First Driverless Taxi Service Begins Limited Trials in Singapore

[MacsRgr8]

Funny isn't it... that back in the late 40's people were afraid of big brother. That humans were to be controlled... see 1984.
Even Apple used that in their ads in the 80's referring to IBM.

But, high-tech which is intended to make our lives easier is more and more taking away the fun of freedom of life. Self-driving cars is one of those things. Also your iPhone knows where you are, so others can find out too, especially of the device is corporately managed (yes, I know Apple's privacy statements on iOS management, but in practice most users are stunned once they see what an IT-admin can see regarding the managed iPhone or iPad).
A self-driving car also always knows where it is (hopefully...lol) and can easily send data of usage of that car to anyone who is interested.

No way a self-driving car will slide sideways out of a 90 degree turn, doing 30 mph more than the sign said you may. No way a self-driing car will do anything which:
a) is remotely unsafe, or
b) use more power than necessary.

Being "connected" anyplace and anytime really makes Big Brother possible. And it's voluntary too.

It's all so boringly logical:
Self driving cars should be safer and more friendly to the environment. So, it must be sane way to go.

But, where's the fun?!?

 

[dsburdette]

I'm going to bet someone said that in the early 1900's about the new flying craft that flew through the skies.......

Sure, but are planes programmed to fly side-by-side in 6 lanes of traffic? Why don't they have planes take off in triplet and fly in unison from NY to LA all while approaching unknown variables like drones, parachuters, etc. etc. etc.

 

[amirite]

Ha! I'm trying to get my head around the fact that so many fans of Apple (of all companies) are afraid of innovative technology.

Autonomous cars are going to happen. And they *will* be safer than human drivers.

Now I'm not so sure they're ready for real world usage. A few accidents will surely be blown out of proportion (while ignoring the frightening number of everyday accidents already happening) so I guess we'll have to watch.

My advice buy shares of those heavily invested in autonomous cars after each accident. This tech is inevitable but so are short sighted investors with little clue of where we're headed.
[doublepost=1472149985][/doublepost]

Sure, but are planes programmed to fly side-by-side in 6 lanes of traffic? Why don't they have planes take off in triplet and fly in unison from NY to LA all while approaching unknown variables like drones, parachuters, etc. etc. etc.

Well you better let Apple Google Uber et al. know they're wasting tens of billions of dollars right now!

 

[shartypants]

There is always a person inside to "take control if necessary" - um, so there is a driver. I am looking forward to the day when it is truly driverless.

 

[producer26]

-Siri be careful.There is a granny trying to cross the street
BANG
-I'm sorry, i couldn't find a post for "Grammy Crosby"

 

[ArtOfWarfare]

Not to fear monger, but...
A fully autonomous car programmed to drive to a populated area... loaded with explosives. Kinda makes terrorism and easier proposition without the prospect of sacrifice.

I thought about this a few weeks ago, and I concluded it's a nonissue. There's a lot of reasons:
#1 - Cars cost a lot of money.
#2 - Autonomous cars are connected cars. Once your car blows up, authorities immediately know who you are. They'll be hunting for you, and you're definitely never going to be able to buy a car again.
#3 - Autonomous cars can't be stolen. A few weeks ago two Teslas were successfully stolen. Part of that process was cutting off its cellular connection, which means they've lost some of their partial autonomy. They can still stay within a lane and avoid obstacles, but the GPS is gone, meaning they don't know where they are within the world, and they won't know how to get to their destination.

Possibly terrorists could acquire non-autonomous cars and turn them into autonomous cars. But this would require them to hire teams of smart people. I think it's unlikely they could recruit enough people. If they can, they still need to acquire GPS receivers. Given the GPS network is operated by the US army, I'd be shocked if the army didn't have a database of who owns each receiver and where that receiver is in the world. They can then go to the manufacturer and start figuring out who's purchasing all the GPSs.

 

[69Mustang]

I thought about this a few weeks ago, and I concluded it's a nonissue.

You concluded incorrectly, and you're waaaaaaaaaay overthinking a simple hypothetical problem.

There's a lot of reasons:

There may be a lot of reasons, but none of them are listed below.

#1 - Cars cost a lot of money.

Wait... someone would have money for explosives but not a car? Fake ID Rental. Take hostage with autonomous car. Remember terrorist.

