Apple Inks Deal With Volkswagen for Self-Driving Employee Shuttles

[IJ Reilly]

I didn't mean to be unpleasant. It's just that if you've worked with technology, esp. in some kind of development capacity, you'll know that it is just as problematic as dealing with people.

Software, when it is developed, goes through multiple stages of verification and testing, before it is released into production and made available to customers. Despite best efforts, software is almost always released with bugs. It is an ongoing process to keep software free from major bugs, and that process is never really complete.

So, first of all, imagine how complex driving software would be. How many use cases and scenarios to test. How many possible bugs. It is insane to think that software can ever be trusted to drive a car, no matter how many iterations of testing and fixing it goes through.

Yes, like you said, people can be idiotic drivers too, and they are. Some get behind the wheel drunk, some text while driving, some engage in oral sex while driving. But, they have an ability for creative problem solving that no computer will ever match.

I am old enough to remember when it was the conventional wisdom that desktop computers would never be powerful enough to be of any real use to anyone. Yes, it was insane to think that they ever could. Imagine that.

The other flaw in your thinking is that "creative problem solving" is the issue that needs most to be addressed here. It isn't. Your pointing to extreme human behavior while driving is both hyperbolic and reductive.

You should be able to prove to yourself both empirically and by observation that human beings on a whole are simply very bad at piloting cars. Because we enjoy the activity, our egos tell us we have skills we don't possess, they urge us to rationalize why the rules of the road don't apply to us, and they give us an excuse to be ignorant of the basic physics of the entire exercise. Computers don't need to be creative to be safer drivers, they mainly need to be more rational than we, which is the easy part, since we are fundamentally not rational where our egos are in charge.

The only reason we continue to allow ourselves to engage in the highly hazardous behavior of driving is that we have yet to develop a better way. That way is coming.

It is not insane, it is inevitable.

 

[0002378]

I am old enough to remember when it was the conventional wisdom that desktop computers would never be powerful enough to be of any real use to anyone. Yes, it was insane to think that they ever could. Imagine that.

And I am experienced enough with software and computers to know that they f**k up, all the time. Your age doesn't make you an expert.

Your desktop computers example is entirely without merit. We are talking about computers driving cars, not about computers able to play Solitaire or Minesweeper. You are trivializing the subject of computers driving cars, with your example.

The other flaw in your thinking is that "creative problem solving" is the issue that needs most to be addressed here. It isn't. Your pointing to extreme human behavior while driving is both hyperbolic and reductive.

Wrong ! Creative problem solving IS the biggest issue here. Everyday driving requires constant adaptation to new circumstances that have never been anticipated before - new drivers on the road with unique driving styles, a fallen tree/rock, hazards caused by bad weather, road closures due to construction, pedestrians crossing the street unlawfully, etc, etc, etc. No two trips to the supermarket you ever take will be the same. Try teaching a computer (i.e. a machine with FINITE power and resources) something of INFINITE scope. Can you do that ? No, didn't think so.

You should be able to prove to yourself both empirically and by observation that human beings on a whole are simply very bad at piloting cars. Because we enjoy the activity, our egos tell us we have skills we don't possess, they urge us to rationalize why the rules of the road don't apply to us, and they give us an excuse to be ignorant of the basic physics of the entire exercise. Computers don't need to be creative to be safer drivers, they mainly need to be more rational than we, which is the easy part, since we are fundamentally not rational where our egos are in charge.

Yes, the human ego definitely does get in the way of rational thinking. That is where I agree with you. And, I agree that people can be and are often idiots on the road. However, their ability to easily anticipate, understand, integrate, and resolve an entirely new situation (again, creative thinking) in real-time, FAR outweighs the negatives of the human ego.

The only reason we continue to allow ourselves to engage in the highly hazardous behavior of driving is that we have yet to develop a better way. That way is coming.

Yes, that better way is to educate drivers and/or more stringently enforce the rules/laws. And even if humans continue to be the idiots they are, they will be infinitely safer drivers than the most powerful computer on Earth at the end of time.

