Self-Driving Cars: Half-Assed Automation Is Stupid and Dangerous
Also, Uber self-driving project launches in Pittsburgh

It's crazy to ride/drive in a car that you need to take control over only when it's about to crash. You may reading a map; talking on the phone; even have drifted off a bit, lulled into a dangerous complacency by the fact that the semi-automated car works just fine in most situations. Then the klaxon rings telling you to take over RIGHT NOW. Of course, you haven't been paying attention, so it will take at least a few vital seconds for you to figure out why your car's computer has panicked and summoned you to take the wheel. Good luck with that!
Yet, that is just the future that a lot of automakers seem to be aiming for as they inch toward full self-driving capability. A nice article in the Washington Post describes this fork in the road toward the self-driving car future. Some manufacturers like Audi will keep adding features that allow for mostly hands-free (but not attention-free) driving on, say, limited access highways. On the other hand, Ford, teaming up with Google, aims for fully autonmous vehicles. Initially their vehicles will operate within specified areas, such as, ride hailing services in cities, but eventually they will safely roam free on America's roads and streets.
With regard to the imperative of full automation, in my July article, "Will Politicians Block Our Driverless Future?," I reported:
[University of Texas engineer Kara] Kockelman argues that semi-autonomous vehicles, or what NHTSA calls "limited self-driving automation," present a big safety problem. With these so-called Level 3 vehicles, drivers cede full control to the car for the most part but must be ready at all times to take over if something untoward occurs. The problem is that such semi-autonomous cars travel along safely 99 percent of the time, allowing the attention of their bored drivers to falter. In an August 2015 study NHTSA reported that depending on the on-board alert, it took some drivers as long as 17 seconds to regain manual control of the semi-autonomous car. "The radical change to full automation is important," argues Kockelman. "Level 3 is too dangerous. We have to jump over that to Level 4 full automation, and most manufacturers don't want to do that. They want protection; they want baby steps; they want special corridors; they won't get that."
Consequently the first law of the robocar revolution, according to [former Google consultant Brad] Templeton, is "that you don't change the infrastructure." Whatever functionality is needed to drive safely should be on board each individual vehicle. "Just tell the software people that this is the road you have to drive on and let them figure it out," Templeton says. "Everything you must do is in software or you lose." Some self-driving shuttles confined to specific areas—airports, pedestrian malls, college campuses—will be deployed, but they are not the future of this technology.
In other news, Uber announced that it will begin deploying a fleet of 100 self-driving Volvos on the streets of Pittsburgh next month. While the cars will drive themselves, each car will have a "chaperone" to make sure that all goes smoothly and an engineer to record what is going on. As Bloomsberg reports:
In Pittsburgh, customers will request cars the normal way, via Uber's app, and will be paired with a driverless car at random. Trips will be free for the time being, rather than the standard local rate of $1.05 per mile. In the long run, [Uber CEO] Kalanick says, prices will fall so low that the per-mile cost of travel, even for long trips in rural areas, will be cheaper in a driverless Uber than in a private car. "That could be seen as a threat," says Volvo Cars CEO Hakan Samuelsson. "We see it as an opportunity."
In the meantime, the MIT startup NuTonomy has just launched the first ever public test of ride-hailed fully self-driving cars in Singapore. And like Uber's Pittsburgh Project, NuTonomy's cars will, for the time-being, have an engineer onboard to monitor and take over driving if necessary.
These are steps in the right direction. As far as I am concerned, the self-driving future can't happen too soon.
Editor's Note: As of February 29, 2024, commenting privileges on reason.com posts are limited to Reason Plus subscribers. Past commenters are grandfathered in for a temporary period. Subscribe here to preserve your ability to comment. Your Reason Plus subscription also gives you an ad-free version of reason.com, along with full access to the digital edition and archives of Reason magazine. We request that comments be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of reason.com or Reason Foundation. We reserve the right to delete any comment and ban commenters for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please
to post comments
"Self-Driving Cars: Half-Assed Automation Is Stupid and Dangerous"
Well yeah, but it's not clear anyone is really pushing that standard. Or did I fall for a click bait headline. Doh!
From the article:
The manufacturers are pushing it.
I wonder if the manufacturers realize that if we only took baby steps, we'd all be driving horse-drawn carriages with anti-lock brakes.
I bet governments don't want "Level 4" either.
Because then they will have to raise taxes to make up for all that lost traffic violation revenue.
So true, pretty sure 30% of Ferguson, Missouri's annual budget was dependent upon ticket revenue. I have a feeling that is more or less status quo across the country.
