"Kalanick began courting Levandowski this spring…"
March 13, 2017 12:43 PM Subscribe
Uber trying to get ahead by doing something illegal and underhanded? color me shocked.
posted by Dr. Twist at 12:48 PM on March 13, 2017 [9 favorites]
posted by Dr. Twist at 12:48 PM on March 13, 2017 [9 favorites]
Fascinating. The billions of dollars (hundreds of billions? trillions?) that are going to be sucked out of the manned driving, delivery and long-haul sectors over the next 25 years are definitely going to end up in someone's pockets. I'm not surprised to see shenanigans on steroids (übershenanigans?) in the early fight to determine who gets to control ownership of those pockets.
posted by darkstar at 12:51 PM on March 13, 2017 [12 favorites]
posted by darkstar at 12:51 PM on March 13, 2017 [12 favorites]
I thought corporate espionage was all sexy people in catsuits avoiding lasers, but I suppose I've been mislead again.
posted by GenjiandProust at 12:52 PM on March 13, 2017 [39 favorites]
posted by GenjiandProust at 12:52 PM on March 13, 2017 [39 favorites]
I am a little surprised to see that money -- at this stage, anyway -- going to tech companies, and not to the established hardware companies (i.e., the truck manufacturers).
I mean, the trucks are the platforms, and my expectation is that they will simply modify their existing designs to remove the seat and maybe do some more aerodynamic shapes once the meat is out of the cab. :7)
posted by wenestvedt at 12:53 PM on March 13, 2017
I mean, the trucks are the platforms, and my expectation is that they will simply modify their existing designs to remove the seat and maybe do some more aerodynamic shapes once the meat is out of the cab. :7)
posted by wenestvedt at 12:53 PM on March 13, 2017
Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
posted by gottabefunky at 12:53 PM on March 13, 2017 [53 favorites]
posted by gottabefunky at 12:53 PM on March 13, 2017 [53 favorites]
Uber: getting the meat out of the cab.
posted by blue_beetle at 12:58 PM on March 13, 2017 [6 favorites]
posted by blue_beetle at 12:58 PM on March 13, 2017 [6 favorites]
wenestvedt: we know how to make cheap trucks
we don't know how to make cheap lidar
well, waymo does, but...
posted by hleehowon at 12:58 PM on March 13, 2017 [1 favorite]
we don't know how to make cheap lidar
well, waymo does, but...
posted by hleehowon at 12:58 PM on March 13, 2017 [1 favorite]
Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
Probably. I maintain that truck driving is not a good job - it's dangerous and it's more physically demanding than people think and there's not a lot of value added by having a person do it. The only think it has going for it is that it has a low training requirement. Unlike coal mining I think there will still be truck drivers in a variety of situations, but there will be a lot of long-haul warehouse-to-warehouse automated trucks or truck trains which will reduce the demand for drivers.
posted by GuyZero at 12:59 PM on March 13, 2017 [10 favorites]
Probably. I maintain that truck driving is not a good job - it's dangerous and it's more physically demanding than people think and there's not a lot of value added by having a person do it. The only think it has going for it is that it has a low training requirement. Unlike coal mining I think there will still be truck drivers in a variety of situations, but there will be a lot of long-haul warehouse-to-warehouse automated trucks or truck trains which will reduce the demand for drivers.
posted by GuyZero at 12:59 PM on March 13, 2017 [10 favorites]
Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
We don't employ truckers, but my company does have its own fleet of corporate black cars and employs in-house drivers to do airport runs and the like (we do a lot of travel). Or, rather, we will have this service until October - the company decided to axe the ground transportation service in favor of partnering with Lyft.
It's understandable in a way - the car fleet is aging and it's expensive to maintain them and keep the fleet fresh. However, it also means a bunch of people are losing full- or part-time jobs that all include insurance benefits, paid vacation, and all the other stuff you get working for a reasonably sized, established company. It seems like a small thing on the whole (and really, what company offers this kind of benefit nowadays?) but it's kind of depressing seeing these jobs disappear.
posted by backseatpilot at 1:00 PM on March 13, 2017 [23 favorites]
We don't employ truckers, but my company does have its own fleet of corporate black cars and employs in-house drivers to do airport runs and the like (we do a lot of travel). Or, rather, we will have this service until October - the company decided to axe the ground transportation service in favor of partnering with Lyft.
It's understandable in a way - the car fleet is aging and it's expensive to maintain them and keep the fleet fresh. However, it also means a bunch of people are losing full- or part-time jobs that all include insurance benefits, paid vacation, and all the other stuff you get working for a reasonably sized, established company. It seems like a small thing on the whole (and really, what company offers this kind of benefit nowadays?) but it's kind of depressing seeing these jobs disappear.
posted by backseatpilot at 1:00 PM on March 13, 2017 [23 favorites]
I mean, the trucks are the platforms
Trucks are 100 year-old technology.
LIDAR, detailed centimeter-level road mapping and driving algorithms are not the core competencies of the average truck manufacturer.
posted by GuyZero at 1:00 PM on March 13, 2017 [11 favorites]
Trucks are 100 year-old technology.
LIDAR, detailed centimeter-level road mapping and driving algorithms are not the core competencies of the average truck manufacturer.
posted by GuyZero at 1:00 PM on March 13, 2017 [11 favorites]
the reason why this is important is the comparative cost of everything.
the car, they can do deals with existing car manufacturers. cars can get pretty cheap, if you have the bullies that a multibillion dollar company does for buyers. that's ok because it's multibillion dollar companies on the other side, so they get some unbullyable folks
the lidar currently at the resolution and quality that self-driving cars need, is something like this. randos buying those things buy at $150k or something like that: if you have bulk stuff and good negotiators, can get it down to $75k to $100k.
waymo's design costs $4k, nowadays. they did this by moving heaven and earth.
posted by hleehowon at 1:02 PM on March 13, 2017 [7 favorites]
the car, they can do deals with existing car manufacturers. cars can get pretty cheap, if you have the bullies that a multibillion dollar company does for buyers. that's ok because it's multibillion dollar companies on the other side, so they get some unbullyable folks
the lidar currently at the resolution and quality that self-driving cars need, is something like this. randos buying those things buy at $150k or something like that: if you have bulk stuff and good negotiators, can get it down to $75k to $100k.
waymo's design costs $4k, nowadays. they did this by moving heaven and earth.
posted by hleehowon at 1:02 PM on March 13, 2017 [7 favorites]
Does Anthony Levandowski have any relation to the Third Corey?
posted by ocschwar at 1:05 PM on March 13, 2017 [1 favorite]
posted by ocschwar at 1:05 PM on March 13, 2017 [1 favorite]
Uh oh! It's Otto!
posted by jim in austin at 1:12 PM on March 13, 2017 [1 favorite]
posted by jim in austin at 1:12 PM on March 13, 2017 [1 favorite]
This is waymo brazen than we thought. They otto be ashamed.
posted by miyabo at 1:12 PM on March 13, 2017 [41 favorites]
posted by miyabo at 1:12 PM on March 13, 2017 [41 favorites]
GuyZero: LIDAR, detailed centimeter-level road mapping and driving algorithms are not the core competencies of the average truck manufacturer.
Well, you're right -- but neither were refrigeration, GP, or satellite dishes, and those are bolted onto frames from Peterbilt and Kenworth all over America's highways. That's what I mean about the truck being the platform: the driver is a "feature" that they can just swap out for, say, extra batteries or more coolant or something. The fact of the drivers disappearing doesn't have to impact the truck companies' business.
posted by wenestvedt at 1:15 PM on March 13, 2017 [3 favorites]
Well, you're right -- but neither were refrigeration, GP, or satellite dishes, and those are bolted onto frames from Peterbilt and Kenworth all over America's highways. That's what I mean about the truck being the platform: the driver is a "feature" that they can just swap out for, say, extra batteries or more coolant or something. The fact of the drivers disappearing doesn't have to impact the truck companies' business.
posted by wenestvedt at 1:15 PM on March 13, 2017 [3 favorites]
Slight derail question here, but once all the roads are full of LIDAR-packing driverless vehicles, doesn't that add up to a non-trivial amount of class-1 laser energy being beamed partly at pedestrians' eyeballs?
posted by dowcrag at 1:23 PM on March 13, 2017 [3 favorites]
posted by dowcrag at 1:23 PM on March 13, 2017 [3 favorites]
refrigeration: >100yrs old
GPS: >40
satellite dishes: >40
They're also usually all subcontracted.
Deep learning is actually also a solid 30-40yrs old, but only practical for industry for about 10 years. It still takes a lot of expertise and dark magic, that you basically have to steal PHD's from academia for, for about what a lot of these truck companies pay C-level execs. Same story for the LIDAR.
posted by hleehowon at 1:24 PM on March 13, 2017 [3 favorites]
GPS: >40
satellite dishes: >40
They're also usually all subcontracted.
Deep learning is actually also a solid 30-40yrs old, but only practical for industry for about 10 years. It still takes a lot of expertise and dark magic, that you basically have to steal PHD's from academia for, for about what a lot of these truck companies pay C-level execs. Same story for the LIDAR.
posted by hleehowon at 1:24 PM on March 13, 2017 [3 favorites]
Self-driving trucks played a non-insignificant role in the latest Wolverine movie, IIRC.
posted by gottabefunky at 1:25 PM on March 13, 2017 [4 favorites]
posted by gottabefunky at 1:25 PM on March 13, 2017 [4 favorites]
Slight derail question here, but once all the roads are full of LIDAR-packing driverless vehicles, doesn't that add up to a non-trivial amount of class-1 laser energy being beamed partly at pedestrians' eyeballs
Where we're going, we won't need eyes to drive.
posted by overeducated_alligator at 1:26 PM on March 13, 2017 [12 favorites]
Where we're going, we won't need eyes to drive.
posted by overeducated_alligator at 1:26 PM on March 13, 2017 [12 favorites]
Refrigerator manufacturers were not actively plotting to replace trucking companies using their $100B of cash in the bank as a way to generate billions of dollars a year in new revenue.
Waymo may end up being the next Mopar but its goals are to be the next GM.
posted by GuyZero at 1:26 PM on March 13, 2017 [3 favorites]
Waymo may end up being the next Mopar but its goals are to be the next GM.
posted by GuyZero at 1:26 PM on March 13, 2017 [3 favorites]
Custom software for novel usage is not a commodity. Lots of orgs clueless about software try to implement large systems and piss 100 million, 500 million, 1 billion into the wind and don't have shit to show for it. Lots of boutique shops who can write good software, but there are only a few really huge companies that have a fighting chance at writing working good really really complex softwar for cheap, and those companies still have 100 million, 1 billion boondoggles.
Custom software for AI usage is even less of a commodity. To my knowledge and expertise, there are about 20,000 actual competent working installations of DL systems anywhere in the world, nearly all in the Valley, with radical inequity in impact and size: undoubtedly largest system is Google Search, which is now backed by a bunch of DL shit. Probably second largest is Facebook news feed.
posted by hleehowon at 1:29 PM on March 13, 2017 [2 favorites]
Custom software for AI usage is even less of a commodity. To my knowledge and expertise, there are about 20,000 actual competent working installations of DL systems anywhere in the world, nearly all in the Valley, with radical inequity in impact and size: undoubtedly largest system is Google Search, which is now backed by a bunch of DL shit. Probably second largest is Facebook news feed.
posted by hleehowon at 1:29 PM on March 13, 2017 [2 favorites]
GuyZero: Waymo may end up being the next Mopar but its goals are to be the next GM.
Ooooohhhhhhhhhh, now I get it when you put it that way! In that case: good luck, Peterbilt.
One downside to that scenario is that the Waymo logo is going to look a lot stupider when rendered in chrome than the Mack Truck bulldog.
posted by wenestvedt at 1:36 PM on March 13, 2017 [2 favorites]
Ooooohhhhhhhhhh, now I get it when you put it that way! In that case: good luck, Peterbilt.
One downside to that scenario is that the Waymo logo is going to look a lot stupider when rendered in chrome than the Mack Truck bulldog.
posted by wenestvedt at 1:36 PM on March 13, 2017 [2 favorites]
I would love to eventually find out that the board-house engineer that "accidentally" emailed the Otto board files to Waymo/Google did it because he/she saw what was going on and decided to provide the smoking gun.
Could we nominate them for the MIT Disobedience Award?
"With this award, we honor work that impacts society in positive ways, and is consistent with a set of key principles."
posted by JoeZydeco at 1:44 PM on March 13, 2017 [7 favorites]
Could we nominate them for the MIT Disobedience Award?
"With this award, we honor work that impacts society in positive ways, and is consistent with a set of key principles."
posted by JoeZydeco at 1:44 PM on March 13, 2017 [7 favorites]
> the Waymo logo is going to look a lot stupider when rendered in chrome than the Mack Truck bulldog.