#2 - Autonomous cars are connected cars. Once your car blows up, authorities immediately know who you are. They'll be hunting for you, and you're definitely never going to be able to buy a car again.

See answer above. Ter-ror-ist. I seriously doubt a terrorist is worried about being able to buy a car again.

#3 - Autonomous cars can't be stolen. A few weeks ago two Teslas were successfully stolen. Part of that process was cutting off its cellular connection, which means they've lost some of their partial autonomy. They can still stay within a lane and avoid obstacles, but the GPS is gone, meaning they don't know where they are within the world, and they won't know how to get to their destination.

Stick a gun in a car owners face and take their car. Why would they need to do anything you mentioned? The hypothetical is just blowing up the car, not trying to keep it.

Possibly terrorists could acquire non-autonomous cars and turn them into autonomous cars. But this would require them to hire teams of smart people. I think it's unlikely they could recruit enough people. If they can, they still need to acquire GPS receivers. Given the GPS network is operated by the US army, I'd be shocked if the army didn't have a database of who owns each receiver and where that receiver is in the world. They can then go to the manufacturer and start figuring out who's purchasing all the GPSs.

Too much TV my friend, too much TV. Why would this hypothetical terrorist need to do any of this when:
1. See AoW in his lovely new Model 3. Follow him home.
2. Take AoW hostage. Steal his Model 3 (leave AoW in it - knocked out)
3. Load it up. Blow it up.
4. Go buy Model 3 since no one knows who you are.

 

[SteveW928]

IN-SANE!!! Can you believe this is happening? The future is here!

Insane is right! Lack of common sense is what's here.

Also, as others have said, driverless cars dependency on technology scares the bejeebus out of me. From software glitches to malicious hacks...

To the fact that it just won't ever work the way the futurists are imagining. Much better sensors, all automated, take out the unexpected factors... and it would work fairly well. Aside from that, it will be mayhem. There are only a few of these things around right now, and the problems are already showing up.... and *MOST* of them are driving around an extremely well mapped area with perfect weather, at under 20 mph like drunken first-day drivers-ed students.

Of course I understand the safety concerns, and the overpopulation issues that mean we can't keep these things go uncontrolled.

Overpopulation is baloney... we're actually facing the opposite problem in the relatively near future. And, safety concerns, yes, *some* of these technologies *could* be used to improve safety a lot. Fully autonomous vehicles, though, are a huge safety hazard.

One advantage they have is better reaction times than humans. They become smarter as more driverless cars are added to the roads because they could be programmed to talk to each other, automating each other. And you can't seriously tell me you don't believe failsafes won't be installed.

Heh. Yes, they can detect something with their sensors quickly and look at the programming to see what action to take in response. The problem is in the programs being correct and complete enough. Yes, more automated vehicles will help each other, but that won't help much for the non-automated vehicles, pedestrians, animals, etc.

Failsafes? Like what? I'm thinking maybe you're thinking these things are actually thinking. They'll do what they are programmed to do. If that's move forward 20 feet, and the sensor doesn't pick up you up, it will move 20 feet right over you.

It's expected that Tesla will release a massive update to the system within a month that'll resolve most of the blindspots in the version that's currently released (although they still don't expect to have a perfect system for another year or two.)

There won't ever be a perfect one. It will just be better... whatever that means.

There will be deaths but the robocars don't need to be perfect. They just need to be better than humans and humans suck at driving.

I hope this goes widespread while I am still young enough to enjoy it. Studying on my commute. Or sleeping. Having kids picked up from school by robocars.

Downtowns reclaimed from
Parking lots, but people suburbs booming because commute time does not matter. Options to choose schools outside districts because commute to school not matter. Housing prices detached from school districts. The potential changes from robocars are absolutely amazing. It is probably the next most disruptive technology since cell phones.

Yikes.... Kurzweil must have gotten ahold of this one. They will be better at humans *in some regards* but they'll never be as good a human driver in other regards.

I still cannot believe how anyone who has ever worked with computers and been in traffic can honestly believe that autonomous cars could work.

I might be persuaded to see a tiny chance if all manually steered cars were replaced at once, but otherwise...

Bingo.

Ha! I'm trying to get my head around the fact that so many fans of Apple (of all companies) are afraid of innovative technology.

Maybe it's because we understand a bit about it, and I don't know, have some COMMON SENSE!

I thought about this a few weeks ago, and I concluded it's a nonissue. There's a lot of reasons:
#1 - Cars cost a lot of money....