The price to register one of those vehicles will have to make up for it.
Oh, yeah.
And then there are the inspection fees, the semi-annual safety certifications and whatever else they can come up with.
I don't have a particular problem with self-driven cars in principle or practically. But I can't imagine that the regulatory situation for such vehicles will make all the safety/emissions regulatory crap we're getting with regular cars now look like nothing.
And, without some massive mandate, it's not as if self driven cars are going to replace regular cars very quickly or at all.
New technologies can't develop without government mandates and heavy taxes and regulation. It is known.
In the case of cars, anyway, that seems to be true.
Backup cameras are a fine idea. As are many other new safety features. But all this crap ends up being mandated, making new cars less and less affordable (and annoying to people like me who just aren't that into new tech gadgets).
Had a car salesman touting the backup camera mandate because, he said, over 300 children a year die from getting backed over. He didn't quite know how to respond when I was unimpressed, and told him my mother's solution to the issue. She wouldn't back the car out of the garage until all 5 of us stood where she could see us. And thus she never backed over any of us, despite the lack of a camera.
This is exactly what Tesla is pursuing. Google called this out as a bad idea over a year ago.
What, you don't want to drive a car that randomly decides to brake without any warning because a plastic bag on the highway wafted in front of the range sensor? What are you, a luddite?!
Will the chinks beat us to having Johnny Cabs on Mars?
More importantly will they still use Robert Picardo as the template?
We'll always beat them, because we have cities full of black men with varying amounts of kids to feed who can man our fleet of cabs.
Need a ride? Robert Picardo. Need a doctor? Robert Picardo. Need someone to fuck a transgender gremlin? Robert Picardo has you covered.
I am going to be pissed if the first major self-driving cab service isnt called Johnny Cab.
Prepare to be pissed. Uber has a good brand and they will leverage it accordingly.
UberJohnny?
Johnny Ub?
Well, they are already beating us in the secret war on the far side of the moon.
I'd rather not cede control over my comings and goings to third parties, nor would I be particularly confident in the security of these things from hackers.
As long as everything is onboard, security is can be reasonably assured. Unfortunately, government will likely try and create "smart" infrastructure, which is a disaster waiting to happen.
It would be really difficult to make autonomous cars work without some kind of coordinating interconnectivity between vehicles.
They'll have to. Networks cannot be guaranteed, so the cars will have to have some ability to navigate on their own.
Networks cannot be guaranteed, so the cars will have to have some ability to navigate on their own.
I think you misunderstand networks and the ability to generate and use them.
I think you miss the point that you are the backup if the network fails. Yet another reason why it's an absolutely terrible idea.
Short range intercommunication from adjacent vehicles and connecting vehicles to the internet are two vastly different things.
The first is a huge benefit, the second is a security horror.
Short-range interconnectivity is still a security risk. Any connected system is.
And consumers are going to want internet connectivity in their autonomous cars so they can watch VR pr0n on their way to work.
It's like everyone's already forgotten the lessons of Battlestar Galactica.
Will Smith warned us. HE WARNED US!
WILL WE NEVER LEARN?
+1 chunky, upside-down telephone receiver
It's like everyone's already forgotten the lessons of Battlestar Galactica.
If you're gonna bone yer car, make sure to get model #6 or #8?
So more like Brazzers Cab than Johnny Cab.
WHAT ABOUT ON THE WAY HOME FROM WORK, HUH?!
That's when the onboard bar opens up.
That is a good possibility. Then again, they may just as well want to play Forza or GTA or other driving game, like using the now disconnected actual controls in the car to play the game.
I know for a fact that they are. I attended a stakeholder meeting last year that the Virginia state government held regarding autonomous vehicles, and there was much gum flapping about smart infrastructure.
Okay, "reasonably assured"? What is that to you exactly?
Information security for an automatic system which could literally have thousands of peoples' lives in its hands around the world at any one time has never been done before. I'm honestly not sure what a reasonably secure system will look like that I would be comfortable with; every system of this complexity and size will have innumerable exploitable weaknesses.