If you squint, you can see that it's the cross section of a spiral (like a spring). It could look quite good as a solid spring in two-tone metal.
As for the story - I have no love for Google and its tentacles, but this lawsuit couldn't happen to a nicer set of people. And the facts of the case look utterly damning at this (early) stage, may least.
posted by RedOrGreen at 1:52 PM on March 13, 2017 [1 favorite]
If you squint, you can see that it's the cross section of a spiral (like a spring). It could look quite good as a solid spring in two-tone metal.
As for the story - I have no love for Google and its tentacles, but this lawsuit couldn't happen to a nicer set of people. And the facts of the case look utterly damning at this (early) stage, may least.
posted by RedOrGreen at 1:52 PM on March 13, 2017 [1 favorite]
Custom software for AI usage is even less of a commodity.
Yes, right now, maybe five years after GPU acceleration brought neural networks back into popularity, AI is not a commodity. And genuinely cutting edge stuff will always be rare. But five years from now will you be able to apt-get deep-magic and feed in some marginal quality training data and get back something reasonable? Probably.
Of course the tech companies all like to posture that they've got unreplicable secret-sauce, but the car manufacturers know they have the advantage of time. Twenty years from now Ford is still going to be around. Can the same be said of any of these autonomous car startups? Will even Google or Facebook be around in twenty?
posted by Pyry at 1:56 PM on March 13, 2017 [5 favorites]
Yes, right now, maybe five years after GPU acceleration brought neural networks back into popularity, AI is not a commodity. And genuinely cutting edge stuff will always be rare. But five years from now will you be able to apt-get deep-magic and feed in some marginal quality training data and get back something reasonable? Probably.
Of course the tech companies all like to posture that they've got unreplicable secret-sauce, but the car manufacturers know they have the advantage of time. Twenty years from now Ford is still going to be around. Can the same be said of any of these autonomous car startups? Will even Google or Facebook be around in twenty?
posted by Pyry at 1:56 PM on March 13, 2017 [5 favorites]
I am a little surprised to see that money -- at this stage, anyway -- going to tech companies
There is actually a massive amount of investment going on at the traditional manufacturers. Kenilworth and Peterbilt are more "systems integrators" than actual engineering firms, but their big suppliers like Here and Bosch are absolutely investing billions of dollars building out their own autonomous navigation teams. John Deere is actually a major robotics powerhouse at this point with 250,000 vehicles in operation. The big old companies just don't get all the press that the tech companies do.
posted by miyabo at 1:57 PM on March 13, 2017 [14 favorites]
There is actually a massive amount of investment going on at the traditional manufacturers. Kenilworth and Peterbilt are more "systems integrators" than actual engineering firms, but their big suppliers like Here and Bosch are absolutely investing billions of dollars building out their own autonomous navigation teams. John Deere is actually a major robotics powerhouse at this point with 250,000 vehicles in operation. The big old companies just don't get all the press that the tech companies do.
posted by miyabo at 1:57 PM on March 13, 2017 [14 favorites]
Slight derail question here, but once all the roads are full of LIDAR-packing driverless vehicles, doesn't that add up to a non-trivial amount of class-1 laser energy being beamed partly at pedestrians' eyeballs
Not really. Eye safety is a requirement for the sort of LIDAR you use in public, and the total energy output of a huge pile of these things is peanuts compared to, say, the sun reflecting off a sea of windshields on a bright day. But yeah, these things are designed to be stared at directly without any safety precautions, so this is not something you need to worry about.
posted by phooky at 2:03 PM on March 13, 2017 [1 favorite]
Not really. Eye safety is a requirement for the sort of LIDAR you use in public, and the total energy output of a huge pile of these things is peanuts compared to, say, the sun reflecting off a sea of windshields on a bright day. But yeah, these things are designed to be stared at directly without any safety precautions, so this is not something you need to worry about.
posted by phooky at 2:03 PM on March 13, 2017 [1 favorite]
Calling it now: Uber will be the first Megacorp to employ for-reals Shadowrunners.
posted by Phobos the Space Potato at 2:04 PM on March 13, 2017 [14 favorites]
posted by Phobos the Space Potato at 2:04 PM on March 13, 2017 [14 favorites]
Slight derail question here, but once all the roads are full of LIDAR-packing driverless vehicles, doesn't that add up to a non-trivial amount of class-1 laser energy being beamed partly at pedestrians' eyeballs
Don't look at the laser with your remaining eye.
posted by Pogo_Fuzzybutt at 2:06 PM on March 13, 2017 [10 favorites]
Don't look at the laser with your remaining eye.
posted by Pogo_Fuzzybutt at 2:06 PM on March 13, 2017 [10 favorites]
I thought corporate espionage was all sexy people in catsuits avoiding lasers, but I suppose I've been mislead again.
You might be thinking of my club, Corporate Espionage. You have to find out how to get in touch with me and I text you the location at the last minute.
I've always had an irrational distrust and dislike of Uber, it's weird and satisfying to see it confirmed recently.
posted by bongo_x at 2:08 PM on March 13, 2017 [3 favorites]
You might be thinking of my club, Corporate Espionage. You have to find out how to get in touch with me and I text you the location at the last minute.
I've always had an irrational distrust and dislike of Uber, it's weird and satisfying to see it confirmed recently.
posted by bongo_x at 2:08 PM on March 13, 2017 [3 favorites]
While Deep Neural Networks are obviously got some strong examples in the search and recommendation market as well as the image/pattern recognition market it's not entirely clear that we are going to solve all of the issues that will be needed to clear the regulatory hurdles necessary to get self-driving trucks and cars fully into the marketspace.
Long haul truck traffic seems like it's relatively easy to solve because highways have way less variables to track (limited entrances, relatively fixed routes, traffic patterns are much easier to monitor, etc). The idea of trying to do automated package delivery in urban areas with self-driving vehicles would be a massive challenge.
And that's before you get into the challenges of rolling out the technology. Millions of Americans are employed as drivers and most of them aren't going to want to have to find new jobs especially when truck driving is a relatively well paying but low skilled profession.
posted by vuron at 2:13 PM on March 13, 2017 [3 favorites]
Long haul truck traffic seems like it's relatively easy to solve because highways have way less variables to track (limited entrances, relatively fixed routes, traffic patterns are much easier to monitor, etc). The idea of trying to do automated package delivery in urban areas with self-driving vehicles would be a massive challenge.
And that's before you get into the challenges of rolling out the technology. Millions of Americans are employed as drivers and most of them aren't going to want to have to find new jobs especially when truck driving is a relatively well paying but low skilled profession.
posted by vuron at 2:13 PM on March 13, 2017 [3 favorites]
GuyZero: "Refrigerator manufacturers were not actively plotting to replace trucking companies using their $100B of cash in the bank as a way to generate billions of dollars a year in new revenue."
Interestingly automakers were big investors in early refrigeration companies. Kelvinator was funded by a Buick executive and later merged with Nash (the car company), Frigidaire was funded by Durant (founder of GM) and a few years later bought by GM. The two companies essentially owned the home refrigeration market until GE introduced the Monitor-Top in the late 20s.
posted by Mitheral at 2:13 PM on March 13, 2017 [6 favorites]
Interestingly automakers were big investors in early refrigeration companies. Kelvinator was funded by a Buick executive and later merged with Nash (the car company), Frigidaire was funded by Durant (founder of GM) and a few years later bought by GM. The two companies essentially owned the home refrigeration market until GE introduced the Monitor-Top in the late 20s.
posted by Mitheral at 2:13 PM on March 13, 2017 [6 favorites]
I have no love for Google and its tentacles, but this lawsuit couldn't happen to a nicer set of people.
When Google is The Good Guys, then The Bad Guys are very, very bad.
But when I first saw the company name "Uber", the Nazi Alarm went off in my head at full volume.
posted by oneswellfoop at 2:16 PM on March 13, 2017 [11 favorites]
When Google is The Good Guys, then The Bad Guys are very, very bad.
But when I first saw the company name "Uber", the Nazi Alarm went off in my head at full volume.
posted by oneswellfoop at 2:16 PM on March 13, 2017 [11 favorites]
I thought corporate espionage was all sexy people in catsuits avoiding lasers, but I suppose I've been mislead again.
Instead it's good old John and Jane Q Public avoiding laser guided automatic trucks, but y'know
This is incredible. It's always amazing to see the incredibly blatant shenanigans people get up to when their greed gets the better of them. Levandowski basically asked the company intranet "how do i access all the key drawings and IP" and then downloaded them all. Did he think the company wasn't keeping logs of all that data?
It would do my bitter old heart some good to see Uber go down the damn toilet. Kalanick is such a massive piece of slime.
posted by Existential Dread at 2:16 PM on March 13, 2017 [6 favorites]
Instead it's good old John and Jane Q Public avoiding laser guided automatic trucks, but y'know
This is incredible. It's always amazing to see the incredibly blatant shenanigans people get up to when their greed gets the better of them. Levandowski basically asked the company intranet "how do i access all the key drawings and IP" and then downloaded them all. Did he think the company wasn't keeping logs of all that data?
It would do my bitter old heart some good to see Uber go down the damn toilet. Kalanick is such a massive piece of slime.
posted by Existential Dread at 2:16 PM on March 13, 2017 [6 favorites]
Here is a Velodyne manual [pdf] with fairly detailed safety numbers, if you're interested. Worst case scenario seems to be about 1/5th of the exposure limit, which suggests that if you got really unlucky and were in the proximity of five carefully synchronized velodynes you might be in trouble.
posted by Pyry at 2:25 PM on March 13, 2017
posted by Pyry at 2:25 PM on March 13, 2017
Don't look at one with binoculars from closer than a foot, though. In general, it's probably a good idea to avoid looking at laser devices with binoculars no matter what the safety calculations say.
posted by Pyry at 2:31 PM on March 13, 2017 [1 favorite]
posted by Pyry at 2:31 PM on March 13, 2017 [1 favorite]
How do those Lidars handle interference from all the other (projected) Lidars?
posted by Mei's lost sandal at 2:34 PM on March 13, 2017
posted by Mei's lost sandal at 2:34 PM on March 13, 2017
if you got really unlucky and were in the proximity of five carefully synchronized velodynes you might be in trouble.
No need to worry, nobody could or would synchronize lasers that carefully unless they were some kind of malevolent networked AI bent on destroying its pitiful human masters
posted by theodolite at 2:35 PM on March 13, 2017 [6 favorites]
No need to worry, nobody could or would synchronize lasers that carefully unless they were some kind of malevolent networked AI bent on destroying its pitiful human masters
posted by theodolite at 2:35 PM on March 13, 2017 [6 favorites]
And that's before you get into the challenges of rolling out the technology. Millions of Americans are employed as drivers and most of them aren't going to want to have to find new jobs especially when truck driving is a relatively well paying but low skilled profession.
Amazon and Walmart will jump on it first thing, and the Republicans want to kill the post office anyway.
their greed gets the better of them. Levandowski basically asked the company intranet "how do i access all the key drawings and IP" and then downloaded them all. Did he think the company wasn't keeping logs of all that data?
It hasn't occurred to me before, but of all the companies to try and pull shenanigans on....I mean, would you not pull up at some point when you're like, "I wonder how I can download something of the private server? I guess I'll just google--hunh."
Kalanivk reminds me of Reacher Gilt. /obscure reference.
posted by Diablevert at 2:40 PM on March 13, 2017 [4 favorites]
Amazon and Walmart will jump on it first thing, and the Republicans want to kill the post office anyway.
their greed gets the better of them. Levandowski basically asked the company intranet "how do i access all the key drawings and IP" and then downloaded them all. Did he think the company wasn't keeping logs of all that data?
It hasn't occurred to me before, but of all the companies to try and pull shenanigans on....I mean, would you not pull up at some point when you're like, "I wonder how I can download something of the private server? I guess I'll just google--hunh."
Kalanivk reminds me of Reacher Gilt. /obscure reference.
posted by Diablevert at 2:40 PM on March 13, 2017 [4 favorites]
It's amazing this was all done with the subtlety of a bodega robbery. If there's any justice we'll see jail terms.
posted by xammerboy at 2:41 PM on March 13, 2017 [1 favorite]
posted by xammerboy at 2:41 PM on March 13, 2017 [1 favorite]
John Deere is actually a major robotics powerhouse at this point with 250,000 vehicles in operation. The big old companies just don't get all the press that the tech companies do.