Umm, terrorists have lots of money. And, that's even w/o Hillary funding them!

 

[kdarling]

Downtowns reclaimed from Parking lots, but people suburbs booming because commute time does not matter. Options to choose schools outside districts because commute to school not matter. Housing prices detached from school districts. The potential changes from robocars are absolutely amazing.

Of course, a good mass transit system also works, is greener and takes up less road room.

It is probably the next most disruptive technology since cell phones.

Remember when the inventor said that about Segways? That they'd change the way cities are built?

--

That said, my wife hates driving and because of allergies would love a self-driving car of her own. In fact, I think that owning one is better than renting one on the spot. If you're going to rent one each time, then there's little reason not to call a vehicle with a regular driver.

I'd also like to thank Google for spending the last decade lobbying to get laws changed to allow autonomous vehicles in the US.

 

[SteveW928]

Of course, a good mass transit system also works, is greener and takes up less road room.

I wish we'd work on that first. Or, at least *first* making all these technologies do something useful in an assistive manner (i.e. radar or sensors to detect the moose about to cross the road and warn).

Remember when the inventor said that about Segways? That they'd change the way cities are built?

Heh, yea. But, this time, there are entire industries who seem to think this will be the case, and they are dropping $billions on it. The problem is they've been influenced more by folks like Kurzweil and Darwin than understanding a bit of philosophy (i.e.: mind-body problem).

 

[kdarling]

Loooooooooooong ago, I delivered pizzas in college. The company used some of the very first tiny lightweight Hondas, and groups of guys would wait for us to go inside a dorm, then they'd pick up the car and hide it behind some tall bushes. We'd come out and wonder where our car was. Cute.

But it makes me visualize exactly what's going to happen with autonomous cars:

Criminal gangs will call them to isolated areas, where they promptly disable, dismantle and sell them for parts!

--

Next up: laws will be passed allowing autonomous cars to install non-lethal protection measures

 

[VideoFreek]

Forget about opinion though. Once there is enough data, accident and fatality rates could be compared between traditional and driverless cars. Time will tell whether you are right or wrong.

One would hope for such a rational and data-driven approach, but I fear you are wrong. While the technical hurdles to achieving autonomous driving remain considerable, I think the societal and legal hurdles are far more daunting.

The problem begins with the attitude exemplified here by dsburdette and others. Essentially, many will demand that driverless cars achieve perfection, when in fact the system they'd be replacing is FAR from perfect. Traffic fatalities in the USA peaked in the late 1960s/ early 1970s, with 50-55 thousand deaths/year, or a rate of about 4-5 deaths/ 100 million vehicle miles traveled (VMT). Since that time, fatalities have declined sharply even as miles driven have soared, such that in 2014 there were fewer than 33 thousand deaths with a rate of 1.08 deaths/ 100 million VMT! Statistically, driving is today far safer than it has ever been.

One can imagine that, properly implemented, autonomous vehicles should be able to achieve a substantial reduction in accidents--my (semi-educated) guess would be by at least an order of magnitude, or a further 90% reduction in fatalities and serious injuries. If that were to happen, we should all happily get behind the technology, right? Anything that can save thousands of lives per year is a good thing.

Sadly, I fear that won't be the case. You can be sure that each and every serious accident involving an autonomous vehicle will be front-page national news for quite some time, and God help the manufacturers when someone's kid gets run over. Public outcry will be deafening. Reactionary politicians will seek to capitalize on the fear, with calls to "take back our streets from the machines." Few in our society are capable of thinking rationally or statistically when it comes to matters of life and death.

We will also need to establish a legal framework to handle this. When an autonomous vehicle injures or kills someone, who is responsible? The owner? The automobile manufacturer? The maker of the guidance and sensor systems? All of the above? Enacting the needed legal reforms will be a political quagmire--who would want to be the politician who has to explain to one of her grieving constituents why she supported legislation that restricts their right to sue?

 

[SteveW928]

One would hope for such a rational and data-driven approach, but I fear you are wrong. While the technical hurdles to achieving autonomous driving remain considerable, I think the societal and legal hurdles are far more daunting.

Too late for rationality... they are already working on it and these things are on the streets for real. If there were rationality involved, they'd be limited to test facilities. The futurists have already sold the dream to ignorant politicians to get this far.