Heck, hackers regularly gain access to financial records and those are among the most secure systems in the world....but at least nobody dies when that happens.
b: With regard to car hacking, I reported:
One of the biggest hacking vulnerabilities will be through the same V2V communications systems that some in Congress want to mandate. Hackers could seek to smuggle in malicious code as cars talk to one other. In a 2015 study, UCLA computer researchers simulated a vehicular botnet attack via vehicular ad hoc networks that could snarl autonomous vehicles into a massive full-stop traffic jam within 20 minutes of being launched. Researchers are busy working on ways to make sure that information transmitted between cars can be validated. In any case, autonomous vehicles will have their own on-board sources of data from their radars, laser range finders (lidars), cameras, altimeters, gyroscopes, tachymeters, GPS, and map databases. If V2V information conflicts with what the car is sensing, it will rely on what it sees rather than on what it is being told.
While acknowledging that V2V might be useful, Templeton is basically against promiscuous chatting between vehicles. "The first thing you teach your children is not to talk to strangers," he said. So why would you want your car to talk with a jalopy with an unknown reputation? ...
Even as lurid worst-case scenarios are conjured, it is good to keep in mind that our already vulnerable traffic system has never been significantly hacked.
Many of the really serious hacks I've seen are through vectors the designers didn't even know existed.
Predictions are fun tho.
No way self-driving jalopies will be a thing.
You may technically own your self-driving car, but you'll have little choice on how to maintain it.
At least I have a hard time imagining it working any other way.
I am less concerned about hacking than I am government. Rest assured, once all the autonomous cars are networked, government will insist on being able to override any given vehicle. Maybe they stop it and lock you in to wait for officers to arrive. Maybe they lock you in and drive you to the nearest police garage.
B: You're entirely correct. As I reported:
Another infrastructure mistake would be mandating the deployment of "smart roadside infrastructure" such as traffic lights and sensors to monitor conditions like icing on bridges and communicate the information via radio to autonomous cars. In 2015, Sens. Debbie Stabenow (D?Mich.), Gary Peters (D?Mich.), and Lamar Alexander (R?Tenn.) embraced this idea when they introduced the Vehicle Innovation Act, which included spending more than $300 million on various favored tech, including vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications systems.
Before embracing such external information systems, keep in mind that the U.S. Department of Transportation estimated in 2007 that 75 percent of the nation's 330,000 traffic lights are mistimed or use obsolete control systems. "If city and county street and road agencies can't keep traffic signals up-to-date, how long would it take them to install and upgrade smart road systems?" Randal O'Toole asked in a 2014 Cato Institute study, "Policy Implications of Autonomous Vehicles." It's all most states and cities can do to fix potholes, much less deploy and maintain sophisticated networks of roadway sensors.
B: And also this:
Templeton cites the internet as a model for how to roll out the technologies that enable self-driving cars. "The internet is a dumb network that connects smart devices," he explains. "You want smart cars running on stupid roads." Dumb networks push innovation to the edge, giving end-users control over the speed and direction of change.
Personally, I can't wait for 'smart' war devices like autonomous tanks and terrorist-hating missiles that independently select and eliminate hostile forces without putting our lads and ladies into harms way!
Imagine the glorious future of America, where we can wage war without ever putting ourselves in harms way! I see no downsides!
And before someone calls me a kook, think about it for a second. The government would absolutely do this. We're already halfway there with Drone warfare. Can you imagine what we would realistically do with this type of technology, as a species? Self-driving cars are just the tip of that iceberg. Thankfully, it's basically an impossible wet dream but I'm fairly confident that DARPA is watching these projects like a hawk.
The internet is a dumb network that connects smart devices
And those smart devices fuck up all the time. Building a system where the individual devices take pre-eminence over the inputs from other sources (V2V and V2I) will open manufacturers to final liability in each and every conceivable incident that happens.
If I'm Ford, there is no way I'm playing that game.
I loled that people are still this naive.
I like driving, it's very relaxing. Anyone who wants to make it so I'm not allowed to drive can fuck right off.
How much I like driving is inversely proportional to the number of other people on the road. Confounding variables: people who don't know how to merge, and people who intentionally cut off mergers.
"Zipper, people! Zipper!"
Nuts or meat?
Frank or beans
The one thing that pisses me off is people who race down the wrong lane along a line of backed up cars then stop when the lanes diverge and they couldn't find a place to jump in.
Funny thing is, the state of Missouri is spending money trying to convince people to do just this.
It's wildly inefficient to leave long stretches of open lane just out of some sort of communist ideal that everyone should wait in line. By far the more efficient use of time and space is to fill every lane, and zipper at the point of merge.
I routinely see people pull halfway out of the line to block the empty lane and prevent anyone from using it.
Holy crap dude, you'd get shot here for pulling something like that.
I lane split on my motorcycle, and I've never even had anybody pull something like that, let alone do it for other cars.