Deere is probably a great example of where the entire transportation industry is going and it's not all that great. You're basically paying for field-plowing-as-a-service except you have to front all the capital to buy the tractor although you don't actually own it like you probably think you do. This will be trucking in a few decades - there will be a lot fewer operators and the ownership of capital assets will be highly concentrated. I expect that they won't repeat Deere's mistake and even fewer truck drivers will be owner-operators then there are now.
posted by GuyZero at 2:46 PM on March 13, 2017 [11 favorites]
Deere is probably a great example of where the entire transportation industry is going and it's not all that great. You're basically paying for field-plowing-as-a-service except you have to front all the capital to buy the tractor although you don't actually own it like you probably think you do. This will be trucking in a few decades - there will be a lot fewer operators and the ownership of capital assets will be highly concentrated. I expect that they won't repeat Deere's mistake and even fewer truck drivers will be owner-operators then there are now.
posted by GuyZero at 2:46 PM on March 13, 2017 [11 favorites]
Also: Intel in $15 Billion Deal for Self-Driving Tech Firm Mobileye happened today. The stakes are insanely high for some reason that I can't quite understand beyond it being an investment bubble.
posted by GuyZero at 2:49 PM on March 13, 2017 [1 favorite]
posted by GuyZero at 2:49 PM on March 13, 2017 [1 favorite]
Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
More like the canaries. Your job is next.
posted by leotrotsky at 2:50 PM on March 13, 2017 [18 favorites]
More like the canaries. Your job is next.
posted by leotrotsky at 2:50 PM on March 13, 2017 [18 favorites]
gottabefunky: Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
Federal Motor Carrier Safety Administration (FMCSA), the relatively young Federal agency whose primary mission is to prevent commercial motor vehicle-related fatalities and injuries, have their questions. As said by another safety advocate,
The problem is: how do you go from reading a book or napping to reacting to something that the computer system can't handle? How much warning does the driver need? How much are they likely to get?
We'll get to level 5 automation, but again the human is the hurdle, because that hand-off is hell, especially when your drivers don't drive at all ... unless it's an emergency. It's like knowing how to drive a motorcycle, but it's been years since you actually drove one, but you've been there in a side-car the whole time.
posted by filthy light thief at 2:58 PM on March 13, 2017 [8 favorites]
Federal Motor Carrier Safety Administration (FMCSA), the relatively young Federal agency whose primary mission is to prevent commercial motor vehicle-related fatalities and injuries, have their questions. As said by another safety advocate,
“I think this is going to change so many aspects of all of our lives,” said committee member Steve Owings, Atlanta financial advisor and co-founder of Road Safe America. “The difficult part is how we get from where we are now to what it promises ultimately. There are a whole lot of questions and problems from here to there.”Emphasis mine. "Here to there" means moving from zero to 5 in the scale of autonomous vehicles, where zero is fully human-controlled and five is fully autonomous. In between you have different amounts of driver assistance, and the point that causes safety advocates and agencies the most heartburn is level 3 and 4, where human is tasked to take over in certain scenarios.
The problem is: how do you go from reading a book or napping to reacting to something that the computer system can't handle? How much warning does the driver need? How much are they likely to get?
We'll get to level 5 automation, but again the human is the hurdle, because that hand-off is hell, especially when your drivers don't drive at all ... unless it's an emergency. It's like knowing how to drive a motorcycle, but it's been years since you actually drove one, but you've been there in a side-car the whole time.
posted by filthy light thief at 2:58 PM on March 13, 2017 [8 favorites]
GuyZero: The stakes are insanely high for some reason that I can't quite understand beyond it being an investment bubble.
It all depends on what tech they get for that purchase. Is it new AI? Novel sensors? More nimble processing? Self-driving vehicles contain multitudes, which can be utilized for other tasks and markets.
posted by filthy light thief at 3:00 PM on March 13, 2017 [1 favorite]
It all depends on what tech they get for that purchase. Is it new AI? Novel sensors? More nimble processing? Self-driving vehicles contain multitudes, which can be utilized for other tasks and markets.
posted by filthy light thief at 3:00 PM on March 13, 2017 [1 favorite]
This is what happens when an amateur attempts to engage in industrial espionage. He was just a dude who thought he'd get away with it, but without any real knowledge of what countermeasures they might have in place (or how they could forensically tie him to his [alleged] crime).
Real corporate espionage are done by specialists, who work for firms that do this sort of thing, that are manned by ex-intelligence officers. They get away with it. They also don't use the knowledge they steal (for their clients) in such a blatant and obvious way.
Yeah, I hope there will be some jail time, but really this is a white collar version of some meth head who thinks he's going to rob a bank and get away with it.
posted by el io at 3:03 PM on March 13, 2017 [2 favorites]
Real corporate espionage are done by specialists, who work for firms that do this sort of thing, that are manned by ex-intelligence officers. They get away with it. They also don't use the knowledge they steal (for their clients) in such a blatant and obvious way.
Yeah, I hope there will be some jail time, but really this is a white collar version of some meth head who thinks he's going to rob a bank and get away with it.
posted by el io at 3:03 PM on March 13, 2017 [2 favorites]
Mei's lost sandal: How do those Lidars handle interference from all the other (projected) Lidars?
Stack Exchange question: Do multiple LIDAR systems in same area interfere? An answer: One could use lasers with slightly different wave length - just as different channels for Wi-Fi signal.
Also of note: LiDAR is only one of a number of technologies to be present in "smarter" vehicles. There's also V2V, vehicle to vehicle, and V2I, vehicle to infrastructure, both of which augment and further inform vehicles. V2V means cars can be warned of something up ahead on the road, and be informed of approaching vehicles. V2I wil inform vehicles from the public infrastructure, something which is being heavily promoted and coordinated from the National (U.S.) level.
posted by filthy light thief at 3:07 PM on March 13, 2017 [3 favorites]
Stack Exchange question: Do multiple LIDAR systems in same area interfere? An answer: One could use lasers with slightly different wave length - just as different channels for Wi-Fi signal.
Also of note: LiDAR is only one of a number of technologies to be present in "smarter" vehicles. There's also V2V, vehicle to vehicle, and V2I, vehicle to infrastructure, both of which augment and further inform vehicles. V2V means cars can be warned of something up ahead on the road, and be informed of approaching vehicles. V2I wil inform vehicles from the public infrastructure, something which is being heavily promoted and coordinated from the National (U.S.) level.
posted by filthy light thief at 3:07 PM on March 13, 2017 [3 favorites]
I maintain that truck driving is not a good job - it's dangerous and it's more physically demanding than people think and there's not a lot of value added by having a person do it.
I'll say. This troubled, covert agency is responsible for trucking nuclear bombs across America each day:
posted by indubitable at 3:10 PM on March 13, 2017 [6 favorites]
I'll say. This troubled, covert agency is responsible for trucking nuclear bombs across America each day:
The increased workload will hit an agency already struggling with problems of forced overtime, high driver turnover, old trucks and poor worker morale — raising questions about its ability to keep nuclear shipments safe from attack in an era of more sophisticated terrorism."Yeah so not only are you driving this giant vehicle for long stretches on the boring sameness of an interstate highway, but your cargo is a nuclear weapon. Have fun!"
posted by indubitable at 3:10 PM on March 13, 2017 [6 favorites]
Real corporate espionage are done by specialists, who work for firms that do this sort of thing, that are manned by ex-intelligence officers. They get away with it. They also don't use the knowledge they steal (for their clients) in such a blatant and obvious way.
Possibly a derail, but I'd love to hear more about this. What should he have done?
You'd think Uber, of all companies, would be into hiring professional mercenaries to steal tech.
posted by schadenfrau at 3:11 PM on March 13, 2017 [1 favorite]
Possibly a derail, but I'd love to hear more about this. What should he have done?
You'd think Uber, of all companies, would be into hiring professional mercenaries to steal tech.
posted by schadenfrau at 3:11 PM on March 13, 2017 [1 favorite]
So the big money bet is on re-inventing trains via automated trucks that will still fail at the short haul delivery problem of the awkward loading docks in urban centers?
posted by srboisvert at 3:13 PM on March 13, 2017 [4 favorites]
posted by srboisvert at 3:13 PM on March 13, 2017 [4 favorites]
Google has an incredibly sophisticated system for doing IT forensics and their security people are considered to be the very best. This guy never stood a chance.
posted by Foci for Analysis at 3:34 PM on March 13, 2017 [7 favorites]
posted by Foci for Analysis at 3:34 PM on March 13, 2017 [7 favorites]
Long-haul truckers are going to be the coal miners of the 2020s, aren't they?
They already are. My uncle was a long-haul trucker who lost his job a long time ago... hung himself two years ago.
posted by palomar at 3:50 PM on March 13, 2017 [1 favorite]
They already are. My uncle was a long-haul trucker who lost his job a long time ago... hung himself two years ago.
posted by palomar at 3:50 PM on March 13, 2017 [1 favorite]
(wow, sorry to hear that palomar)
Interesting drama, this Waymo/Uber suit.
It's a nifty engineering challenge, but I still don't see self-driving vehicles as something we really really need. I think there are safety, traffic-flow and regulatory issues that will make adoption and refinement much harder than anticipated. We need less private single-driver vehicles on the roads and in our cities, not more.
Can't remember where I saw it, but one of the commercial proponents of autonomous vehicles was just gushing about ALL THE DATA they will be getting from your connected car. This isn't about safety, folks.
Also...unless Basic Personal Income is an entrenched policy when self-driving trucks are rolled out in volume, you will have millions of out-of-work drivers with a grudge and time on their hands.
posted by Artful Codger at 3:54 PM on March 13, 2017 [7 favorites]
Interesting drama, this Waymo/Uber suit.
It's a nifty engineering challenge, but I still don't see self-driving vehicles as something we really really need. I think there are safety, traffic-flow and regulatory issues that will make adoption and refinement much harder than anticipated. We need less private single-driver vehicles on the roads and in our cities, not more.
Can't remember where I saw it, but one of the commercial proponents of autonomous vehicles was just gushing about ALL THE DATA they will be getting from your connected car. This isn't about safety, folks.
Also...unless Basic Personal Income is an entrenched policy when self-driving trucks are rolled out in volume, you will have millions of out-of-work drivers with a grudge and time on their hands.
posted by Artful Codger at 3:54 PM on March 13, 2017 [7 favorites]
Self-driving trucks played a non-insignificant role in the latest Wolverine movie, IIRC.
Came here to mention this myself; the highway scene in Logan was very striking and some thought went into it. The technology was obviously mature enough that the trucks in question didn't have cabs or even tractors at all; they were basically self-propelled trailers capable of carrying shipping containers. There was no recognizable branding. Someone really thought this through; once you don't need the driver or any of the support services to interface him to the platform, you really don't need a cab. If you focus on shipping containers you don't have to worry about matching trailers to tractors to tow them; the thing carrying the shipping container can simply be the truck, with the engine and other hardware slung underneath. This would create enormous scaling economies and probably justify the complete absence of contemporary style trucks on the road by 2029.
posted by Bringer Tom at 4:03 PM on March 13, 2017 [5 favorites]
Came here to mention this myself; the highway scene in Logan was very striking and some thought went into it. The technology was obviously mature enough that the trucks in question didn't have cabs or even tractors at all; they were basically self-propelled trailers capable of carrying shipping containers. There was no recognizable branding. Someone really thought this through; once you don't need the driver or any of the support services to interface him to the platform, you really don't need a cab. If you focus on shipping containers you don't have to worry about matching trailers to tractors to tow them; the thing carrying the shipping container can simply be the truck, with the engine and other hardware slung underneath. This would create enormous scaling economies and probably justify the complete absence of contemporary style trucks on the road by 2029.
posted by Bringer Tom at 4:03 PM on March 13, 2017 [5 favorites]
This is all meaningless drama. If you read back a year or so in metafilter you'll see that People In the Know Because they Work with Computers are quite confident it's going to be at least another 30 years before we have self-driving cars anyway.
posted by lastobelus at 4:06 PM on March 13, 2017 [1 favorite]
posted by lastobelus at 4:06 PM on March 13, 2017 [1 favorite]
I still don't see self-driving vehicles as something we really really need
Given your handle, you're almost certainly old enough to remember that being a ceaseless refrain about personal computers. If not, you are certainly old enough to remember it being a constant refrain about smart phones.
So even if you don't "see" it, I'm mystified that your not seeing it wouldn't seem highly suspect to yourself.
posted by lastobelus at 4:09 PM on March 13, 2017 [4 favorites]
Given your handle, you're almost certainly old enough to remember that being a ceaseless refrain about personal computers. If not, you are certainly old enough to remember it being a constant refrain about smart phones.
So even if you don't "see" it, I'm mystified that your not seeing it wouldn't seem highly suspect to yourself.
posted by lastobelus at 4:09 PM on March 13, 2017 [4 favorites]
I'm old and I honestly can't remember a single person I knew arguing we didn't need PCs or smartphones (well, in the healthy psychological sense, we really shouldn't call it "need," as these are consumer products we're talking about here, not basic human drives; but I can't even remember anybody who didn't want one at least at some point).
posted by saulgoodman at 4:18 PM on March 13, 2017 [1 favorite]
posted by saulgoodman at 4:18 PM on March 13, 2017 [1 favorite]
at least another 30 years before we have self-driving cars anyway.
