Armed Polite Society
Main Forums => The Roundtable => Topic started by: Ben on May 24, 2017, 12:44:59 PM
-
I've been getting a lot of links (see the "Why?" thread here as to why!) to stories on how we are on the verge of a major transportation changes in the next decade. All regarding autonomous vehicles, people giving up their own vehicles, etc.
The below story asks many of the same questions that I do. When I was looking at property in Idaho last week, a lot if it was wandering around. Further, neither Google nor Entune could get me to the right friggin' road in the rural area I was exploring. If I didn't take a hands on approach, I would have never gotten to a few places.
I was also thinking about things like fishing trips, or as the article says, simply exploring, especially offroad, etc. How would we do such things in the autonomous car era? There will perhaps be a workaround for it, but I would hate the idea of not being able to just aimlessly go places.
Also, if we're just going to be calling "public" cars as we do taxis and Ubers, I wonder how that will affect rural America, where things are spread out? All the discussion on autonomous vehicles sounds great -- if you live in a city or other populated area.
http://jalopnik.com/autonomous-cars-must-let-us-wander-and-explore-without-1795445877/amp
-
https://www.youtube.com/watch?v=FAvQSkK8Z8U
-
While having conventional driving controls makes sense for the near-term, as autonomous cars become more and more common, it becomes less and less likely many, many people will be interested or even able to drive a car manually.
Ability is one thing, but I don't know how realistic it is to think that people will lose interest in the ability or opportunity to control where they're going. Unless he means that people will be afraid to go somewhere unless a machine is showing them how to get there. That would be sad.
-
Ability is one thing, but I don't know how realistic it is to think that people will lose interest in the ability or opportunity to control where they're going. Unless he means that people will be afraid to go somewhere unless a machine is showing them how to get there. That would be sad.
I agree it would be sad, but I think it is realistic. Maybe not in the 10-20 year time frame they're talking about, but 20-30 years? I suspect by then it will be difficult to even get a license to drive manually, or perhaps even to buy a car for yourself. Possibly only something the very wealthy can afford to do. At least in the US, Europe, and other first world areas.
As an old guy, I hate the idea, because to me, automobile = independence. but I suspect a generation or two down the road, many young men and women will not care about that independence factor.
-
https://www.youtube.com/watch?v=eWgrvNHjKkY
-
Another consideration that I think will weigh heavily on how quick autonomous vehicles are accepted is whether people will accept not just giving up control to their car's AI, but giving it up to the collective traffic AI.
Consider that to get the max benefit of this tech, cars will be built to work together and to communicate with each other. A following distance of less than a second would be wholly reasonable if all of the cars in the area are constantly communicating their positions and upcoming maneuvers to each other. But what about when the traffic community of AI gets together to make an emergency decision? Your car has a blowout. Let's say your car decides that it can take one of two options to deal with the blowout. It can swerve to the left, cutting off oncoming traffic and possibly causing a church van full of 14 pregnant nuns to go off road to avoid you and probably roll over. The other option is for your car to swerve to the right, off a cliff, involving no others, yet killing you. Which decision do you think the collective traffic AI will choose. Will the motoring public be happy with giving up that decision making duty to the collective?
-
Another consideration that I think will weigh heavily on how quick autonomous vehicles are accepted is whether people will accept not just giving up control to their car's AI, but giving it up to the collective traffic AI.
Consider that to get the max benefit of this tech, cars will be built to work together and to communicate with each other. A following distance of less than a second would be wholly reasonable if all of the cars in the area are constantly communicating their positions and upcoming maneuvers to each other. But what about when the traffic community of AI gets together to make an emergency decision? Your car has a blowout. Let's say your car decides that it can take one of two options to deal with the blowout. It can swerve to the left, cutting off oncoming traffic and possibly causing a church van full of 14 pregnant nuns to go off road to avoid you and probably roll over. The other option is for your car to swerve to the right, off a cliff, involving no others, yet killing you. Which decision do you think the collective traffic AI will choose. Will the motoring public be happy with giving up that decision making duty to the collective?
If you were responsible for impregnating the nuns, then going off the cliff might be the correct solution...Unless they were headed to a Planned Parenthood Clinic...
The question becomes one of two fold:
1) Can the AI make the decision that quickly ?? (That a lot of computing power, given all the other decisions it has to make)
2) What happens when it locks up/BSD's/or needs to be restarted. Cars do not stop on a dime.
-
Obviously there would have to be some pretty serious redundancy involved with the operating controllers. I'd be skeptical of anything less than 3 layers with an unhackable manual over-ride.