The only time I see people do that is if there's a backup and people start using the shoulder as a lane.
There's one spot on my commute where that happens. The merge lane itself technically ends at a point: the far right is unmarked for 500-1000 ft past it but fits two cars comfortably. Some people try to pass traffic in this unmarked lane, others occupy the space to prevent it. I don't have a good answer to this one.
That's why I loved living in Texas.
Not only do I half-block the empty lane, I stick my middle finger out of the window to anyone who wants to get by. Fuck you, this is America, wait in line.
No, no, you're thinking of Britain.
But what you're doing in effect is moving the choke point from A, to A minus X yards. You're forcing more cars onto less road, exacerbating the backup.
If the lanes truly do merge, then you are just being an asshole blocking one. Both are there to be used, fucking use them.
That doesn't apply of course to people who try to merge in at the last second from a lane that continues on.
When did courtesy and foresight become communist?
Besides, I'm not convinced you're right about it being more efficient, that sounds suspicipously like an assertion not a fact.
I can try to find the article I was reading regarding Missouri's efforts, but do you really logically think it's more efficient to leave miles of lane unused, rather than fill them with cars? It's pretty universally accepted that zippering is the most efficient way to merge traffic, not waiting in line unnecessarily.
Here you go:
http://www.pressherald.com/201.....e-squeeze/
Read the whole thing. Many assertions, no real evidence demonstrating efficiency. What did I miss?
I'll wait for you to find the article, as "pretty universally accepted " is meaningless to me.
By that logic, completely blocking a road maximizes efficiency, since it leads to tightly packed bumper to bumper traffic where there was once sparse free flowing cars.
And the opposite retarded extreme is that every highway should consist of two lanes, one going in each direction. That way no one will ever get in front of anyone ever again!
I would imagine that zippering only works up to a certain number of cars. Then you're just the asshole that's making the line that much longer.
The science is conclusive on the zipper merge.
http://www.virginiadot.org/vtr...../05-r6.pdf
'Zippering' doesn't involve one tooth jumping over 50 other teeth and works best if controlled at a single point. The state in MO was often the culprit that fucked up the zippering/merge by telling people to do it 1/2 mile before it needs to be done and then leading to people jumping back out and start leap-frogging. That really fucks up the back-log.
That's why I always pulled out and blocked in MO; because the state was the one that caused the fuck-up.
Let me state in plainly.
When you wait til the last moment to do something, then inconvenience people who had the foresight to do it already, you aren't being efficient, you're being an asshole.
So that's not just a bald assertion?
Let me put it plainly:
If you're so goddamned stupid and lily-livered that you meekly get in line and wait unnecessarily, and then have the gall to get upset about people who have more sense and balls than you, you're an asshole a milk-toast nobody who doesn't have the guts to do anything about it anyway except bitch on the internet so it doesn't really matter.
What happens when many people zipper? I know what happens. I end up pissing in a Starbucks cup.
I'm comfortable with the assertion that you're an asshole if you inconvenience people while chasing an unproven efficiency gain.
That reminds me it's almost lunch-time.
Milk toast goes great with milk steak.
Backs slowly away from the psychopath?
Nah, it really isn't a dick-measuring contest, I just like to be unaccountably hostile sometimes.
We won't agree. Somehow you think that leaving vast stretches of lane open and unused is an efficient use of resources. I can't see how you figure that at all, and your arguments keep coming back to vague unquantifiable things such as "inconvenience".
I really think you're more concerned with the politeness aspect of it, at least that's how it comes across to me. I find nothing polite or necessary about waiting in line if I don't have to.
I'm curious, would you apply this same logic to waiting in line at the supermarket? You're technically able to form your own line to the right or left of the pinch point and "zipper" in, saving some line space, but I suspect that would be frowned upon by the people around you.
That's not even remotely the same thing when the street lanes are DESIGNED TO MERGE. The supermarket analogy is more like someone trying to merge in from a lane that continues on, e.g. if the right line splits off at a highway exit. And in that case, the merger is being an asshole. But if the lanes are, again, DESIGNED TO MERGE, not using one of them is just plain stupid.
The measure of a road's efficiency isn't how much of it is full of cars, it's how many cars pass through per unit of time.
It's a wasted lane. You're going to have to merge in anyway, so might as well do it early so that you have more of a chance to get access to the lanes that are going a decent speed. Also, by merging in early, you don't have to slow down when you don't get the opening you expect at the end of the ramp. I rather enjoy tucking up on the back bumper of the guy in front of me so that the guy who tried to pass me on the ramp has to slot in somewhere behind me.