Self-driving long-haul trucks =/= self-driving cars in the way you envision. The near-term long-haul trucks are not navigating crowded city streets, but highways between distribution centers near to but definitely not within urban areas. Self-driving cars that can navigate constantly changing, crowded areas with pedestrians, pets, cyclists, other cars, and constantly changing traffic management patterns are way higher in terms of sophistication and potential liability than a long-haul truck.
It's worth considering, though, that the investments in railways that the US refused to make over the last 40 years would have gone a long way to making these self-driving trucks not nearly as necessary or attractive. These are basically rail cars adapted for roads, with requirements in advanced sensor systems and computing that would be less necessary if we had more/better railways. Whether or not that's a good thing, I'll leave up to you.
posted by Existential Dread at 4:20 PM on March 13, 2017 [16 favorites]
Self-driving long-haul trucks =/= self-driving cars in the way you envision. The near-term long-haul trucks are not navigating crowded city streets, but highways between distribution centers near to but definitely not within urban areas. Self-driving cars that can navigate constantly changing, crowded areas with pedestrians, pets, cyclists, other cars, and constantly changing traffic management patterns are way higher in terms of sophistication and potential liability than a long-haul truck.
It's worth considering, though, that the investments in railways that the US refused to make over the last 40 years would have gone a long way to making these self-driving trucks not nearly as necessary or attractive. These are basically rail cars adapted for roads, with requirements in advanced sensor systems and computing that would be less necessary if we had more/better railways. Whether or not that's a good thing, I'll leave up to you.
posted by Existential Dread at 4:20 PM on March 13, 2017 [16 favorites]
Artisan use of heavy trucks isn't going to go away and those trucks still need to carry people.
posted by Mitheral at 4:22 PM on March 13, 2017
posted by Mitheral at 4:22 PM on March 13, 2017
Well, horses and buggies never went away, they just attenuated 1000x. Same will hold true for hand driving, whenever self-driving cars come.
Take the extremally bad estimate for that time, tho.
posted by hleehowon at 4:24 PM on March 13, 2017 [1 favorite]
Take the extremally bad estimate for that time, tho.
posted by hleehowon at 4:24 PM on March 13, 2017 [1 favorite]
It's also worth noting that truck drivers are likely to be required between distribution centers and destination, or the so-called "last mile." This will be the case for some time, so you'll see waves of layoffs, rather than all at once.
posted by Existential Dread at 4:24 PM on March 13, 2017 [3 favorites]
posted by Existential Dread at 4:24 PM on March 13, 2017 [3 favorites]
Yeah, I'm expecting that the first usage will be essentially road trains -- a half dozen or so trucks with one driver accompanying them to deal with any AI-breaking issues. Not actually driving except in that case.
posted by tavella at 4:24 PM on March 13, 2017 [3 favorites]
posted by tavella at 4:24 PM on March 13, 2017 [3 favorites]
Convoys are illegal everywhere in Canada for, arguably, the safety of other road users (they are difficult to pass and take a long time to pass others among other things). Obviously legality is a minor stumbling block to money but it'll be interesting to watch the double speak on this issue.
posted by Mitheral at 4:32 PM on March 13, 2017 [4 favorites]
posted by Mitheral at 4:32 PM on March 13, 2017 [4 favorites]
at least another 30 years before we have self-driving cars anyway.
Or, for certain values of we and have, maybe next year sometime?
From the California DMV: Deployment of Autonomous Vehicles for Public Operation
"The department is amending Article 3.7 to include the testing of vehicles that do not require the presence of a driver inside the vehicle and ensure the testing of such vehicles is conducted on California public roads in a safe manner."
posted by Grimgrin at 5:09 PM on March 13, 2017 [3 favorites]
Or, for certain values of we and have, maybe next year sometime?
From the California DMV: Deployment of Autonomous Vehicles for Public Operation
"The department is amending Article 3.7 to include the testing of vehicles that do not require the presence of a driver inside the vehicle and ensure the testing of such vehicles is conducted on California public roads in a safe manner."
posted by Grimgrin at 5:09 PM on March 13, 2017 [3 favorites]
I thought corporate espionage was all sexy people in catsuits avoiding lasers, but I suppose I've been mislead again.
In this case they weren't avoiding the lasers, precisely the opposite.
posted by ill3 at 5:28 PM on March 13, 2017 [4 favorites]
In this case they weren't avoiding the lasers, precisely the opposite.
posted by ill3 at 5:28 PM on March 13, 2017 [4 favorites]
Insurance companies are going to keep drivers employed, because nobody is going to insure a driverless truck with millions of dollars in inventory. They just won't have to be to be driving the entire time, which is really dangerous working conditions with serious public health externalities.
posted by politikitty at 5:44 PM on March 13, 2017 [2 favorites]
posted by politikitty at 5:44 PM on March 13, 2017 [2 favorites]
The hardest thing about reading the Susan J Fowler post on the harassment, bias and exclusion she had to deal with at Uber was knowing that engineers at Uber get to work on some of the most interesting problems in tech, of a kind that you need to have massive bankroll to work on. The company is excluding women through attrition and retaliation from being able to work on those problems because, its not at all clear why, they just want to privilege the ability of male IT managers to engage in casual misogyny?
So now it looks like, ok, the company is funded to work on the sickest problems in tech, so they just lazily try to steal the answers from Google. Great. I hope they lose it all! They don't deserve problems that fun.
(Also it is so incredibly interesting to hear that Google got their sensor costs down to ~$4k. I didn't realize that was possible.)
posted by ProtoStar at 5:45 PM on March 13, 2017 [10 favorites]
So now it looks like, ok, the company is funded to work on the sickest problems in tech, so they just lazily try to steal the answers from Google. Great. I hope they lose it all! They don't deserve problems that fun.
(Also it is so incredibly interesting to hear that Google got their sensor costs down to ~$4k. I didn't realize that was possible.)
posted by ProtoStar at 5:45 PM on March 13, 2017 [10 favorites]
Yeah, I'm expecting that the first usage will be essentially road trains -- a half dozen or so trucks with one driver accompanying them to deal with any AI-breaking issues.
I think the first road trains will be the intermodal freight. There's a lot of containerized freight that gets moved from a port to a rail yard, then from a rail yard to a distribution center. The big advantages to automated trucking is that it's short hops, a predictable route (the same all the time), and predictable cargo. There's a better payoff per mile for high-resolution scanning for the route to create a software railroad. There's predictable scheduling so inspections, maintenance, and fueling the trucks can happen at set intervals so physical labor can be available at both ends to do the hands-on work that is required in trucking.
For the rest of over the road trucking I see automation having a much slower uptake. There is so much hands-on work that truckers do without getting paid for. They get paid cents per mile that they drive, but they don't get paid for every pre-trip inspection, they don't get paid to fuel the truck, to make sure the cargo or fuel isn't stolen, for washouts, weight checks, etc. Most long haul truckers are only earning money if the wheels are turning, their rules let them work 14 hours a day, only 10 of those hours can be driving, so they perform a lot of uncompensated labor. All of these tasks are much harder to automate than driving, so you either pay a labor force to be available at every shipper and receiver, or you pay a driver to do it. That's just for dry van and reefer trucks, for flatbed trucking you have load securement; there are legal requirements for how you secure the load and how often you have to manually check your chains or straps (something like every 3 hours or 150 miles). How do you automate that? And I don't think think self-driving trucks will ever drive tanker, hazmat, oversized or heavy haul; that stuff will kill people/destroy infrastructure so easily that self driving trucks are going to have to be a very mature technology before it's trusted there.
Which bring up the issue of legal liability; even if the AI driver works perfectly, machines breakdown in unpredictable ways at unpredictable times. If failed brakes pass a pre-trip inspection, who is held liable? Now it's the driver, whose life and livelihood also depends on the brakes functioning properly so the incentives for safe operation are aligned. If that liability transferred to a multi-billion robotransport company, safe operation is no longer aligned with the profit motives of industry, there will always be financial incentive to minimize expenses for inspection and maintenance leading either to more death on the road or stronger regulatory powers for the state and federal DOTs (our robotransport companies and their lobbyists will prefer the carnage over regulation).
Anyway, driving jobs will continue to get much crappier and the pay will get much lower before they go away. There will be a small minority who specialize in niche fields who will continue to do okay.
posted by peeedro at 5:54 PM on March 13, 2017 [22 favorites]
I think the first road trains will be the intermodal freight. There's a lot of containerized freight that gets moved from a port to a rail yard, then from a rail yard to a distribution center. The big advantages to automated trucking is that it's short hops, a predictable route (the same all the time), and predictable cargo. There's a better payoff per mile for high-resolution scanning for the route to create a software railroad. There's predictable scheduling so inspections, maintenance, and fueling the trucks can happen at set intervals so physical labor can be available at both ends to do the hands-on work that is required in trucking.
For the rest of over the road trucking I see automation having a much slower uptake. There is so much hands-on work that truckers do without getting paid for. They get paid cents per mile that they drive, but they don't get paid for every pre-trip inspection, they don't get paid to fuel the truck, to make sure the cargo or fuel isn't stolen, for washouts, weight checks, etc. Most long haul truckers are only earning money if the wheels are turning, their rules let them work 14 hours a day, only 10 of those hours can be driving, so they perform a lot of uncompensated labor. All of these tasks are much harder to automate than driving, so you either pay a labor force to be available at every shipper and receiver, or you pay a driver to do it. That's just for dry van and reefer trucks, for flatbed trucking you have load securement; there are legal requirements for how you secure the load and how often you have to manually check your chains or straps (something like every 3 hours or 150 miles). How do you automate that? And I don't think think self-driving trucks will ever drive tanker, hazmat, oversized or heavy haul; that stuff will kill people/destroy infrastructure so easily that self driving trucks are going to have to be a very mature technology before it's trusted there.
Which bring up the issue of legal liability; even if the AI driver works perfectly, machines breakdown in unpredictable ways at unpredictable times. If failed brakes pass a pre-trip inspection, who is held liable? Now it's the driver, whose life and livelihood also depends on the brakes functioning properly so the incentives for safe operation are aligned. If that liability transferred to a multi-billion robotransport company, safe operation is no longer aligned with the profit motives of industry, there will always be financial incentive to minimize expenses for inspection and maintenance leading either to more death on the road or stronger regulatory powers for the state and federal DOTs (our robotransport companies and their lobbyists will prefer the carnage over regulation).
Anyway, driving jobs will continue to get much crappier and the pay will get much lower before they go away. There will be a small minority who specialize in niche fields who will continue to do okay.
posted by peeedro at 5:54 PM on March 13, 2017 [22 favorites]
I think peedro gets a lot of this right. I came up through the construction ready mix industry. I spent the first few years of my career, specifically, doing GPS and routing code for ready mix.
Drivers do a lot more than just being 'meat in a can' and going point A to B. I think autonomous vehicles are super cool, don't get me wrong, and it wows me to no end that there's a LIDAR package out there for $4,000. And yes, this sort of engineering theft, if true, is super gross.
But yeah, drivers are a critical 'service point' in the shipping industry, and I think that this whole thing is a bit of the ol' 'smoke and mirrors show' considering a robot that consistently delivers a solid hamburger has proven illusive so far. There's a lot of huge, incremental steps between 'neat-o' gee-whiz Google Car and 'I want one of those suckers passing me under load at 75mph between here and San Diego.'
posted by mrdaneri at 6:21 PM on March 13, 2017 [6 favorites]
Drivers do a lot more than just being 'meat in a can' and going point A to B. I think autonomous vehicles are super cool, don't get me wrong, and it wows me to no end that there's a LIDAR package out there for $4,000. And yes, this sort of engineering theft, if true, is super gross.
But yeah, drivers are a critical 'service point' in the shipping industry, and I think that this whole thing is a bit of the ol' 'smoke and mirrors show' considering a robot that consistently delivers a solid hamburger has proven illusive so far. There's a lot of huge, incremental steps between 'neat-o' gee-whiz Google Car and 'I want one of those suckers passing me under load at 75mph between here and San Diego.'
posted by mrdaneri at 6:21 PM on March 13, 2017 [6 favorites]
It's a nifty engineering challenge, but I still don't see self-driving vehicles as something we really really need. I think there are safety, traffic-flow and regulatory issues that will make adoption and refinement much harder than anticipated. We need less private single-driver vehicles on the roads and in our cities, not more.
Truly autonomous cars, if feasible, would lead to fewer cars on the road. Why do you have a car? So you can easily travel from point A to point whenever you want to. You do not have a car because you enjoy filling it with gas, taking it to the mechanic, finding parking.