It would also have to be connected to the traffic network (which doesn't yet exist) which would allow external control. Which would make it hackable/hijackable as well as allowing the police to disable a suspect vehicle.
Lots of Sci-fi stories had auto drive vehicles with traffic control systems, Heinlein had a decent model in his universe.
It's doable with current or near term tech but I doubt I'll see it universally implemented in my lifetime, and I'm not yet 55 years old.
-
Ben remarked,
As an old guy, I hate the idea, because to me, automobile = independence. but I suspect a generation or two down the road, many young men and women will not care about that independence factor.
Doggy Daddy added in more or less the same vein:
Another consideration that I think will weigh heavily on how quick autonomous vehicles are accepted is whether people will accept not just giving up control to their car's AI, but giving it up to the collective traffic AI.
Exactly. And the use of the word "collective" is very appropriate.
All of us remember how it was when we first got our driver's licenses. Driving a car meant exactly that: Independence.
If I may add to Ben's "old guy" remark, both of us have probably noticed the subtle but very general shift over the decades toward a "collective" attitude in the U.S., especially in urban areas.
And if my AI car uploads to the whole world my routes to my favorite fishin' hole and prairie dog hunting grounds, I wlll not be happy about it.
Nor will I be happy if it uploads the routes to my GF's or mistress's houses, or my favorite rubdown parlor.
Terry
-
........cutting off oncoming traffic and possibly causing a church van full of 14 pregnant nuns to go off road to avoid you and probably roll over. The other option is for your car to swerve to the right, off a cliff, involving no others, yet killing you. Which decision do you think the collective traffic AI will choose. Will the motoring public be happy with giving up that decision making duty to the collective?
I know when I introduce myself to my car I am going to self identify as 14 pregnant nuns to increase my chance of survival.
-
Ben remarked,
Doggy Daddy added in more or less the same vein:
Exactly. And the use of the word "collective" is very appropriate.
I'm glad you liked that choice of words
I know when I introduce myself to my car I am going to self identify as 14 pregnant nuns to increase my chance of survival.
"Van full of pregnant nuns."
A term I frequently use in new driver class when I need to wake them up/see who's paying attention.
-
If I may add to Ben's "old guy" remark, both of us have probably noticed the subtle but very general shift over the decades toward a "collective" attitude in the U.S., especially in urban areas.
Even more interesting to me, prior to this article, most of the articles I've read talking about rapid implementation as if it were a done deal have been coming from business publications. A couple of articles were talking about the recent Ford shakeup, and both kept mentioning the "individual driver" model as some archaic way of thinking that Ford had better wake up and get away from. I was somewhat surprised to see a capitalistic-oriented pull away from the individual driver and the individually owned vehicle.
-
So as someone who rides bicycles and motorcycles on the public roads, I wonder how I and my colleagues will fit into the grand scheme of autonomous vehicles.
-
So as someone who rides bicycles and motorcycles on the public roads, I wonder how I and my colleagues will fit into the grand scheme of autonomous vehicles.
(https://armedpolitesociety.com/proxy.php?request=http%3A%2F%2Fwww.lyndenhuggins.com%2Fhunting%2Finfo%2Ftargets%2Fbullseye.JPG&hash=3a049de40811756b33f4b3b24cc1d7320310f7c3)
=D
-
For as long as I can remember, liberals have wanted to control our travel. Buses, trains, and electronic trolleys or trains give them the ability to dictate when you leave, where you go, and when you arrive. Autonomous cars seem to be a further step in that direction.
A mobile, armed, educated public is dangerous to a totalitarian government.
-
. . . "Van full of pregnant nuns."
A term I frequently use in new driver class when I need to wake them up/see who's paying attention.
I went to Catholic school for grades 1-8. In one of the early grades - still taught by nuns - a classroom topic came up about license plate numbers. #1 was the governor, and there was much discussion about who got the other single digit plates - mayor for #2, cardinal for #3, etc. I think I suggested #5 should go to the cardinal's daughter.
The nun teaching the class was not amused. :O
As for autonomous cars . . . there are LEGIONS of personal injury lawyers simply salivating over the opportunities for personal enrichment at the expense of people & businesses with deep pockets . . .
-
I suspect a generation or two down the road, many young men and women will not care about that independence factor.
It's certainly starting. There are lots of 16 & 17 year olds (and older =|) that are in no hurry to get their license. They walk, bike, take the bus when necessary\possible but for the most part digital interaction just replaces going anywhere in person. It's very foreign to me but being able to drive just isn't the priority it used to be.