It is not insane, it is inevitable.

Come back and tell me that after they run over someone you care about.

(Hopefully, I won't be around when that happens.)
[doublepost=1527208834][/doublepost]

Computers don't need to be creative to be safer drivers

Have you ever driven a vehicle ?

 

[IJ Reilly]

And I am experienced enough with software and computers to know that they f**k up, all the time. Your age doesn't make you an expert.

Your desktop computers example is entirely without merit. We are talking about computers driving cars, not about computers able to play Solitaire or Minesweeper. You are trivializing the subject of computers driving cars, with your example.



Creative problem solving IS the biggest issue here. Everyday driving requires constant adaptation to new circumstances that have never been anticipated before - new drivers on the road with unique driving styles, hazards caused by bad weather, road closures due to construction, pedestrians crossing the street unlawfully, etc, etc, etc. No two trips to the supermarket you ever take will be the same. Try teaching a computer (i.e. a machine with FINITE power and resources) something of INFINITE scope. Can you do that ? No, didn't think so.



Yes, the human ego definitely does get in the way of rational thinking. That is where I agree with you. And, I agree that people can be and are often idiots on the road. However, their ability to easily anticipate, understand, integrate, and resolve an entirely new situation (again, creative thinking) in real-time, FAR outweighs the negatives of the human ego.



Yes, that better way is to educate drivers and/or more stringently enforce the rules/laws. And even if humans continue to be the idiots they are, they will be infinitely safer drivers than the most powerful computer on Earth at the end of time.



Come back and tell me that after they run over someone you care about.

(Hopefully, I won't be around when that happens.)
[doublepost=1527208834][/doublepost]

Have you ever driven a vehicle ?

Again, you are engaging in reductive logic. Software doesn't need to be perfect, except in the ego-defined world where only perfection beats imperfection. In the world of logic, the correct criterion is better, not perfect. Human beings are far from perfect, so why are they driving?

You don't need to recite the massive complexity of the problem. I think we're all entirely aware of it. So are the many high technology companies that are spending billions trying to solve it. Because it will be worth it. Or perhaps they are just too dumb to realize that their entire effort is futile?

Human beings are clearly not good at this task. Again, the empirical evidence of the many thousands who die every year at it should help you get this point. So would be the futility of trying to "educate" drivers better. This has been going on for a century to little avail, and even the huge improvements in the safety of automobiles has done little to stem the tide. We are not good at it. Period. Some are prepared to acknowledge this fact, others are in denial. Ego is the culprit.

I write in compete thoughts, so if you wish to discuss this with me further, please don't dice them into chopped up disconnected thoughtlets. Thanks.

 

[0002378]

I think we're all entirely aware of it. So are the many high technology companies that are spending billions trying to solve it. Or perhaps they are just too dumb to realize that their entire effort is futile?

Yes, they are too dumb ! They, like you, are only looking at the surface layer of reality - the "coolness" of driving software, the $$$, and how closely they can make the world resemble a videogame or a Sci-Fi movie. You think they give a $#!T about the people who are going to die as a result of their actions ?

They are unable to see beyond their selfish motives, and you are unable to see beyond your loyalty to them.

I wish those geeks sitting at their desks working on this shameful software would get out into the real world once in a while ... then they'd see how wrong what they're doing is.

I write in compete thoughts, so if you wish to discuss this with me
further, please don't dice them into chopped up disconnected thoughtlets.

No thanks. It's been a pleasure.

 

[szw-mapple fan]

Uh, whaaaaa? Development phase? Every major automaker has an electric vehicle for sale. Some have multiple. I'm not even including hybrids. I'm talking strictly battery electric vehicles (BEV's).

Almost all are compliance vehicles in California and most have mediocre range. At least in the US, most automakers are aiming for 2019 or 2020 for full market introduction of longer range models.
[doublepost=1527247078][/doublepost]

This article claims that talks with BMW and Mercedes broke down because Apple wanted control of design. I’m assuming that includes the look of the car. It’s not like Apple to buy fully fledged consumer products. Usually it’s buying technology components that go into building their own products. Beats of course is an exception but nobody thinks Apple bought Beats for the headphones.