In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done. We're already seeing the precursor to this in cities with the likes of Uber and Zipcar. Cut out the cost of paying the driver, reduce the cost of paying insurance, and these options become even cheaper. Supposedly 1/3 of land in cities is devoted to parking. In an autonomous vehicle world, you need much much less parking because far fewer people own their own cars and the cars that do exist spend much more of their time in use on the road, rather than parked and idle. The existence of the car is the reason LA looks different than New York. Truly autonomous vehicles could reshape our landscape just as profoundly as the invention of the car itself.
posted by Diablevert at 6:27 PM on March 13, 2017 [3 favorites]
Truly autonomous cars, if feasible, would lead to fewer cars on the road. Why do you have a car? So you can easily travel from point A to point whenever you want to. You do not have a car because you enjoy filling it with gas, taking it to the mechanic, finding parking.
In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done. We're already seeing the precursor to this in cities with the likes of Uber and Zipcar. Cut out the cost of paying the driver, reduce the cost of paying insurance, and these options become even cheaper. Supposedly 1/3 of land in cities is devoted to parking. In an autonomous vehicle world, you need much much less parking because far fewer people own their own cars and the cars that do exist spend much more of their time in use on the road, rather than parked and idle. The existence of the car is the reason LA looks different than New York. Truly autonomous vehicles could reshape our landscape just as profoundly as the invention of the car itself.
posted by Diablevert at 6:27 PM on March 13, 2017 [3 favorites]
Lastobelus: Given your handle, you're almost certainly old enough to remember that being a ceaseless refrain about personal computers. If not, you are certainly old enough to remember it being a constant refrain about smart phones.
So even if you don't "see" it, I'm mystified that your not seeing it wouldn't seem highly suspect to yourself.
Step up! Explain exactly how the widespread use of self-driving vehicles is going to Make Things Better.
Me - all I see is doubling-down on an unsustainable level of car culture when we need to be weaning ourselves away from them. This will be life-support for automakers. And yet another platform for data harvesting.
posted by Artful Codger at 6:28 PM on March 13, 2017 [7 favorites]
So even if you don't "see" it, I'm mystified that your not seeing it wouldn't seem highly suspect to yourself.
Step up! Explain exactly how the widespread use of self-driving vehicles is going to Make Things Better.
Me - all I see is doubling-down on an unsustainable level of car culture when we need to be weaning ourselves away from them. This will be life-support for automakers. And yet another platform for data harvesting.
posted by Artful Codger at 6:28 PM on March 13, 2017 [7 favorites]
Diablevert:
Truly autonomous cars, if feasible, would lead to fewer cars on the road....
In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done.
... I don't see that soon, certainly not in my lifetime. Your vision is like something from a Disney "World of Tomorrow" reel. Everyone sharing cars? Riiiight. And cyclists are notable for their complete absence from that urban vision.
You do know that Uber's still losing money hand over fist, right?
Seriously, I dig the technical challenge and all, but proper investment in public transit is the answer to urban traffic problems, not adding smarts to cars.
posted by Artful Codger at 6:43 PM on March 13, 2017 [8 favorites]
Truly autonomous cars, if feasible, would lead to fewer cars on the road....
In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done.
... I don't see that soon, certainly not in my lifetime. Your vision is like something from a Disney "World of Tomorrow" reel. Everyone sharing cars? Riiiight. And cyclists are notable for their complete absence from that urban vision.
You do know that Uber's still losing money hand over fist, right?
Seriously, I dig the technical challenge and all, but proper investment in public transit is the answer to urban traffic problems, not adding smarts to cars.
posted by Artful Codger at 6:43 PM on March 13, 2017 [8 favorites]
Diablevert: "In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done."
A pool of cars that you can pull from on demand is great but I can't see it significantly reducing the number of cars, in the shortish term at least, because everyone wants to get to work/school at the same time (hence traffic jams/grid lock/rush "hour"). The pool would have to own nearly as many cars as the people commuting already do and most of those cars will sit doing nothing most of the day just like now. To reduce the number of cars businesses have to restructure their scheduling to lower the demand peak by spreading the commute over more hours.
It'll also be interesting to see if a pool of cars can support special needs; rental companies now certainly don't in any but the biggest markets. Lots of people I know own a truck they don't need 50 weeks out of the year so they can pull their 5th wheel on vacation; or own a 4wd/awd vehicle for the 2 days year that sort of traction is useful; or own a third recreational car for summer cruising or track days. If the pool can't support those needs people will still own those cars personally and once you've invested the capital in supporting those needs with a personally owned vehicle then the incremental operating costs of using that vehicle to commute is much less of an obvious win for the pool.
posted by Mitheral at 6:46 PM on March 13, 2017 [1 favorite]
A pool of cars that you can pull from on demand is great but I can't see it significantly reducing the number of cars, in the shortish term at least, because everyone wants to get to work/school at the same time (hence traffic jams/grid lock/rush "hour"). The pool would have to own nearly as many cars as the people commuting already do and most of those cars will sit doing nothing most of the day just like now. To reduce the number of cars businesses have to restructure their scheduling to lower the demand peak by spreading the commute over more hours.
It'll also be interesting to see if a pool of cars can support special needs; rental companies now certainly don't in any but the biggest markets. Lots of people I know own a truck they don't need 50 weeks out of the year so they can pull their 5th wheel on vacation; or own a 4wd/awd vehicle for the 2 days year that sort of traction is useful; or own a third recreational car for summer cruising or track days. If the pool can't support those needs people will still own those cars personally and once you've invested the capital in supporting those needs with a personally owned vehicle then the incremental operating costs of using that vehicle to commute is much less of an obvious win for the pool.
posted by Mitheral at 6:46 PM on March 13, 2017 [1 favorite]
i'm with atrios. ain't never gonna work, for any useful definition of 'work'. have you seen the drop-off line at an elementary school?
posted by j_curiouser at 6:49 PM on March 13, 2017 [4 favorites]
posted by j_curiouser at 6:49 PM on March 13, 2017 [4 favorites]
These are basically rail cars adapted for roads, with requirements in advanced sensor systems and computing that would be less necessary if we had more/better railways. Whether or not that's a good thing, I'll leave up to you.
The reason trucking has curb-stomped rail is that it is direct to dock for facilities that aren't located on rail lines. Rail is great for the 3,000 mile part of the trip but if you're not ordering a whole rail car full of stuff or you aren't on the line, you need a truck to get it from the rail terminal to you anyway. Long haul trucking uses more fuel per pound on the overall trip but it's cheaper and faster because it skips the transfer and extra step.
posted by Bringer Tom at 7:19 PM on March 13, 2017 [3 favorites]
The reason trucking has curb-stomped rail is that it is direct to dock for facilities that aren't located on rail lines. Rail is great for the 3,000 mile part of the trip but if you're not ordering a whole rail car full of stuff or you aren't on the line, you need a truck to get it from the rail terminal to you anyway. Long haul trucking uses more fuel per pound on the overall trip but it's cheaper and faster because it skips the transfer and extra step.
posted by Bringer Tom at 7:19 PM on March 13, 2017 [3 favorites]
You do know that Uber's still losing money hand over fist, right
I've only read six of the nine part series a transport finance analyst has written for Naked Capitalism on the subject, true.
Uber is losing money because its current fares don't cover the costs of labor and maintenance. What Uber has proved is that there is a very big group of people who will happily rely on cabs for most of their transport needs, provided the price is cheap enough and the service quick enough. Hell, they'll even share cabs with strangers. If Uber could just ax all its drivers it could actually make a profit on a $3 cab ride and would actually be worth billions.
. If the pool can't support those needs people will still own those cars personally
Depends what happens with insurance, I'd say. Posit that the proponents are right and that the tech can work just like they say it will: fully networked, able to talk to other cars and react far quicker than a human. At that point, doesn't insurance for human-driven vehicles skyrocket? If, say, you're 10X more likely to get in an accident than a robot, isn't liberty mutual going to charge you 10X more than the owner of a robot car? Maybe 20X more, since the robo-car owner is probably a taxi service with a whole fleet of the things? Hell, maybe human drivers get cordoned off into slow lanes or restricted to surface street so they don't impede the flow of the overall auto-network, in such a world. If truly autonomous cars become a thing, it will definitely affect the price of human car ownership in big ways. There's always going to be some people that will want/need to own their own cars --- if you live in the sticks or just love driving itself. But I gotta think that while there may be a bunch of people who don't mind paying another $50 a month to use a truck three weekends a year, how many are there willing to pay $500 a month? Jay Leno ain't going to break a sweat, but there's a price point where even yuppies would balk.
have you seen the drop-off line at an elementary school?
Sure. I've also seen school buses. I'm not really sure what your point is here; even if you think that there's something people treasure about having their kids dropped off from a car instead of a bus, surely nobody's thrilled to have to spend 20 minutes of their own day waiting in that line? Not having to play chauffeur to your kids is a compelling reason to want an AV.
Again --- I'm not saying that this tech is there right now, today. Maybe it never will be; we could end up in some permanent in-between, like with planes, where you have to have a meat sack behind the wheel for emergency and /or liability reasons, and in that case many of the more transformative possibilities will be lost. But companies aren't diving into this for no reason. Changing from cars to AV could be just as impactful as changing from horses to cars in the first place. That only took ~20 years...
posted by Diablevert at 8:30 PM on March 13, 2017 [2 favorites]
I've only read six of the nine part series a transport finance analyst has written for Naked Capitalism on the subject, true.
Uber is losing money because its current fares don't cover the costs of labor and maintenance. What Uber has proved is that there is a very big group of people who will happily rely on cabs for most of their transport needs, provided the price is cheap enough and the service quick enough. Hell, they'll even share cabs with strangers. If Uber could just ax all its drivers it could actually make a profit on a $3 cab ride and would actually be worth billions.
. If the pool can't support those needs people will still own those cars personally
Depends what happens with insurance, I'd say. Posit that the proponents are right and that the tech can work just like they say it will: fully networked, able to talk to other cars and react far quicker than a human. At that point, doesn't insurance for human-driven vehicles skyrocket? If, say, you're 10X more likely to get in an accident than a robot, isn't liberty mutual going to charge you 10X more than the owner of a robot car? Maybe 20X more, since the robo-car owner is probably a taxi service with a whole fleet of the things? Hell, maybe human drivers get cordoned off into slow lanes or restricted to surface street so they don't impede the flow of the overall auto-network, in such a world. If truly autonomous cars become a thing, it will definitely affect the price of human car ownership in big ways. There's always going to be some people that will want/need to own their own cars --- if you live in the sticks or just love driving itself. But I gotta think that while there may be a bunch of people who don't mind paying another $50 a month to use a truck three weekends a year, how many are there willing to pay $500 a month? Jay Leno ain't going to break a sweat, but there's a price point where even yuppies would balk.
have you seen the drop-off line at an elementary school?
Sure. I've also seen school buses. I'm not really sure what your point is here; even if you think that there's something people treasure about having their kids dropped off from a car instead of a bus, surely nobody's thrilled to have to spend 20 minutes of their own day waiting in that line? Not having to play chauffeur to your kids is a compelling reason to want an AV.
Again --- I'm not saying that this tech is there right now, today. Maybe it never will be; we could end up in some permanent in-between, like with planes, where you have to have a meat sack behind the wheel for emergency and /or liability reasons, and in that case many of the more transformative possibilities will be lost. But companies aren't diving into this for no reason. Changing from cars to AV could be just as impactful as changing from horses to cars in the first place. That only took ~20 years...
posted by Diablevert at 8:30 PM on March 13, 2017 [2 favorites]
Also how do pool vehicles handle stuff like the need to chain up/down?
Diablevert: "Posit that the proponents are right and that the tech can work just like they say it will: fully networked, able to talk to other cars and react far quicker than a human. At that point, doesn't insurance for human-driven vehicles skyrocket? If, say, you're 10X more likely to get in an accident than a robot, isn't liberty mutual going to charge you 10X more than the owner of a robot car?"
The insurance system isn't punitive. So while autonomous cars may have fewer accidents that just means their insurance would be cheaper; human driver insurance won't get more expensive because the risk is already fully accounted for (IE: they are still having the same number of accidents and incurring the same damages). If anything if autonomous cars become a significant percentage of the cars on the road at any time and they actually end up having lower accident rates human drivers should see cheaper insurance as autonomous cars end up avoiding some accidents that would otherwise be caused by the humans.
posted by Mitheral at 8:44 PM on March 13, 2017 [1 favorite]
Diablevert: "Posit that the proponents are right and that the tech can work just like they say it will: fully networked, able to talk to other cars and react far quicker than a human. At that point, doesn't insurance for human-driven vehicles skyrocket? If, say, you're 10X more likely to get in an accident than a robot, isn't liberty mutual going to charge you 10X more than the owner of a robot car?"