And, no, the technical hurdles are bigger, because they are infinite in a sense. Driverless cars will always be driverless. Bug-ridden code hurtling tons of machine around in public places. (And, the argument that it's better than a crazy-drunk doesn't comfort me much... maybe, maybe not. The solution is to get the drunk off the streets, not replace them with autono-drunk.)

The problem begins with the attitude exemplified here by dsburdette and others. Essentially, many will demand that driverless cars achieve perfection ...

No, not perfection, a driver. If we're serious about reducing vehicle deaths, there are a LOT more effective things we could do. Yes, we've made big advances in terms of auto-safety through design, but we've a way to go on stuff like road design and more importantly, policing, training, and licensing. Yes, a lot of people shouldn't be driving, but the solution isn't to substitute a slightly better inept 'pilot'.

One can imagine that, properly implemented, autonomous vehicles should be able to achieve a substantial reduction in accidents...

And, that's the problem.... lots of imagining going on without real understanding of the core problem. The list of utopian ideas that backfired is long and painful.

... God help the manufacturers when someone's kid gets run over. Public outcry will be deafening.

Not nearly deafening enough so far, apparently. And, hey, Tesla has that beta-clause people had to click through... sheesh!

When an autonomous vehicle injures or kills someone, who is responsible? The owner? The automobile manufacturer? The maker of the guidance and sensor systems? All of the above?

I'd say all of the above, and hopefully they all get the living-$*@#($% sued out of them. This is irresponsibility at the highest level.

 

[kdarling]

We will also need to establish a legal framework to handle this. When an autonomous vehicle injures or kills someone, who is responsible? The owner? The automobile manufacturer? The maker of the guidance and sensor systems? All of the above? Enacting the needed legal reforms will be a political quagmire--who would want to be the politician who has to explain to one of her grieving constituents why she supported legislation that restricts their right to sue?

Because of Google, laws are already being considered and passed in the US about who takes responsibility.

For example, some states are considering making the passenger liable (!).

--

Also, researchers have noted that before long, people are going to expect autonomous vehicles to exhibit some human-like decision making capabilities.

Consider if a car runs over a small child who darts out in front (making braking not possible in time), instead of instantly swerving into a minor crash with a pole. Most people would instinctively try to save the child even at their own risk. Especially they were a parent, or had a little sibling of their own. I would hate to be riding in a car that did not swerve.

One poster here suggested that a car should only be programmed to stop as safely and quickly as it can. I don't think that'll fly with the mass public. It'll be seen as valuing the passenger over pedestrians.

 

[SteveW928]

For example, some states are considering making the passenger liable (!).

Yea, that's a good question. If someone buys a product that's dangerous to others, not realizing it is dangerous, what happens? For example, if you buy a laptop that has a defect, and brings down a jet... is it the person who bought the laptop, or the laptop manufacturer.

In light of the marketing of this stuff... while I'd like to believe the average person is intelligent enough to recognize using it as negligence, most people clearly don't understand the danger.

Also, researchers have noted that before long, people are going to expect autonomous vehicles to exhibit some human-like decision making capabilities.

For sure. And since that's impossible, we're just going to have to adjust those expectations (and accept that software bugs or programming lapses are just going to cost lives), or not do it. I personally don't care who is responsible... I don't want to get killed by one either way.

I know driving is a risk, in general. I can't guarantee someone won't get drunk and crash into me (though I'd like to see everything possible done to prevent that). But, this is purposely putting something into action we know to be problematic. We have a choice not to do it.

 

[amirite]

How can you compare autonomous cars to Segways?

Segway failed for countless reasons. The thing was a new mode of transport that looked ridiculous and couldn't even be used in many places. Not to mention most Americans drive cars everywhere anyway.

Autonomous cars are vastly improved versions of something we already have. They are, or will soon be, better than manual cars by every objective measure. All you need to do is sit in the car as if it were a taxi or a train, and do what you need to do (work or leisure) instead of focusing on the road, manually driving the vehicle and trying not to die.

They don't require you to learn to ride a new vehicle, it's not a new form of transport. We don't need new roads as manual and autonomous cars will co-exist.

The connection you have to manual cars is purely emotional, and the fear of new technology—and speculation of all the ways it can go wrong—is nothing new.

Will they create new opportunities for criminals? Maybe, as will any new technology. People once said "Online shopping? Online banking? But...hackers!". Whatever happens, we'll learn, we'll fix the flaws, and the net benefit to society will be extremely positive. People will adjust, like they always do.