What I find interesting is people focus on the software as if the hardware is the hard part. Honestly I’d have more confidence in Apple being able to design/build a physical car than I would Apple being able to design the autonomous/self-driving software platform the car would run on.

When Tesla was making its first cars, people also thought the internal software was the hard part. They hired the right people and did their best to find parts manufacturers. Even so , when production started, the bottleneck would almost always turn out to be a production issue. Apple is a consumer electronics company with no background in automotive manufacturing and it has no parts infrastructure to support one. Building from zero would be a very difficult task, even for a company like Apple.

 

[Vjosullivan]

Apple has definitely lost their way

Sounds like they need a device that can find its own way. One they can climb aboard.

 

[69Mustang]

Almost all are compliance vehicles in California and most have mediocre range. At least in the US, most automakers are aiming for 2019 or 2020 for full market introduction of longer range models.

This is an entirely unrelated and irrelevant argument. Whether the vehicles were created for CAFE compliance or have mediocre range is immaterial to the argument of development vehicles. None of the electric vehicles on sale in the US are development vehicles. There's no definition of that terminology that fits BEV's on car lots today.

 

[truthertech]

Apple had to settle for a reject. Not even Volvo but Volkswagen which is like bottom of the list of desirable auto manufacturer. At least they both share one thing in common with scandals from emissions and battery throttling.

Ha! Too funny, too trolled yourself. Volkswagen is the key company to partner with as they are the largest car manufacturer in the world.

 

[IJ Reilly]

Yes, they are too dumb ! They, like you, are only looking at the surface layer of reality - the "coolness" of driving software, the $$$, and how closely they can make the world resemble a videogame or a Sci-Fi movie. You think they give a $#!T about the people who are going to die as a result of their actions ?

They are unable to see beyond their selfish motives, and you are unable to see beyond your loyalty to them.

I wish those geeks sitting at their desks working on this shameful software would get out into the real world once in a while ... then they'd see how wrong what they're doing is.



No thanks. It's been a pleasure.

This desperate baloney is hardly surprising considering the lack of thought that went into your original arguments.
[doublepost=1527262028][/doublepost]

Almost all are compliance vehicles in California and most have mediocre range. At least in the US, most automakers are aiming for 2019 or 2020 for full market introduction of longer range models.

I know it's a thing to dismiss most of the EVs on the market today as "compliance cars," but it is clear that the manufacturers are committed to the technology now, and that should be your main takeaway. FWIW (and I think, a lot) Chevy has been selling the BoltEV (EPA range of 238 miles) for a year and a half now. I don't know if you'd dismiss that range as "mediocre" but for sure the car is fully introduced to the market. Unlike the Model 3, a person could walk into a Chevy dealership today and buy one for a reasonable price.
[doublepost=1527262829][/doublepost]

This is an entirely unrelated and irrelevant argument. Whether the vehicles were created for CAFE compliance or have mediocre range is immaterial to the argument of development vehicles. None of the electric vehicles on sale in the US are development vehicles. There's no definition of that terminology that fits BEV's on car lots today.

He did say compliance, and to an extent this is true, they help the carmakers comply with CAFE and pollution standards. The real question though is whether the manufacturers are committed to EVs outside of the compliance window. I think it's pretty clear that they see them as a big part of the future.

 

[kanki1985]

It's just that if you've worked with technology, esp. in some kind of development capacity, you'll know that it is just as problematic as dealing with people.

Software, when it is developed, goes through multiple stages of verification and testing, before it is released into production and made available to customers. Despite best efforts, software is almost always released with bugs. It is an ongoing process to keep software free from major bugs, and that process is never really complete.

So, first of all, imagine how complex driving software would be. How many use cases and scenarios to test. How many possible bugs. It is insane to think that software can ever be trusted to drive a car, no matter how many iterations of testing and fixing it goes through.

Yes, like you said, people can be idiotic drivers too, and they are. Some get behind the wheel drunk, some text while driving, some engage in oral sex while driving. But, they have an ability for creative problem solving that no computer will ever match.