The insurance system isn't punitive. So while autonomous cars may have fewer accidents that just means their insurance would be cheaper; human driver insurance won't get more expensive because the risk is already fully accounted for (IE: they are still having the same number of accidents and incurring the same damages). If anything if autonomous cars become a significant percentage of the cars on the road at any time and they actually end up having lower accident rates human drivers should see cheaper insurance as autonomous cars end up avoiding some accidents that would otherwise be caused by the humans.
posted by Mitheral at 8:44 PM on March 13, 2017 [1 favorite]
The reason trucking has curb-stomped rail is that it is direct to dock for facilities that aren't located on rail lines.
Well, there's also the massive public subsidy of trucking rather than rail. Not that convenience isn't a huge factor, but it's definitely not the only one.
posted by asperity at 8:52 PM on March 13, 2017 [8 favorites]
Well, there's also the massive public subsidy of trucking rather than rail. Not that convenience isn't a huge factor, but it's definitely not the only one.
posted by asperity at 8:52 PM on March 13, 2017 [8 favorites]
In a truly autonomous car world, every car is a taxi. You don't own one; you hail it when needed, go where you want, and dismiss it when you're done.
Okay, so this revolution you predict is a taxi. But we've had taxis for a century, and they haven't meant the end of the personally owned vehicle. You may be thinking they're too expensive, and if the price came down everybody would use them. I'd like to point out that there are people around who have five or ten times as much money as you do, doctors, high end lawyers - and even though they are today living in a world where relative to income the price of a taxi is 10-20% of what you're paying, they mostly still own their own cars, in fact they own more, nicer cars.
Outside of places like Manhattan (so dense and congested and expensive to store a car that it's truly a luxury), most 1%ers drive themselves even though taxis are available and cheap relative to their income. (I will note the super rich 0.01% who have actual chauffeurs and the like, but that's the exception.)
People use cars for many things, including status symbols. People also have stuff in their cars. My aging mother was looking for her last car and part of the criteria was that she had to easily be able to put her golf clubs in the trunk and leave them there all season, so she only has to deal with them at the course. Have you ever been in a parent's car that was totally showroom empty? Or do they all have diaper bags, car seats, toys, DVDs, electronics or whatever else is age appropriate?
I'm sure we will eventually be impoverished out of owning cars and it might even be a good thing, but it's not because people are nothing more than point A to B movers.
posted by Homeboy Trouble at 9:23 PM on March 13, 2017 [6 favorites]
Okay, so this revolution you predict is a taxi. But we've had taxis for a century, and they haven't meant the end of the personally owned vehicle. You may be thinking they're too expensive, and if the price came down everybody would use them. I'd like to point out that there are people around who have five or ten times as much money as you do, doctors, high end lawyers - and even though they are today living in a world where relative to income the price of a taxi is 10-20% of what you're paying, they mostly still own their own cars, in fact they own more, nicer cars.
Outside of places like Manhattan (so dense and congested and expensive to store a car that it's truly a luxury), most 1%ers drive themselves even though taxis are available and cheap relative to their income. (I will note the super rich 0.01% who have actual chauffeurs and the like, but that's the exception.)
People use cars for many things, including status symbols. People also have stuff in their cars. My aging mother was looking for her last car and part of the criteria was that she had to easily be able to put her golf clubs in the trunk and leave them there all season, so she only has to deal with them at the course. Have you ever been in a parent's car that was totally showroom empty? Or do they all have diaper bags, car seats, toys, DVDs, electronics or whatever else is age appropriate?
I'm sure we will eventually be impoverished out of owning cars and it might even be a good thing, but it's not because people are nothing more than point A to B movers.
posted by Homeboy Trouble at 9:23 PM on March 13, 2017 [6 favorites]
At that point, doesn't insurance for human-driven vehicles skyrocket?
Not just that. Once robo-cars are common, what happens to parking? Outside my house, there's a line of parked cars up and down the street. If I use a shared robo-car, why would I want the space in front of my house to be used as storage for some guy's property instead of a nice tree? If they want a place to store their vehicle 21 hours a day, they can very well pay for it themselves.
Likewise, would people still require the train station to spend $50 million in public funds to provide parking for people's non-autonomous cars? Should the hospital? Should the supermarket? Once it's no longer a default expectation, and people start having to pay for the space they use, the cost of driving will go way up.
The insurance system isn't punitive. So while autonomous cars may have fewer accidents that just means their insurance would be cheaper; human driver insurance won't get more expensive because the risk is already fully accounted for
The risk isn't really fully accounted for, though. Required insurance for causing someone's death is as low as $10,000. Property damage $3000. Obviously the actual damages are often orders of magnitude greater. There are also a lot of damages-- the costs of emergency services, of thousands of people's time lost in the resulting traffic jams-- which are currently ignored or externalized or swallowed by the government. There are also huge infrastructure costs-- traffic lights, road widenings-- that are necessary only because of bad drivers, but are paid for by everybody.
We accept that, because we figure everyone needs to drive a car, so we can't really go around requiring people to pay for all the damage they cause, or no one could afford to drive and the entire system would collapse. Once there's a viable alternative, though... If it's the case that "while 75% of people use robo-cars, the 200 children who were killed last year were killed by the other 25%", I suspect that there will be 75%+ public support for increasing those $10,000 minimums.
I can't see it significantly reducing the number of cars, in the shortish term at least, because everyone wants to get to work/school at the same time (hence traffic jams/grid lock/rush "hour")
Well that's exactly why it would reduce the number of cars. Like flights, if you want a seat at a busy time, you're going to pay more. Or you can save by sharing a car. Some people will still be willing to pay for their own car, just as people pay for first class, but most won't.
even though they are today living in a world where relative to income the price of a taxi is 10-20% of what you're paying, they mostly still own their own cars
In many locations, calling a taxi is unreliable and requires a wait of uncertain time, and the experience is often not very pleasant. In many more locations, taxis are simply unavailable at all, unless you hire a driver to be on-call, which is far more expensive than even expensive luxury cars.
posted by alexei at 9:39 PM on March 13, 2017 [2 favorites]
Not just that. Once robo-cars are common, what happens to parking? Outside my house, there's a line of parked cars up and down the street. If I use a shared robo-car, why would I want the space in front of my house to be used as storage for some guy's property instead of a nice tree? If they want a place to store their vehicle 21 hours a day, they can very well pay for it themselves.
Likewise, would people still require the train station to spend $50 million in public funds to provide parking for people's non-autonomous cars? Should the hospital? Should the supermarket? Once it's no longer a default expectation, and people start having to pay for the space they use, the cost of driving will go way up.
The insurance system isn't punitive. So while autonomous cars may have fewer accidents that just means their insurance would be cheaper; human driver insurance won't get more expensive because the risk is already fully accounted for
The risk isn't really fully accounted for, though. Required insurance for causing someone's death is as low as $10,000. Property damage $3000. Obviously the actual damages are often orders of magnitude greater. There are also a lot of damages-- the costs of emergency services, of thousands of people's time lost in the resulting traffic jams-- which are currently ignored or externalized or swallowed by the government. There are also huge infrastructure costs-- traffic lights, road widenings-- that are necessary only because of bad drivers, but are paid for by everybody.
We accept that, because we figure everyone needs to drive a car, so we can't really go around requiring people to pay for all the damage they cause, or no one could afford to drive and the entire system would collapse. Once there's a viable alternative, though... If it's the case that "while 75% of people use robo-cars, the 200 children who were killed last year were killed by the other 25%", I suspect that there will be 75%+ public support for increasing those $10,000 minimums.
I can't see it significantly reducing the number of cars, in the shortish term at least, because everyone wants to get to work/school at the same time (hence traffic jams/grid lock/rush "hour")
Well that's exactly why it would reduce the number of cars. Like flights, if you want a seat at a busy time, you're going to pay more. Or you can save by sharing a car. Some people will still be willing to pay for their own car, just as people pay for first class, but most won't.
even though they are today living in a world where relative to income the price of a taxi is 10-20% of what you're paying, they mostly still own their own cars
In many locations, calling a taxi is unreliable and requires a wait of uncertain time, and the experience is often not very pleasant. In many more locations, taxis are simply unavailable at all, unless you hire a driver to be on-call, which is far more expensive than even expensive luxury cars.
posted by alexei at 9:39 PM on March 13, 2017 [2 favorites]
Can't remember where I saw it, but one of the commercial proponents of autonomous vehicles was just gushing about ALL THE DATA they will be getting from your connected car.
Then Otto would be dumb to start by focusing on the trucking industry. In trucking they are already collecting all the data they can, all the big companies use Omnitracs or PeopleNet to manage their fleet. They are hoovering up their own data and other companies are already established in that space to make the most of it. I doubt the data they trucking industry generates is nearly as valuable as the juicy consumer information that can be had from autonomous cars.
Then the other big win for autonomous vehicles, the 100% utilization goal, isn't such a big deal for the trucking industry. The trucks now are probably at at least 50% utilization, there is efficiency to be gained but it's nothing like a private car that sits idle for 23 hours a day. Some companies use team driving, where there are two driver assigned per truck so (I don't know why this isn't so common but I assume it has to do with two people living in a space smaller than a prison cell). Anyhow, unlike Uber, the trucking industry is profitable today, no moonshot into complete automation is required for their continued existence.
With these factors plus the more onerous regulation environment for trucking, it seems to me that autonomous cars make a whole lot more sense than autonomous heavy commercial vehicles. After thinking about it, I'm onboard with the idea that Otto was just a head fake and a legal fig leaf so Uber could launder stolen technology.
posted by peeedro at 10:02 PM on March 13, 2017
Then Otto would be dumb to start by focusing on the trucking industry. In trucking they are already collecting all the data they can, all the big companies use Omnitracs or PeopleNet to manage their fleet. They are hoovering up their own data and other companies are already established in that space to make the most of it. I doubt the data they trucking industry generates is nearly as valuable as the juicy consumer information that can be had from autonomous cars.
Then the other big win for autonomous vehicles, the 100% utilization goal, isn't such a big deal for the trucking industry. The trucks now are probably at at least 50% utilization, there is efficiency to be gained but it's nothing like a private car that sits idle for 23 hours a day. Some companies use team driving, where there are two driver assigned per truck so (I don't know why this isn't so common but I assume it has to do with two people living in a space smaller than a prison cell). Anyhow, unlike Uber, the trucking industry is profitable today, no moonshot into complete automation is required for their continued existence.
With these factors plus the more onerous regulation environment for trucking, it seems to me that autonomous cars make a whole lot more sense than autonomous heavy commercial vehicles. After thinking about it, I'm onboard with the idea that Otto was just a head fake and a legal fig leaf so Uber could launder stolen technology.
posted by peeedro at 10:02 PM on March 13, 2017
there are legal requirements for how you secure the load and how often you have to manually check your chains or straps (something like every 3 hours or 150 miles). How do you automate that?
Doesn't seem too hard. Automated trucks will pull into a bay in a truck stop, where an employee will fuel the truck, run down the inspection checklist, wash it out and weigh it as necessary, sign off on it, and send it on its way. The truck stop company will bill the trucking company. Makes a lot more sense to pay someone for the ten minutes of work you need him to do than to pay someone for hours so that they will also do those ten minutes of work.
Which bring up the issue of legal liability; even if the AI driver works perfectly, machines breakdown in unpredictable ways at unpredictable times. If failed brakes pass a pre-trip inspection, who is held liable?
The trucking company and/or the inspector, presumably. The profit motive is the same. When you depend on the driver's profit motive, you still have the same alignment problem. Drivers certainly have the incentive to drive dangerously and skimp on safety. After all, that's why we need so much regulation regarding how many hours they're allowed to work, how frequently they need to perform safety checks, and so on.
Anyway, driving jobs will continue to get much crappier and the pay will get much lower before they go away. There will be a small minority who specialize in niche fields who will continue to do okay.
Well, driving jobs are already pretty crappy, aren't they? Heck, that's a big part of the reason they pay well. If the jobs were done by inspectors in fixed locations and workers in warehouses, at least they'd be able to have consistent schedules and go home to their families every day.
posted by alexei at 10:03 PM on March 13, 2017 [3 favorites]
Doesn't seem too hard. Automated trucks will pull into a bay in a truck stop, where an employee will fuel the truck, run down the inspection checklist, wash it out and weigh it as necessary, sign off on it, and send it on its way. The truck stop company will bill the trucking company. Makes a lot more sense to pay someone for the ten minutes of work you need him to do than to pay someone for hours so that they will also do those ten minutes of work.
Which bring up the issue of legal liability; even if the AI driver works perfectly, machines breakdown in unpredictable ways at unpredictable times. If failed brakes pass a pre-trip inspection, who is held liable?
The trucking company and/or the inspector, presumably. The profit motive is the same. When you depend on the driver's profit motive, you still have the same alignment problem. Drivers certainly have the incentive to drive dangerously and skimp on safety. After all, that's why we need so much regulation regarding how many hours they're allowed to work, how frequently they need to perform safety checks, and so on.