Another consideration that I think will weigh heavily on how quick autonomous vehicles are accepted is whether people will accept not just giving up control to their car's AI, but giving it up to the collective traffic AI.
Which decision do you think the collective traffic AI will choose. Will the motoring public be happy with giving up that decision making duty to the collective?
I see this come up all the time and to me it's really not a big deal. People don't currently have the kind of information to make an informed decision in that time - they don't know if it's nuns or puppies or terrorists in that van, and most people wouldn't be prepared to process and act on that information in that split second anyway. We have people eating, texting, shaving, reading, groping each other, and falling asleep behind the wheel right now so if we can have AI take over I can't imagine it's going to be any worse, even if the AI doesn't make the 'better' decision in a truly unavoidable accident.
-
I find it amusing that the people who are circle-jerking about autonomous cars are often also the ones most worried about .gov & corporate interference in their daily lives. The cognitive dissonance is off the charts at times.
-
Hmmmmmm.......I'm holding out for a real version of: [tinfoil]
-
What about changing your mind. how badly will that screw up the traffic AI ??
"Car, let's go to Burger King."
Car: "Yes Dave, We'll go to Burger King."
Car begins to drive to BK. And on the way there you see the sign for Taco Bell.
"Car, let's go to Taco Bell instead."
Car: "I'm afraid I can't do that Dave."
-
As for autonomous cars . . . there are LEGIONS of personal injury lawyers simply salivating over the opportunities for personal enrichment at the expense of people & businesses with deep pockets . . .
This. No longer will being hit by a car entail trying to figure out if the driver and/or their minimum liability insurance can cover your costs; you're going after the car manufacturers, software companies, etc. that can afford to pay for all the pain and suffering you can imagine.
Ability is one thing, but I don't know how realistic it is to think that people will lose interest in the ability or opportunity to control where they're going. Unless he means that people will be afraid to go somewhere unless a machine is showing them how to get there. That would be sad.
(https://armedpolitesociety.com/proxy.php?request=http%3A%2F%2F1.bp.blogspot.com%2F-LjC4Gw-FbKI%2FUj3bNiSVpMI%2FAAAAAAAABkU%2FdNv6XVzMqIA%2Fs1600%2Fwalle7.jpg&hash=c9f98896188525c300a9aac332d16bf6d79bf312)
Consider that to get the max benefit of this tech, cars will be built to work together and to communicate with each other.
And how often do we still lose cell signal, or even just the data signal, on even relatively major roads? Even when it's not completely lost, how well is a control system going to work over that connection that takes 3-4 minutes to load your Facebook feed? An per-car backup that can function without the central computer just means even more liability for the car manufacturer.
Personally, I would like something along the lines of a basic autopilot for long trips; going from here to Weatherford, for example, is 27 miles of 75mph US281, turn right onto I20 and go another 15 miles. Even if driver interaction was necessary for that interchange, you could still automate 42 miles of a 55 mile trip with nothing more than simple collision avoidance, speed control and lane holding.
-
It's certainly starting. There are lots of 16 & 17 year olds (and older =|) that are in no hurry to get their license. They walk, bike, take the bus when necessary\possible but for the most part digital interaction just replaces going anywhere in person. It's very foreign to me but being able to drive just isn't the priority it used to be.
I see this come up all the time and to me it's really not a big deal. People don't currently have the kind of information to make an informed decision in that time - they don't know if it's nuns or puppies or terrorists in that van, and most people wouldn't be prepared to process and act on that information in that split second anyway. We have people eating, texting, shaving, reading, groping each other, and falling asleep behind the wheel right now so if we can have AI take over I can't imagine it's going to be any worse, even if the AI doesn't make the 'better' decision in a truly unavoidable accident.
Kids are not isolated in the home anymore. They can interact with their friends in relative privacy while sitting on the couch even with parents a few feet away. When I was a kid, you either picked up the land-line phone or drove to where they were. If you couldn't do those, you were on your own. That doesn't explain everything, but maybe part of it.
-
1. I can see city folks going to self-driving cars especially people who commute every day for more than 1 hour. Maybe not all of them, but quite a few would like to be able to sleep or do work or something while someone else drives. If you could control when and where you gave over control and took it back, it might not be such a bad thing.
2. For me, I agree with the idea of a fully autonomous cruise control option on long trips. That might be useful. However, I would need it seldom enough that any extra cost couldn't be justified.
-
so if we can have AI take over I can't imagine it's going to be any worse, even if the AI doesn't make the 'better' decision in a truly unavoidable accident.