You must have the same commute I do. I'll be cutting you off later. You can recognize me as the guy with the Giant Meteor 2016! bumper sticker.
On my rear bumper, which is what you'll see.
Not when I cut your ass off in tbe lane youre speeding through.
(I don't do that)
No, you're too polite. Happy traffic jam!
You've convinced me.
I'll start cutting you off.
This is Reason, where a lot of the commenters think "libertarian" mean "being as big an asshole as possible to as many people as possible and if anyone complains, they're a commie"
Or being as big an asshole as possible to the people who are the root cause of traffic jams.
This is Reason the Internet, where a lot of the commenters think "libertarian" mean "being as big an asshole as possible to as many people as possible and if anyone complains, they're a commie"
Shut up, you Marxist fag.
Stormy, just because we think you're commie, doesn't mean we think everyone is a commie.
Here's what happens on the highway I take to get home. There are four lanes, two to go south and two to peel off to head east. There is construction going on southbound and one of the lanes closes off right at the same point that the eastbound lanes diverge. Nobody uses the far left lane, the southbound lane that get blocked, to travel along then merge. Everyone zooms along the second lane trying to find a place to merge left. When they can't find a place, they stop. Just stop right there on the highway. So now there are three lanes blocked.
Yeah, but often enough the only reason the line exists is because of the assholes trying to zipper in at the front of the line.
What I think would really help would be electronic signs explaining to drivers why there is a backup so people know if it's a smart idea to try the other lane or not.
Worst is gappers, people who leave about 5 car lengths between them and the car stopped in front of them at the light, neatly preventing anyone from getting into the turn lane for the arrow.
If you have a Clinton or Trump bumper sticker, be assured that I'll intentionally cut off your merge.
Where do people do this? That might be a shootable offense around here.
Every single fucking place along my daily commute.
1. Move.
2. Why would people do this, does it provide them some benefit?
1. I wish I could. I greatly dislike Chicago.
2. Because people are fucking stupid.
It allows them to check up on Pokemon Go.
Worst is gappers, people who leave about 5 car lengths between them and the car stopped in front of them at the light
There's an invisible 18-wheeler there, you just can't see it.
Almost equally as bad are the people who have to wait for the car ahead of them to clear the intersection before commencing their left turn in a left turn signal lane.
They're probably from a major city, where you can actually get a ticket and points for doing that (google "blocking the box").
I'm sure you're wrong. I live in Mass and we don't have that rule here.
Mass may not, bust Boston does.
That's the thing, in most places (suburbs and rural area), the rule doesn't make sense. It's sometimes necessary in dense city centers though to avoid a "deadlock" sort of situation where you can't move because your intersection is blocked, but the block can't move because their intersection is also blocked, etc.
The problem is when people from cities head elsewhere and don't change their driving habits.
Massachusetts General Laws, Chapter 89, Section 9:
Let me put it this way. I don't live in Boston and people aren't getting stuck in the intersection, douchebag. The first driver waits five seconds after the light turns green before moving. The second person waits until the first person is out of the intersection. The third person doesn't even start moving until the second person is all the way through. Then the light turns red and nobody else can go. Three people have just used enough time to get through an intersection that eight people could have gotten through.
*quotes the rule they supposedly don't have*
Good thinking doubling down on your stupidity. You may notice the rule you cited specifies that it only applies when the vehicle won't be able to exit the intersection and doesn't apply always.
Yes. The problem is that in stop and go traffic, the only way to be sure you will be able to exit the intersection if to let the guy in front of you go all the way through. If you follow him before that, you can't tell if traffic will suddenly stop, leaving you trapped in the middle of the intersection or not.
If you gamble wrong, it's $150 and 2 points.
Someone used to driving that way who moves out to the suburbs becomes the people you're complaining about.
Worst is gappers, people who leave about 5 car lengths between them and the car stopped in front of them at the light, neatly preventing anyone from getting into the turn lane for the arrow.
This happens where I live, I think it's just clueless old people.
As long as we're ranting about annoying drivers, my pet peeve are the people who decide to be "nice" and stop before a parking lot exit or side street to wave in a bunch of drivers who have been waiting to turn....at the expense of all of the drivers behind them who have the right of way. If it's a red light up ahead, fine. But last time someone did this through a green light which we both missed. This idea of being generous for a few people in your line of sight at the expense of all the people you conveniently ignore makes me think the ones who do this are probably liberals.