Anyway, driving jobs will continue to get much crappier and the pay will get much lower before they go away. There will be a small minority who specialize in niche fields who will continue to do okay.
Well, driving jobs are already pretty crappy, aren't they? Heck, that's a big part of the reason they pay well. If the jobs were done by inspectors in fixed locations and workers in warehouses, at least they'd be able to have consistent schedules and go home to their families every day.
posted by alexei at 10:03 PM on March 13, 2017 [3 favorites]
Automated trucks will pull into a bay in a truck stop
Your first step assumes the construction of new pitstop facilities for automated truck every 150 miles over the hundreds of thousands of highway miles. I hope you have VC money to burn.
posted by peeedro at 10:18 PM on March 13, 2017 [1 favorite]
Your first step assumes the construction of new pitstop facilities for automated truck every 150 miles over the hundreds of thousands of highway miles. I hope you have VC money to burn.
posted by peeedro at 10:18 PM on March 13, 2017 [1 favorite]
Alternately, they could use some of the existing network of about 90 bajillion truck stops in the US.
posted by ROU_Xenophobe at 10:39 PM on March 13, 2017 [8 favorites]
posted by ROU_Xenophobe at 10:39 PM on March 13, 2017 [8 favorites]
I think if you're a public transport advocate you should generally be in favour of vehicle automation. The operating cost of buses is going to drop significantly when you no longer need a driver. No longer having to juggle driver schedules and training requirements will allow more flexibility in how frequently buses run and even the routes they take to easily to cater to changes in demand.
You can also run smaller buses on less popular routes without eating another driver salary.
None of this is especially sexy or likely to wow people in rich cities with commuter rail lines but it could be a godsend for less affluent people in small towns and the suburbs.
posted by zymil at 12:57 AM on March 14, 2017 [1 favorite]
You can also run smaller buses on less popular routes without eating another driver salary.
None of this is especially sexy or likely to wow people in rich cities with commuter rail lines but it could be a godsend for less affluent people in small towns and the suburbs.
posted by zymil at 12:57 AM on March 14, 2017 [1 favorite]
> But five years from now will you be able to apt-get deep-magic and feed in some marginal quality training data and get back something reasonable?
The cool kids are already running docker pull tensorflow/tensorflow and feeding it random stuff. Sometimes it's even blue worthy.
As far as hamburger flipping robots go, there was a big thing in the press about a new one less than a week ago. They posted a video though it doesn't sound like it's actually been installed yet.
"The future is already here — it's just not very evenly distributed", as the saying goes. Cruise-control that's on most cars these days is maybe .2 on the zero to five scale for autonomous driving, but Tesla's auto-drive capabilities are closer to two than zero. The price of Tesla's is hoped to come down soon though.
What I'm hearing though, is that truck stop worker is about to become a whole new profession. Rather than driving the truck from point A to B, and doing a bunch of unpaid stuff in between, get paid to do the unpaid stuff, at a regular location; maybe be a designated receiver at the weigh-station to maneuver the semi-autonomous trucks while they're at the station, as well as filling them up and all the other maintenance stuff.
Still, assuming self-driving vehicles are even sold to the public anytime soon after the technology exists (which is a *huge* assumption), I'd "just" buy an RV-sized thing that you'd hate to have to drive in the city, and, um, not have to drive it. Living in a level-five autonomous RV sounds pretty sweet. Roll out of bed, take a shower and get dressed. Walk from the RV into the office. The RV just circles the city all day until you're ready for it to pick you up. Tell it a bit before you get off work and it just shows up at work to pick you up. If you want to go out for dinner or something later, it's like you live right next door to the restaurant because the RV drives you there.
The convenience can't be beat, and it would probably be cheaper than actually buying a condo in the popular parts of Manhattan. Small practical wrinkle though - if everyone actually does that, the city would slow to a digitally managed crawl for vehicles that are simply running 24/7 while the humans are inside the office.
posted by fragmede at 1:45 AM on March 14, 2017 [1 favorite]
The cool kids are already running docker pull tensorflow/tensorflow and feeding it random stuff. Sometimes it's even blue worthy.
As far as hamburger flipping robots go, there was a big thing in the press about a new one less than a week ago. They posted a video though it doesn't sound like it's actually been installed yet.
"The future is already here — it's just not very evenly distributed", as the saying goes. Cruise-control that's on most cars these days is maybe .2 on the zero to five scale for autonomous driving, but Tesla's auto-drive capabilities are closer to two than zero. The price of Tesla's is hoped to come down soon though.
What I'm hearing though, is that truck stop worker is about to become a whole new profession. Rather than driving the truck from point A to B, and doing a bunch of unpaid stuff in between, get paid to do the unpaid stuff, at a regular location; maybe be a designated receiver at the weigh-station to maneuver the semi-autonomous trucks while they're at the station, as well as filling them up and all the other maintenance stuff.
Still, assuming self-driving vehicles are even sold to the public anytime soon after the technology exists (which is a *huge* assumption), I'd "just" buy an RV-sized thing that you'd hate to have to drive in the city, and, um, not have to drive it. Living in a level-five autonomous RV sounds pretty sweet. Roll out of bed, take a shower and get dressed. Walk from the RV into the office. The RV just circles the city all day until you're ready for it to pick you up. Tell it a bit before you get off work and it just shows up at work to pick you up. If you want to go out for dinner or something later, it's like you live right next door to the restaurant because the RV drives you there.
The convenience can't be beat, and it would probably be cheaper than actually buying a condo in the popular parts of Manhattan. Small practical wrinkle though - if everyone actually does that, the city would slow to a digitally managed crawl for vehicles that are simply running 24/7 while the humans are inside the office.
posted by fragmede at 1:45 AM on March 14, 2017 [1 favorite]
cars can get pretty cheap, if you have the bullies that a multibillion dollar company does for buyers. that's ok because it's multibillion dollar companies on the other side, so they get some unbullyable folks
A random aside: one of the supply chain buyers I knew at Google had a standing offer to help coworkers with car shopping. I never used his services, but I heard that car shopping with him was a hoot.
posted by ryanrs at 2:03 AM on March 14, 2017 [2 favorites]
A random aside: one of the supply chain buyers I knew at Google had a standing offer to help coworkers with car shopping. I never used his services, but I heard that car shopping with him was a hoot.
posted by ryanrs at 2:03 AM on March 14, 2017 [2 favorites]
This is good schadenfreude.überschadenfreude
posted by nnethercote at 2:20 AM on March 14, 2017 [4 favorites]
Small practical wrinkle though - if everyone actually does that, the city would slow to a digitally managed crawl for vehicles that are simply running 24/7 while the humans are inside the office.
Also, aside from the ethical problem of all that energy waste, air quality would be horrible. Electric cars don't solve that. There's no way to eliminate air pollution from disintegrating brakes, tires, and roads in use.
posted by asperity at 5:46 AM on March 14, 2017 [2 favorites]
Also, aside from the ethical problem of all that energy waste, air quality would be horrible. Electric cars don't solve that. There's no way to eliminate air pollution from disintegrating brakes, tires, and roads in use.
posted by asperity at 5:46 AM on March 14, 2017 [2 favorites]
There's something about questioning the INEVITABLE SELF DRIVING FUTURE that gets tech bros so damn salty.
posted by Ferreous at 7:05 AM on March 14, 2017 [11 favorites]
posted by Ferreous at 7:05 AM on March 14, 2017 [11 favorites]
I'd be pissed if I were one of the Carnegie Mellon robotics scientists who sold out to join Uber.
posted by benbenson at 7:54 AM on March 14, 2017 [2 favorites]
posted by benbenson at 7:54 AM on March 14, 2017 [2 favorites]
Existential Dread: It's worth considering, though, that the investments in railways that the US refused to make over the last 40 years would have gone a long way to making these self-driving trucks not nearly as necessary or attractive.
Rail is a weird situation - unlike the public roadways, rails are largely privately owned, with some lease and sharing agreements set up. As such, there are seven major companies who own the rails (mildly interactive map - zoom and pan, no legend or on-click information). Any investments into the rails is an investment into a specific company, with the notion that it could benefit freight-reliant or utilizing companies, but industries have a scale of reliance on rail: many products will move by rail or truck, depending on which has the lower cost, and trucks and trains are in ongoing competition in this way, especially for loads that are not too perishable or time-critical. Then there are heavy, cheap bulk goods, like coal, which is more likely to travel by barge than truck.
Many states have anti-donations laws in place, largely due to the railroads who used to lean on states for financial support to ensure their continued operation in a region, so states passed laws that made it harder to pay for improvements that benefit a single company. And even if these laws weren't in place, state roads and other public infrastructure is eternally underfunded (as the eternally dire infrastructure report card* points out). So it's hard to make the case that railway upgrades should be prioritized over the ever-important interstates, or aging dams and other water infrastructure, or anything else for that matter. Sure, if a rail improvement could reduce some congestion on roadways, that's worth investigation, but even that could be tricky to do with public money.
* ... experts say we should approach this figure skeptically. The ASCE is very good at pointing out engineering deficiencies in our infrastructure — but not so good on whether it's actually beneficial to upgrade. "We need this report to point out problems," says Joshua Schank of the Eno Center on Transportation. "But if you're thinking about policy, you have to think more broadly than that." (Washington Post, 2013)
posted by filthy light thief at 7:55 AM on March 14, 2017 [3 favorites]
Rail is a weird situation - unlike the public roadways, rails are largely privately owned, with some lease and sharing agreements set up. As such, there are seven major companies who own the rails (mildly interactive map - zoom and pan, no legend or on-click information). Any investments into the rails is an investment into a specific company, with the notion that it could benefit freight-reliant or utilizing companies, but industries have a scale of reliance on rail: many products will move by rail or truck, depending on which has the lower cost, and trucks and trains are in ongoing competition in this way, especially for loads that are not too perishable or time-critical. Then there are heavy, cheap bulk goods, like coal, which is more likely to travel by barge than truck.
Many states have anti-donations laws in place, largely due to the railroads who used to lean on states for financial support to ensure their continued operation in a region, so states passed laws that made it harder to pay for improvements that benefit a single company. And even if these laws weren't in place, state roads and other public infrastructure is eternally underfunded (as the eternally dire infrastructure report card* points out). So it's hard to make the case that railway upgrades should be prioritized over the ever-important interstates, or aging dams and other water infrastructure, or anything else for that matter. Sure, if a rail improvement could reduce some congestion on roadways, that's worth investigation, but even that could be tricky to do with public money.
* ... experts say we should approach this figure skeptically. The ASCE is very good at pointing out engineering deficiencies in our infrastructure — but not so good on whether it's actually beneficial to upgrade. "We need this report to point out problems," says Joshua Schank of the Eno Center on Transportation. "But if you're thinking about policy, you have to think more broadly than that." (Washington Post, 2013)
posted by filthy light thief at 7:55 AM on March 14, 2017 [3 favorites]
If there is a giant autonomous car fleet you can depend on the rich being ferried around in gigantic RV like rooms full of luxurious furnishings. It would basically be another room in their mansions and will blot out the sunlight when passing the grimy pleb-cars.
posted by chaz at 8:34 AM on March 14, 2017
posted by chaz at 8:34 AM on March 14, 2017
Alexei: Well, driving jobs are already pretty crappy, aren't they? Heck, that's a big part of the reason they pay well.
... truck driving jobs are mostly crappy AND they don't pay very well, considering the hours and paperwork.
Zymil: I think if you're a public transport advocate you should generally be in favour of vehicle automation. The operating cost of buses is going to drop significantly when you no longer need a driver. No longer having to juggle driver schedules and training requirements will allow more flexibility in how frequently buses run and even the routes they take to easily to cater to changes in demand.
You can also run smaller buses on less popular routes without eating another driver salary.
None of this is especially sexy or likely to wow people in rich cities with commuter rail lines but it could be a godsend for less affluent people in small towns and the suburbs.
I accept that autonomous control IS coming, and it will find application in public and mass transit. It's possibly the best use of autonomous control, since you'd be working with fewer, specialized vehicles and a relatively small area, which could be equipped with the transponders and electronic infrastructure to make autonomous vehicles really effective.
I just object to the fact that just about every other aspect of improving transit infrastructure, traffic congestion and urban mobility is being sidelined while we get all gushy over self-driving.
I do get the immediate appeal. For cities, self-driving vehicles seems like a freebie - they hope to see a reduction in traffic problems, without significant spending on infrastructure.
The people who actually NEED a vehicle - rural, small towns, are going to see little to no benefit from self-driving vehicles.
posted by Artful Codger at 8:55 AM on March 14, 2017 [4 favorites]
... truck driving jobs are mostly crappy AND they don't pay very well, considering the hours and paperwork.