Certainly in many instances it will be a good to great thing. For instance I hit a snowstorm driving back from Utah last week. As one of those wimpy Californians, I don't have occasion to do a lot of snow and ice driving. The vehicle traction control came on at least once that I'm aware of when I was on the downhill side of the pass where it was snowing. I was glad it did. In that case, I will trust the computer over me. Also a lot of people have eyesight problems in night driving. The AI is probably safer than them at night.
I do wish we would see more talk about transition, versus the binary of "cut out the human, go straight to AI." It seems like it would make sense, for a lot of reasons, to be able to go back and forth between the human and the computer. KD5 had a great example. There's a lot of highways I drive where it's just basically straight ahead at 70-80MPH for HOURS. I'd love to be able to press a button for the computer to handle that for me and keep me from getting both tired and zoned out.
-
1. I can see city folks going to self-driving cars especially people who commute every day for more than 1 hour. Maybe not all of them, but quite a few would like to be able to sleep or do work or something while someone else drives. If you could control when and where you gave over control and took it back, it might not be such a bad thing.
Most frustrating thing for me is the traffic that creeps forward a car length every 20 seconds or so, and keeps doing that for half an hour or more. If I can, I'll pull over and shut down until it clears. A simple position hold (i.e. hold the lane and an appropriate following distance from the car ahead) with an alert when speed goes over x mph for 10 seconds so I can retake control would cut my stress levels a lot in situations like that. Moving to a designated lane as it becomes feasible would be nice if I can see which one is getting through, or doing that from a short-range control system set up at an accident or construction site would be great, but I'd settle for just not having to sit there and watch somebody's bumper for the next creep-and-stop.
-
So as someone who rides bicycles and motorcycles on the public roads, I wonder how I and my colleagues will fit into the grand scheme of autonomous vehicles.
I wondered about two wheelers, too, in terms of the mechanics. And they smack too much of independence to fit into the collective scheme.
So they'd probably handle it the way they handle any hairy-assed pursuit or hobby. Gradually add restrictions until they were too much of a pain in the ass to own.
-
One of the entrants in the first DARPA challenge was a motorcycle. It can be done, but I agree that the attraction is that bikes are fun to ride, so automating it would have little appeal.
(https://armedpolitesociety.com/proxy.php?request=http%3A%2F%2Fzdnet2.cbsistatic.com%2Fhub%2Fi%2Fr%2F2014%2F08%2F29%2Fd27b2eea-2f58-11e4-9e6a-00505685119a%2Fresize%2F1170x878%2F242ce4f41cd8c58e27b73f3537efa63d%2F22400.jpg&hash=564456ad61f4e31a24ecc16ba6722139f5d1161e)
-
^ Wholly chit, mon !
-
1) where does the rider sit ??
2) what happens if it has to stop (and it will have to at some point) ??
-
1) where does the rider sit ??
2) what happens if it has to stop (and it will have to at some point) ??
(1) They'll miniaturize the rider, of course.
(2) Outriggers will pop out. OR, training wheels will be required.
Or, thinking ahead, maybe they'll have to retrofit gyrostabilizers.
As Ben said, I too, am concerned with the immediacy with which all this is being touted.
Sounds to me like another Lobbyist Boondoggle, where powerful forces are trying to push their products, as usual. Maybe they're rushing it to slip it into us before the more conservative elements in the lawmaking industry can gather their wits.
Such as they are at present. ;/
Terry, 230RN
-
1) where does the rider sit ??
2) what happens if it has to stop (and it will have to at some point) ??
1 The DARPA challenge was to get a driverless vehicle to successfully travel a course of varied terrain. They figured that sort of technology would be useful for military applications since a supply truck takes away a driver from other duties, thus the government interest. So there was never an intention to have a rider.
2 Stopping was an issue. The first year their attempt ended when they tipped over near the beginning. They developed a self righting system, sort of like deployable outriggers. The second year they used those and righted themselves a couple times, but eventually got stuck when they tipped over in an awkward spot. The four wheeled competitors only had to work on "seeing" the terrain instead of worrying about falling over, which ended up being a better strategy.
-
1 The DARPA challenge was to get a driverless vehicle to successfully travel a course of varied terrain. They figured that sort of technology would be useful for military applications since a supply truck takes away a driver from other duties, thus the government interest. So there was never an intention to have a rider.
And if they have to detail 3 more soldiers to maintenance of the driverless vehicle's guidance system?
-
That test site looks familiar. Anyone know where it is?