I have said that I want to right a manual and teach a course on merging.
Of course, the zipper is lesson #1.
#2 is getting to a speed 5 mph faster than the main body of traffic when merging into it, you can always break into a gap, speeding up into a gap is much more difficult.
This. People that slow down when preparing to merge are subhuman.
There's Zipper and then there is fuckknuckle.
What's that? You want to slowly merge into fast moving traffic at the last possible microsecond as the lane vanishes, despite having ample opportunity to merge, hundreds or thousands of feet before? Watch me not care.
I cannot count the number of times I have shouted (to no effect, of course) "put your goddamn foot on the fucking gas pedal" as I am forced to slow down to make an already ample gap even larger lest I get sideswiped by someone who thinks freeways are just larger versions of residential streets.
Well, according to the CEO of Volvo, their cars won't be self-driving-- at least in Los Angeles, until the mayor starts painting some fucking lines.
Over zealous Government is only part of the problem; I fell human nature will doom the self driving cab / car share.
For some a person in a car is a deterrent to, bad behavior theft and vandalism. Technology never takes morality human nature into account, that people can be jerks . There just so many ways one can monitor the inside of the self driving car. What happened when the care is trashed,or puked in by a drunk, some tweak think they can strip the care for stuff for pocket change for meth, or pictures pop up on the net of people having sex in the driver-less car.
This is part of why I'm so baffled by Uber's interest in self-driving cars. That seems like a terrible idea because one of Uber's biggest advantages over traditional cab companies is that it pushes all maintenance, servicing, etc. costs to the independent drivers. Those costs (plus the capital cost of acquiring the vehicles) will be internalized to Uber in a self-driving car system, and I suspect that maintenance and servicing costs will be even higher for the reasons you outlined above.
In an August 2015 study NHTSA reported that depending on the on-board alert, it took some drivers as long as 17 seconds to regain manual control of the semi-autonomous car.
Based on my observations, 17 seconds seems awfully fast.
I saw a couple of these on Tuesday and Wednesday in Pittsburgh. My rides to/from the airport were from Central Asia.
My rides to/from the airport were from Central Asia
That's a long drive.
Sorry, Central Asians. Ulugbek got a little lost, but he was a really nice guy.
Yes, we continue to see more and more features that functionally take away control at unpredictable times in the name of safety. I fail to see how that's anything other than pure madness.
That being said, the odd's of anyone being able to create a fully autonomous vehicle is likely impossible until it's made illegal for a human to drive since that's obviously and clearly the #1 hurdle to such a system.
The entire concept of autonomous vehicles is stupid. Putting a computer in charge of such a variable environment is a fools errand. Aircraft and even trains would be more likely options for full automation given that they have either more room for maneuvering or are linear predictable paths yet you don't see a lot of places that trust the system enough to actually fully automate them.
Now extrapolate that into a far more complex and varied infrastructure of your average city where there are massive potholes, drunk pedestrians, and constant pervasive unmarked road construction and you start to see that this is essentially a pie-in-the-sky total restructure of everything in the transportation market. This isn't something that's going to happen in our lifetimes without massive and pervasive federal dictates that drastically reduce your autonomy.
I want to believe it's possible, but I haven't seen anything that makes that probable without completely removing the human element.
I concur with your statements, but ultimately this technology is destined for warehouses, mining, manufacturing sites, etc. - easy to eliminate jobs, and potentially human error. The "driverless vehicle" is rah-rah pap for the masses to get them excited about the technology and demand funding.
The only place I could logically see something like this working is somewhat ironically the exact opposite of what these companies are trying to do. You could relatively easily train a computer to drive, along with all the other computer driven vehicles, across the country on a massive highway. Just choose your lane, and the vehicle could easily be guided to simply stay in that lane indefinitely at a set speed while adjusting for incoming and current traffic. It would be relatively easy to modify existing cars as well, assuming you aren't driving something made prior to fuel injection.
What's actually being attempted is the far more difficult option of having a computer navigate a concrete jungle. Which, I might add, is difficult for the most powerful organic computers in existence who actually invented the concrete jungles they navigate.
It's madness, top to bottom, and is a transparent attempt to limit the travel autonomy of individuals. This will be the nail the government attempts to hammer in to make any vehicle made before autonomous vehicles came on the market illegal, probably in some misguided attempt to limit greenhouse gasses by mandating carpooling and getting rid of low-efficiency vehicles through the backdoor.
In my view, it's that simple and is the only rational reason I can see for attempting such a massive boondoggle.
