years since rumors began of Apple's car plans, the secretive company has never confirmed Project Titan or its automobile intentions.

 

[blkt]

O battery technology, where art thou?

 

[nutzo]

Maybe they finally realized there there's no market for an over priced Apple car?

 

[morgwon]

After the feedback they received from removing the jack from the iPhone 7 they realized their plan to remove the steering wheel from the car might not be such a good idea so they fired everyone who signed off on that design.

 

[lostinseganet]

Well if they did work on the car I am sure they will sell their story to the tabloids.

 

[DukenukemX]

I bet it's because they put a headphone jack in the car.

 

[evilsofa]

They're not building an Apple Car. That's just a cover for the real project - they're building [redacted].

Edit: I just got a cease and desist so I must be on to something!

 

[DrLobotomy]

They are working on the iCoffin.

 

[-PK-]

But what does building a car mean in today's society? We've removed all of the 12v jacks, dials, and usb plugs from inside the vehicle to simplify the car's user experience. You can now charge your iPhone directly from the gas port. Want to refuel and charge your device at the same time? Our engineers have designed an accessory that allows you to charge a single device and refuel your car at the same time for $199. Have multiple devices? You can refuel and charge 4 devices* with our $399 accessory, which is offered in white or black.

 

[britjh22]

I can see it now:

Apple iCar 7, we removed the doors, because we are just SO BRAVE AND COURAGEOUS.

 

[defaultluser]

I just looked at it as a "just in case" investment. Just in case Google could actually find a large number of customers for their self-driving system, Apple wouldn't be left too far in their dust.

But Google hasn't found much out there. Many car companies are going it alone, because it's not so hard to do. And the rest that are looking for partners have several choices now, which means there's no real advantage for for early movers.

33 Corporations Working On Autonomous Vehicles

I think the last time I checked, only Fiat was going with Google. That's not the massive amount of companies Google was hoping to bag with this 10 year investment.

The full set of non-car SELLING companies on that list:

Google
Apple
Baidu
Mobileye
Intel
Bosch
Delphi
Nvidia
Microsoft
Uber

That's a long list of companies. Also listed in the article were major universities offering their expertise direct to these car companies, proving that there's no first-mover advantage to be had here. It's easy enough for there to be several dozen competitive standards that all seem to work decently.

 

[Alai]

Many car companies are going it alone, because it's not so hard to do.

HAHAHAHAHAHAHAHAHAHAHA

 

[defaultluser]

HAHAHAHAHAHAHAHAHAHAHA

Well, it's not like they have to have the car self-driving tomorrow. Everyone has started with lane keeping assist and self-park, and then they'll add features incrementally. Not so hard that they can't do it over time. And they can go hit up any of the dozen other companies for small stuff, or hit up any of the hundreds of universities working on the field. Selling your soul to a big player like Google could be a costly mistake if your competitors can do it for cheaper because they own the tech.

In other words, you can get help if you go it alone. But if you buy-in totally, then you're at the mercy of another company in a roughshod industry.

That was the mistake Google made, trying to push the state-of-the-art, and assuming everyone will buy your tech. They were hoping to corner the industry, but all they did was light the path for everyone else to follow. If a company with absolutely no experience on the subject can come out of nowhere and make a passable self-driving system in less than a decade, then other companies could do it almost as fast, even if they don't have quite the brain trust.

After all, there's only certain things you can patent. And there's a lot of software cross-pollination in any hot young industry.

 

[Alai]

a company with absolutely no experience on the subject can come out of nowhere and make a passable self-driving system in less than a decade

Are you talking about Google? If so, then you're absolutely wrong. Google has had machine learning specialists (i.e. data scientists) in their employment since forever. Besides the advancement of their search engine algorithms, one of the main uses for machine learning is automatic classifications of text (i.e. spam filters). They developed/obtained one of the best spam filters in the email space ever. And much sooner than any other company (I think mid 2000's?). Machine learning is what self-driving cars use to drive itself.