Yes, pretty much a developer here for last 15 years. I agree its pretty complex to develop. However, my original post was alluding towards level-5 autonomy. To your point, There are bugs in dev, which can be eradicated, then there are use cases that will be discovered after prod, then there are gamers that game the system that need to be fixed as well. I agree nothing is fool proof and things are extremely complicated. What I said is that I would be comfortable with level 5 autonomy. how does that come trough is currently beyond my capabilities (if it were, I would contact Musk ). May be machine learning could help? I don't know. May be crowdsourcing can help? I don't know either or some alien tech...
[doublepost=1527268918][/doublepost]

If self driving cars still have issues to work out where otherwise it would take humans to just know to avoid like puddles, and pot-holes, or even buses and merging lanes, how is that better?

Humans can react to these thing better. I see no time when self driving cars will surpass human thinking. This is why you still nee someone willing to manually take over what a self driving car *can't* do. rather then self driving.. Its just assisting most of the time..

"Assisting" is ok, providing there are no bugs. If there was self driving cars that adaptive, and leant their surrounding/corrections in real time, vs just pure "must push an update out" kind of like self-healing, then that would be something..

...and who knows. That even may be the NEXT big thing..

When I say self driving, I mean level 5 autonomy.

 

[69Mustang]

He did say compliance, and to an extent this is true, they help the carmakers comply with CAFE and pollution standards. The real question though is whether the manufacturers are committed to EVs outside of the compliance window. I think it's pretty clear that they see them as a big part of the future.

No, he didn't say compliance. He changed to compliance when I questioned the claim that most automakers are in the development phase with EV's. The goalpost was obviously on wheels.

Here's what he originally said:

Since electric cars are still in the development phase for most automakers except for Tesla, I’m betting that Apple is waiting to see who will be able to electrify the best and swoop in with an acquisition to get the hardware to implement their software.

Now if you can tie the claim of compliance to that original quote, you're a much better man than me. Well, you may already be a much better man than me. Not an especially high bar if I'm honest. Regardless, he simply moved the goalpost when his original claim was challenged. No Biggy. Smalls even.

 

[IJ Reilly]

No, he didn't say compliance. He changed to compliance when I questioned the claim that most automakers are in the development phase with EV's. The goalpost was obviously on wheels.

Here's what he originally said:


Now if you can tie the claim of compliance to that original quote, you're a much better man than me. Well, you may already be a much better man than. Not an especially high bar if I'm honest. Regardless, he simply moved the goalpost when his original claim was challenged. No Biggy. Smalls even.

I see, I failed to track the intricacies of this discussion. Shame on me.

We can race to the bottom. I'll give you a head start!

 

[szw-mapple fan]

I know it's a thing to dismiss most of the EVs on the market today as "compliance cars," but it is clear that the manufacturers are committed to the technology now, and that should be your main takeaway. FWIW (and I think, a lot) Chevy has been selling the BoltEV (EPA range of 238 miles) for a year and a half now. I don't know if you'd dismiss that range as "mediocre" but for sure the car is fully introduced to the market. Unlike the Model 3, a person could walk into a Chevy dealership today and buy one for a reasonable price.

All makers are committed sure, but almost none of them will have a EV on the market that's not an compliance car for at least one or two years. BoltEV and Leafs have been around for a while, sure, but they are the exception. My point is that we won't know which car manufacturer will come on top in terms of its hardware until everyone starts to have cars competing in the space.
[doublepost=1527294254][/doublepost]

This is an entirely unrelated and irrelevant argument. Whether the vehicles were created for CAFE compliance or have mediocre range is immaterial to the argument of development vehicles. None of the electric vehicles on sale in the US are development vehicles. There's no definition of that terminology that fits BEV's on car lots today.

Being a compliance vehicle means that they are created so the maker can sell gas cars. This means that the manufacturer does not expect consumers to buy these cars in large quantities and subsequently would not spend a lot on its R&D. In fact, most compliance cars actually looses money for the manufacturer. While this is changing with many more vehicles coming that are not for compliance reasons, their tech is still not mature enough for Apple to see which one is the best as an acquisition target.