You are dead on. Bailey is not stupid and surely can see this. I am forced to conclude that he considers it a good thing.
There are two major reasons I like the concept of driverless cars.
1. For long trips. I would like to sleep/read/whatever on a 6-12 hr roadtrip. That is my normal limit before flying, probably extend it some with driverless cars, and it becomes enjoyable.
2. Getting home from the bar.
And I thought of a third:
3. Eliminate massive parking lots and drop the price of land significantly.
I think a compromise can be made. Either special, sequestered lanes to keep the stupid human drivers away from the automated vehicles-- or even roads or thru-ways where only the self-driving are allowed.
I've been critical of self-driving cars, but not because I don't think the self-driving cars are the problem, but the difficulty of dealing with the stupidity of humans, which is limitless.
If you hate the stupidity of humans so much, why do you like freedom? If human stupidity is so bad, maybe people should be controlled and saved from it?
I think everyone should be banned from the roads except me and maybe a race car driver from the 60s. No one else seems to be able to handle the complex operation of keeping a car between two lines and getting their car moving as fast as 30mph. It's a lost art.
#MorePublicTransporationForOtherPeople
I endorse this message - except Diane should be banned and me allowed to drive.
They already have these in most cities, they call it the HOV lane and is in fact feasible. However, this is a fancy way of saying 'we're going to make the literal equivalent of light rail but make you pay for your own car'. It's a likely scenario, but it's also a bullshit far cry from what such a system is being sold as.
I think these are all pretty self evidently true statements, but people are so in love with the concept that they're willing to look past all the massive problems such a system would have or how badly very comparable systems have done already. That's a problem for me, and while I'm not against innovation I am against stupid idea's that won't work.
This is right up there with solar panel roads as an idea that sounds good until you stop and think about it for even a microsecond.
Good point. I'm not at all convinced that the demand for it is sufficient to generate enough interest in overcoming the massive hurdles we've been pointing out.
...yet you don't see a lot of places that trust the system enough to actually fully automate them.
Would you board a plane that was going to fly with no human pilots? It remains a ludicrous concept. We're not there and I don't know when in the near future we will be.
Take for instance Turkish Airlines Flight 1951. One of the 737's altimeters on approach to the airport suddenly and erroneously read -8 feet. The autopilot took this to mean the plane was about to touch down on the runway and throttled the engines to idle. The plane stalled and crashed in a field short of the runway. The autopilot received incorrect information and acted. A human pilot caught the error (his failing, though, was not to keep his hand on the throttle after returning it to power; the system reverted back to idle as soon as he let go), because he has the ability to think independently. There are many examples.
Indeed sir, all fantastic points.
Freedom is in essence control. So when you give up control and the machine drive, you are giving up your freedom. Everything else is just negotiating how much freedom the machine and by extension its creator lets you keep.
J: So no giving up of control to Uber, taxi, bus, train, and airplane drivers?
Well, to some degree. The bus comes at certain times. I can only have limited items on a plane.
My car will go where I want when I want.
Yes Ron, those are all some form of giving up freedom for convenience.
Does your car still have a handcrank to save you from the tyranny of electric starters?
So this is one of those times when even though it's a complete and utter truism, we're going to attack it because John said it.
Seems legit.
His is incorrectly equating automation with loss of control, and then concluding that, say, adding up numbers by hand with pen and paper means I have more "control" than using Excel to do it.
I like the analogy of Excel.
Excel gives the user more control in the sense that they can achieve more with less knowledge. I can set Excel to add up totals and display information way faster than I would be able to do it without Excel. That doesn't mean that I want Excel to budget for me without my input though.
That's why I like automated systems in my vehicle that allow me to brake without standing on the brake pedal, or turn without using a knob on the wheel, or even a system to recommend an efficient route to take. These amplify my ability to drive, rather than attempting to replace it.
That, to me, is the difference.
This is completely opposite of the American-Euro divide in autopilot philosophy between Boeing and companies like Airbus.
Again, the human element will screw the system up rendering it less safe than just driving your car yourself. People will be texting, reading the paper, or drinking their coffee because 'the system will drive me' is what they were promised. People already do this while driving themselves and you literally already see laws making stupid behavior illegal instead of punishing people for the results of their idiocy.
You're going to have people putting total trust in a system that is deeply, deeply flawed because they were promised one thing and received a completely different product. The worst, and I mean the worst, idea's have already been halfway implemented on new vehicles.