If you're talking about startups or some such, I have extreme doubts that any other company can catch up to Google in the self-driving space excluding perhaps Baidu who employs Andrew Ng, a leading researcher in the deep learning field. He pioneered the usage of GPUs for machine learning which made for vastly improved performance over super computers (i.e. merely using multiple CPUs).

Machine learning for self-driving cars has very specific and sharp logistical issues related to the massive amount and varied kinds of data required to properly run.

The closest we're going to get to a self-driving car in the next several years is going to be an automated vehicle that only runs specific routes. We won't see a true self-driving car that can drive along any road and interpret any street signs and follows state-specific laws for a long, long time. And it won't be an incremental hop in technology. It's going to be some monumental paradigm shift in the machine learning and cloud technology space because that's almost true AI.

 

[defaultluser]

Are you talking about Google? If so, then you're absolutely wrong. Google has had machine learning specialists (i.e. data scientists) in their employment since forever. Besides the advancement of their search engine algorithms, one of the main uses for machine learning is automatic classifications of text (i.e. spam filters). They developed/obtained one of the best spam filters in the email space ever. And much sooner than any other company (I think mid 2000's?). Machine learning is what self-driving cars use to drive itself.

If you're talking about startups or some such, I have extreme doubts that any other company can catch up to Google in the self-driving space excluding perhaps Baidu who employs Andrew Ng, a leading researcher in the deep learning field. He pioneered the usage of GPUs for machine learning which made for vastly improved performance over super computers (i.e. merely using multiple CPUs).

Machine learning for self-driving cars has very specific and sharp logistical issues related to the massive amount and varied kinds of data required to properly run.

The closest we're going to get to a self-driving car in the next several years is going to be an automated vehicle that only runs specific routes. We won't see a true self-driving car that can drive along any road and interpret any street signs and follows state-specific laws for a long, long time. And it won't be an incremental hop in technology. It's going to be some monumental paradigm shift in the machine learning and cloud technology space because that's almost true AI.

Right, and Google is no closer to making it work than anyone else, because nobody is ready to throw that kind of money at the problem.

The problem isn't really the algorithms, it's the amount of sensors (tell me, how do you run several thousand radars simultaneously on the same stretch of freeway?), compute power and interconnections that this will require to function , all while finding new ways to remote-control a car accurately. If it were as simple as the algorithms, this would have been completed a decade ago.

Google isn't contributing MOST of the parts I just listed above. They're just good at writing software, and they have some massive amounts of compute power already spoken for. What else do they bring to the table?

It's not like we're still in the dark ages of machine learning, where Google was treading the path alone two decades back. People have seen the light of the Goggles, and they have copied every single feature of The Goggles, and even exceeded them.

This is why you can actually get useful results from Bing today, funny that. Or you can use voice recognition AI systems NOT POWERED BY GOOGLE, like Siri (no-longer powered by Google search), Amazon Alexa, and IBM Watson just to name a few. Itf it really was hard and beyond the expense accounts of big companies, then how the hell did all this happen so fast?

And I don't see e why you give credit to only Baidu because they've done GPU Compute in the past. Today anyone can buy a GPU-compute platform from Nvidia and put it to use, they have a whole platform designed for car compute, with a development system, sensor handling and everything. What once was tough to do, Nvidia now hold your hands. How is this hard anymore?

Autonomous Car Development Platform from NVIDIA DRIVE PX2

Not difficult to get started. The hard part will be to get it working perfectly. But we're nowhere near that level of capital investment required to pull off that miracle.

So I sand behind my original assertion. Getting started with Self-driving cars is EASY today, and that's why there are currently so many different horses in the race.

 

[Kaitian]

Many car companies are going it alone, because it's not so hard to do.

The full set of non-car SELLING companies on that list:

Incorrect. Most of those companies on the list are SUPPLIERS which they want to supply components for the final assembly of a car. For example: Microsoft wants to try and leverage HoloLens to car manufacturers.