 

[0002378]

This desperate baloney is hardly surprising considering the lack of thought that went into your original arguments.

There was plenty of thought in my arguments. Your big ego just won't let you consider anything I said. In any case, if you think driving a car on a road is a videogame, then you are devoid of all intelligence and it is a waste of my precious time talking to you. So, have a great day, Sir. I've got better things to do.

 

[69Mustang]

Being a compliance vehicle means that they are created so the maker can sell gas cars. This means that the manufacturer does not expect consumers to buy these cars in large quantities and subsequently would not spend a lot on its R&D. In fact, most compliance cars actually looses money for the manufacturer. While this is changing with many more vehicles coming that are not for compliance reasons, their tech is still not mature enough for Apple to see which one is the best as an acquisition target.

Development (your original claim) does not equal compliance (your revised claim). It's a completely different argument. Plain and simple. Your compliance argument is lacking factual info.

What vehicles are you considering compliance vehicles?
What evidence do you have that companies are losing money on said vehicles?

Generalized supposition is fine. You're trying to present that (generalized supposition) as defacto truth. It ain't.

 

[szw-mapple fan]

Development (your original claim) does not equal compliance (your revised claim). It's a completely different argument. Plain and simple. Your compliance argument is lacking factual info.

What vehicles are you considering compliance vehicles?
What evidence do you have that companies are losing money on said vehicles?

Generalized supposition is fine. You're trying to present that (generalized supposition) as defacto truth. It ain't.

I am not changing my claim whatsoever. I might have been a little unclear. Most electric cars offered in the states today are compliance vehicles or semi-compliance vehicles that only existed due to California laws. They are made to so that gas cars can be sold and do not make a profit (more on this later). This means that the mass market vehicle that each car maker is making for 1 or 2 years from now will bring out their A game since they actually need to make money selling these. While some automakers are starting to spread EVs to other states, mass market vehicles are mostly still in development and not produced as of yet (Toyota , VW, BMW, Ford). Unlike the current ones on the market, these need to actually be competitive and somewhat profitable. Because the question here is whether they’ll be an acquisition target for Apple (as per my original comment) to gain car hardware expertise and parts manufacturers, Apple will have to wait for these new vehicles to be finished before looking at their options imo.

As for whether the current batch of EVs is making money, Fiat, Chevy, BMW, Ford, and others are loosing money per car. Dealerships are reluctant to sell them. Without the economy of scale to bring battery and other component prices down, electric vehicles (especially compliance vehicles) will not be profitable.

 

[Wanted797]

I agree that Apple is a "we want to change the world" company, and unfortunately, therein lies the problem. They've lost a lot of their focus. Even if Apple were able to afford enough engineers to adequately work on all the projects under their belt, and give them the time to do so, the executive-level management of so many disparate projects would be almost unsustainable.

To put it a simpler way, would you like Ford or Toyota to design and manufacture your personal computer? I wouldn't, because it's not their forte.

If you wanted to get scientific about it, you'd find that in the biological world, evolution via natural selection generally favors mutations which allow the organism to adapt to its specific environment. Very few species can survive outside the narrow region they were designed for; even we humans are limited to a thin couple of miles within the biosphere, under a small range of suitable temperatures.

In other words, Apple can take on as much as they want, of course. Whether or not they can or should succeed in all these disparate endeavors, however, is a question that neither you nor I can answer.

I can agree with the ‘spreading themselves to thin idea’. The mess of products they sell now still gives me an eye twitch. I love how Steve stripped it down.

 

[tex210]

Just think of this as a mental exercise in a.i. Robotics. It may not lead anywhere directly, but it was called one of the hardest problems to solve. Maybe what they learn here will apply elsewhere. As for spread too thin, this isn't the same company that was Apple when it was struggling to survive. There is room to play. Hell, they have the money to shut the doors and just r&d for a few years.
Edit: spelling