Things like autobraking are terrible idea's unless you only ever drive under perfect conditions or are such a terrible driver that you shouldn't be driving. No amount of 'computer assistance' will make a terrible driver 'safe'. If anything it exacerbates the problems that are already inherent in the system.
The best systems are the one's that give the driver more control and more predictability of operation. That is such a truism I find it baffling it even needs to be said.
Aircraft accidents often begin at the point the plane's control transitions from the automatic pilot to the human pilot.
I thought they occurred when the plane slammed into a mountain.
That's generally when they end.
Let me know when your AI is sophisticated enough to tell the sprinklers not to come on when it's raining.
TLPB: OK.
Alarm.com might be able to help with that.
I can't wait for this for two reasons.
1- Drinking
2- Spoofing the sensors of the cars around you so that there aren't cars around you. Free, free at last...
There is no technical reason fully autonomous cars won't/can't work reliably. They will end up being safer than human operated vehicles. They won't be distracted. They won't drive drunk.
Those opposed are fear mongering luddites, plain and simple.
Sure there is. Computer vision still sucks ass compared to the visual processing of a 5 year old child, let alone an adult. There are still times when I have no fucking clue where the lane actually is (downpours, etc.), and I guarantee that without embedding special technology in every inch of pavement across the US, the autonomous car is going to have more trouble with those situations than you or I will.
Let alone when we get into more rural areas where "roads" aren't exactly paved.
Autonomous cars are going to be an urban/close suburbs thing for quite a while. Come back in 30 years and we can start talking about the further suburbs and rural areas.
They don't exist today, so I agree with you. What I am arguing is that any current technical limitations will be overcome, and without having to embed technology or repave all the roads. If humans can navigate the roads as they are then a car can will be able to do so. GPS technology needs improvement too.
This may surprise you, but not all technical limitations can actually be overcome.
I know, it's shocking but true.
What you describe is basically the birth of A.I. since any such system would need to have the senses and judgment of an organic life form to be superior. Or so tightly controlled as to allow no human input.
Pick your poison, skippy. I'm betting on option 2 since I'm not convinced A.I. is even remotely possible.
Yeah, pretty much everything that seems obvious and intuitive to us is a really difficult thing to get to work at all with a computer. Vision in particular. I have been surprised at how far it's come in the past 10 years or so, but I'm not about to trust my life to a vision based self-navigating car.
I work with robots that are pretty much little cars. And to get a fleet to operate well in a fairly controlled indoor environment is a difficult enough challenge (though one that is pretty well under control at this point). I remain fairly skeptical that we will see any self-driving cars that can handle all weather and human drivers at highway speeds on the roads we currently have anytime soon.
And I don't really want them anyway. I'd rather drive than be a passenger most of the time anyway.
You could relatively easily train a computer to drive, along with all the other computer driven vehicles, across the country on a massive highway.
We have an excellent substitute for that, right now. It's called "freight rail".
I absolutely agree, and that's one reason I usually compare and contrast with rail when talking about self driving cars. Right along with all the failures of systems of that sort.
That being said I could see a real market for cross-country or long-haul driving, not the least of which is automated road transport. Even then, it would need to be a special lane or multiple special lanes which is a massive project. And, as you say, there isn't a very good reason for doing it when rail already exists.
Cars are freedom. The two week driving vacation used to be a thing. Ordinary people could pack up the station wagon and just.....go places. Wherever you want at your own pace. Without permission.
There's no such thing as a "driver-less car". A transportation module that you don't control is just a different type of train. And progs love trains. Makes it easier to control the proles.
Cars are freedom.
Where are you going?"
"Out."
What I have really learned from this comments thread is that the old saw about "85% of people consider themselves above-average drivers" is really true.
I had a related convo recently about the efficacy of car pool lanes. My argument was basically only "Top Men" think you increase traffic flow by reducing the number of lanes. My cohort's argument was that it increased flow because, you know, if you are in the car pool lane you might be going faster.
If they really want to reduce congestion then they need to consider paying people to pick up people from their departure area and drop then near their arrival area. Meaning, get people to "carpool" by paying them to do so. Put up systems that make it easy to find a ride to where you want to go, "uber-esque" so people can simply pick someone to ride with and take advantage of the need to go from roughly a to roughly b. Stop wasting money on building more lanes or "encouraging" people with a potentially faster drive IF they happen to have more than 1 person in the car.
It may not cost less than building new roads but it will most likely end up being effective.
Turbocharged, intercooled, direct injected freedom with variable valve timing and a sweet six speed stick.