You won't even need to own a car as companies will have large fleets of vehicles everywhere. Anyone who lives in a major city knows how quickly you'll get an Uber after requesting it. These autonomous vehicles will be even more ubiquitous because of their lower cost. There'll be issues of cleanliness etc, but all these issues can be solved.

You're free to cling on to your manual cars for a few more decades if you wish. That is until they're no longer feasible to use: companies stop manufacturing them and governments ban them from public roads.

It might be scary to some of you, but change is inevitable and it's going to be wonderful.
[doublepost=1472334081][/doublepost]Or maybe you're right. We've been manufacturing cars for a hundred years—PC guys are not gonna just figure this out.

 

[SteveW928]

Segway failed for countless reasons. The thing was a new mode of transport that looked ridiculous and couldn't even be used in many places. Not to mention most Americans drive cars everywhere anyway.

I think the point was that the press made a big deal out of it and some of the 'visionaries' predicted it was going to revolutionize things. But, yea, I agree it isn't a great parallel for a number of reasons.

Autonomous cars are vastly improved versions of something we already have.

That's where I disagree. There's no improvement made. It's a downgrade as you're taking a driver with a mind, to make decisions, and trying to replace it with a very incomplete computer program.

There are sensor technology innovations, but those would better serve drivers. I'm sure some of that will happen, but way too much money is being distracted by the autonomous efforts. And, that's besides the danger it poses.

They are, or will soon be, better than manual cars by every objective measure. All you need to do is sit in the car as if it were a taxi or a train, and do what you need to do (work or leisure) instead of focusing on the road, manually driving the vehicle and trying not to die.

No, they won't. They'll be better in some regards (i.e.: where they plug a hole in human limitations... like a sensor that can check a blind-spot, or sensor that could detect a deer in the brush), but worse in others (like actually figuring out what is a road and what isn't, or 'reading' the intent of another driver or pedestrian, etc.).

Trains (automated) work because they are on tracks and the whole system is automated. Even then, it's possible for a train to hit someone who falls on the tracks, or be disrupted by nature in some way (i.e.: a tree falls in the way, and a sensor isn't scanning for it appropriately).

re: not trying to die - maybe tell that to the guy who had his head lopped off in the Tesla recently.

We don't need new roads as manual and autonomous cars will co-exist.

No, not very well. And, that's the problem. *IF* we installed sensors in all roads, or mapped them to an incredible amount of detail... and *IF* ALL vehicles were automated... and *IF* sensor technology gets good enough to detect ALL obstacles that come up unexpectedly... and *IF* programmers can account for all the possible unpredictable things that can still happen. Then, it would work pretty well.

The above is nearly impossible, though, without incredible changes to the entire system. And, even then, will it ultimately be worth it? If all of the above were accomplished, I'll agree that it would probably reduce vehicle deaths. But, we could *easily* accomplish much the same effect if we used the new technology to assist drivers, and tighten up some on licensing requirements and enforcement of existing regulations.

Even though cars are way safer than they have been in the past, most accidents are pretty avoidable. (i.e.: they are due to people being stupid, lack of training, etc.)

The connection you have to manual cars is purely emotional, and the fear of new technology—and speculation of all the ways it can go wrong—is nothing new.

Look, I've been a new technology adopter my whole life. But new tech and sci-fantasy are two different things. Yes, I do love cars and driving. I realize not everyone shares my enthusiasm about it. But, I also know a good deal about the limitations of technology and mind-body problem, etc.

To put it simply: AI can do many things, but the way the media and many people view it, is a pipe-dream. These 'autonomous' vehicles will only work as well as the sensors and the software. Have you used a computer recently???

You won't even need to own a car as companies will have large fleets of vehicles everywhere.

You know, even as a car enthusiast, that sounds wonderful. I could maybe have my collector's ePorsche in the garage (which might be impractical for daily living) and just hail a vehicle from a service when I need one; take the ePorsche out to the countryside or track on nice days. Sign me up! I'm just telling you that it isn't going to work like you think.

With enough sensor tech advancement, and extremely good mapping in some urban areas, I'm sure the taxi thing will happen. (And feel like riding with a drunken first-day drivers ed student.) As long as they go really slow and are overly cautious, too many people won't get killed. But, as a whole, this isn't going to go well. I just hope too many won't have to be sacrificed until the hype dissolves.