Zymil: I think if you're a public transport advocate you should generally be in favour of vehicle automation. The operating cost of buses is going to drop significantly when you no longer need a driver. No longer having to juggle driver schedules and training requirements will allow more flexibility in how frequently buses run and even the routes they take to easily to cater to changes in demand.
You can also run smaller buses on less popular routes without eating another driver salary.
None of this is especially sexy or likely to wow people in rich cities with commuter rail lines but it could be a godsend for less affluent people in small towns and the suburbs.
I accept that autonomous control IS coming, and it will find application in public and mass transit. It's possibly the best use of autonomous control, since you'd be working with fewer, specialized vehicles and a relatively small area, which could be equipped with the transponders and electronic infrastructure to make autonomous vehicles really effective.
I just object to the fact that just about every other aspect of improving transit infrastructure, traffic congestion and urban mobility is being sidelined while we get all gushy over self-driving.
I do get the immediate appeal. For cities, self-driving vehicles seems like a freebie - they hope to see a reduction in traffic problems, without significant spending on infrastructure.
The people who actually NEED a vehicle - rural, small towns, are going to see little to no benefit from self-driving vehicles.
posted by Artful Codger at 8:55 AM on March 14, 2017 [4 favorites]
think if you're a public transport advocate you should generally be in favour of vehicle automation. The operating cost of buses is going to drop significantly when you no longer need a driver. No longer having to juggle driver schedules and training requirements will allow more flexibility in how frequently buses run and even the routes they take to easily to cater to changes in demand.
If you are a public transit user you know there is no way they can ever replace the bus driver. At least half the people are useless at getting on the bus and paying their fares. Then half the time the fare devices don't work. People run for the bus. People get stuck in doors. People can't open the doors. Some people need help getting on and off. Some people need to be kicked off the bus. Having a driver there is important for passenger safety from other passengers.
No robot will handle this.
posted by srboisvert at 9:53 AM on March 14, 2017 [9 favorites]
If you are a public transit user you know there is no way they can ever replace the bus driver. At least half the people are useless at getting on the bus and paying their fares. Then half the time the fare devices don't work. People run for the bus. People get stuck in doors. People can't open the doors. Some people need help getting on and off. Some people need to be kicked off the bus. Having a driver there is important for passenger safety from other passengers.
No robot will handle this.
posted by srboisvert at 9:53 AM on March 14, 2017 [9 favorites]
The people who actually NEED a vehicle - rural, small towns, are going to see little to no benefit from self-driving vehicles.
What? Rural places have old people and drunk people and little public transportation. The safety and independence gains by self-driving cars will be biggest out there. Plus you can tolerate a longer commute so those communities can tolerate a depression without a brain drain. It will transform affordable housing.
posted by politikitty at 10:00 AM on March 14, 2017 [5 favorites]
What? Rural places have old people and drunk people and little public transportation. The safety and independence gains by self-driving cars will be biggest out there. Plus you can tolerate a longer commute so those communities can tolerate a depression without a brain drain. It will transform affordable housing.
posted by politikitty at 10:00 AM on March 14, 2017 [5 favorites]
If you are a public transit user you know there is no way they can ever replace the bus driver. At least half the people are useless at getting on the bus and paying their fares. Then half the time the fare devices don't work.
Some of this is solvable by the transit agencies. All-door boarding would save a lot of time and confusion (yeah, if the fare devices work -- let's hope for improvements there), and anything that increases transit use generally would increase familiarity with the system, so more people would know what they're doing when they get on. Better and more accessible info about the system up front are also key. Also, better ways of paying fares should save a lot of time counting change.
But yeah -- I really don't want most buses to be unattended for a variety of reasons. Assistance for people with disabilities is the biggest one, and it'd be great if bus attendants were more available to give directions and that sort of thing without the responsibility of driving.
posted by asperity at 10:34 AM on March 14, 2017
Some of this is solvable by the transit agencies. All-door boarding would save a lot of time and confusion (yeah, if the fare devices work -- let's hope for improvements there), and anything that increases transit use generally would increase familiarity with the system, so more people would know what they're doing when they get on. Better and more accessible info about the system up front are also key. Also, better ways of paying fares should save a lot of time counting change.
But yeah -- I really don't want most buses to be unattended for a variety of reasons. Assistance for people with disabilities is the biggest one, and it'd be great if bus attendants were more available to give directions and that sort of thing without the responsibility of driving.
posted by asperity at 10:34 AM on March 14, 2017
Eliminate fares altogether. They contribute only a minor portion of operating costs; are a barrier to entry for poor people; discourage casual ridership; cost money in maintenance of fare collection systems; and (where cash is permitted) consume resources in counting, securing, and depositing the cash. Eliminating fares generally ends up reducing the cost per rider because ridership goes up which increases the demand which should end up increasing service levels.
posted by Mitheral at 10:47 AM on March 14, 2017 [5 favorites]
posted by Mitheral at 10:47 AM on March 14, 2017 [5 favorites]
Australia autonomous bus trial. It's small, it goes slow, and it's completely unattended. I could see this being incredibly valuable for feeding people to larger transit lines, but it won't replace them entirely.
posted by miyabo at 10:52 AM on March 14, 2017 [1 favorite]
posted by miyabo at 10:52 AM on March 14, 2017 [1 favorite]
Truly autonomous cars, if feasible, would lead to fewer cars on the road.So, perhaps car transportation can be made more efficient. That would mean we could provide the same level of service with less resources (cars, parking, etc.). But it *also* means people can use even more of it. When something becomes cheaper and more convenient, we find new ways to use it.
Whether the net result is more of fewer cars isn't clear to me.
posted by floppyroofing at 1:40 PM on March 14, 2017 [1 favorite]
If I could hop in my car and do an all night road trip (while sleeping)... I'd do that practically every weekend and visit every little town or landmark in my part of the country. That would be fun for me, but not so much fun for the earth or for traffic.
posted by miyabo at 1:44 PM on March 14, 2017
posted by miyabo at 1:44 PM on March 14, 2017
politikitty: What? Rural places have old people and drunk people and little public transportation. The safety and independence gains by self-driving cars will be biggest out there. Plus you can tolerate a longer commute so those communities can tolerate a depression without a brain drain. It will transform affordable housing.
Rural places have old unlined roads and uncontrolled intersections and grassy or nonexistent shoulders and unmarked turnoffs, about the most difficult terrain for autonomous vehicles. Will your autonomous farm pickup know which freight door to back up to at the feed store? Does it know where to park to load pressure-treated fenceposts? Can it negotiate washouts and bad ruts? Does it know how to pass a combine or a Mennonite horse and buggy?
What small commmunities need is reliable mass transport (ideally a train), not autonomous vehicles.
posted by Artful Codger at 2:35 PM on March 14, 2017 [1 favorite]
Rural places have old unlined roads and uncontrolled intersections and grassy or nonexistent shoulders and unmarked turnoffs, about the most difficult terrain for autonomous vehicles. Will your autonomous farm pickup know which freight door to back up to at the feed store? Does it know where to park to load pressure-treated fenceposts? Can it negotiate washouts and bad ruts? Does it know how to pass a combine or a Mennonite horse and buggy?
What small commmunities need is reliable mass transport (ideally a train), not autonomous vehicles.
posted by Artful Codger at 2:35 PM on March 14, 2017 [1 favorite]
Trains are quite useless for getting to town to see the doctor. You seem to be ignoring the last mile issue; even if we had a much better system of trains than we do, it does not substitute for door to door service. And no, walking and bicycles won't either, because there are a lot of elderly and handicapped people who can't do either for any distance. Buses and vans are a more realistic option, but poor rural regions rarely have the money to subsidize either sufficiently.
posted by tavella at 2:52 PM on March 14, 2017 [2 favorites]
posted by tavella at 2:52 PM on March 14, 2017 [2 favorites]
In fact, while trains and light rail have many virtues in densely populated areas and for transportation between said areas, they are an absolutely terrible intrarural mass transport option. They are vastly expensive per mile and utterly inflexible.
posted by tavella at 2:55 PM on March 14, 2017 [1 favorite]
posted by tavella at 2:55 PM on March 14, 2017 [1 favorite]
The weight issues associated with freight might be poorly situated for private land and roads. You'll want standardized distribution centers and then deal with last mile issues as we currently do.
But the other issues are ideal for autonomous vehicles. Those are things that people do relatively poorly. They have blind spots and can't respond quickly to changing stimuli. I've personally been in a wreck because I did not accurately realize the other car was speeding, so I couldn't clear a left turn in time. Every time it floods in Texas you've got dozens of cars that get stranded in an underpass because idiots can't tell how deep water is.
Every year thousands of people die in car accidents because our senses are limited. Rural communities go bankrupt, leaving people in inescapable poverty because jobs have moved to greener pastures. People get priced out of their cities because housing has ceased to be affordable. And our current long-haul trucking industry has such razor thin margins that we've created a culture of unsafe working conditions, which affects everyone on the road.
Autonomous cars can help level the playing field that geography is economic destiny. It can transform the lives of elderly or disabled folks. Yes, it has environmental concerns we should discuss and innovate to fix. But the gains are just too immense to ignore.
posted by politikitty at 4:48 PM on March 14, 2017 [1 favorite]
But the other issues are ideal for autonomous vehicles. Those are things that people do relatively poorly. They have blind spots and can't respond quickly to changing stimuli. I've personally been in a wreck because I did not accurately realize the other car was speeding, so I couldn't clear a left turn in time. Every time it floods in Texas you've got dozens of cars that get stranded in an underpass because idiots can't tell how deep water is.
Every year thousands of people die in car accidents because our senses are limited. Rural communities go bankrupt, leaving people in inescapable poverty because jobs have moved to greener pastures. People get priced out of their cities because housing has ceased to be affordable. And our current long-haul trucking industry has such razor thin margins that we've created a culture of unsafe working conditions, which affects everyone on the road.
Autonomous cars can help level the playing field that geography is economic destiny. It can transform the lives of elderly or disabled folks. Yes, it has environmental concerns we should discuss and innovate to fix. But the gains are just too immense to ignore.
posted by politikitty at 4:48 PM on March 14, 2017 [1 favorite]
Truly autonomous cars, if feasible, would lead to fewer cars on the road.
I get very annoyed at the technological determinism espoused about autonomous vehicles.
Autonomous vehicles *may* benefit us, in all kinds of ways.
But without good policy, it will probably be shit. With the kind of minimalist policy compromised by short-term narrow self interest and incumbency, which is what we tend to get most of the time in in most countries these days, it will almost certainly be shit.
Just the commute behavior doubles car volumes, because the car now makes a two-way trip for each direction of the commute, instead of just one. And if everyone shopping downtown has a car circling the block waiting for them, well, that level of congestion will far exceed what's generated by cars circling for parking today. It could pretty well shut down the city. Source.
posted by 8k at 7:39 PM on March 14, 2017 [5 favorites]
I get very annoyed at the technological determinism espoused about autonomous vehicles.
Autonomous vehicles *may* benefit us, in all kinds of ways.
But without good policy, it will probably be shit. With the kind of minimalist policy compromised by short-term narrow self interest and incumbency, which is what we tend to get most of the time in in most countries these days, it will almost certainly be shit.
Just the commute behavior doubles car volumes, because the car now makes a two-way trip for each direction of the commute, instead of just one. And if everyone shopping downtown has a car circling the block waiting for them, well, that level of congestion will far exceed what's generated by cars circling for parking today. It could pretty well shut down the city. Source.
posted by 8k at 7:39 PM on March 14, 2017 [5 favorites]
politikitty: Autonomous cars can help level the playing field that geography is economic destiny. It can transform the lives of elderly or disabled folks. Yes, it has environmental concerns we should discuss and innovate to fix. But the gains are just too immense to ignore.
I have not said that there shouldn't be autonomous cars, or that self-driving technology has no benefits. It's coming, I know this, it has the potential to increase safety, and will help throw more even people out of work because, apparently, we need this. But, FFS, there are too many cars in use now, because, especially in North America we've embraced the personal vehicle whole-hog, while other considerations have been neglected.
We need:
- sensible inter-city mass transit
- sufficient urban public transit
- urban design where people have more rights than cars
THEN - for those who truly need the door to door transport... autonomous cars.
posted by Artful Codger at 8:08 PM on March 14, 2017 [2 favorites]
I have not said that there shouldn't be autonomous cars, or that self-driving technology has no benefits. It's coming, I know this, it has the potential to increase safety, and will help throw more even people out of work because, apparently, we need this. But, FFS, there are too many cars in use now, because, especially in North America we've embraced the personal vehicle whole-hog, while other considerations have been neglected.
We need:
- sensible inter-city mass transit
- sufficient urban public transit
- urban design where people have more rights than cars
THEN - for those who truly need the door to door transport... autonomous cars.
posted by Artful Codger at 8:08 PM on March 14, 2017 [2 favorites]
And BBC reports:
The backroom manoeuvrings could suggest bigger changes at Uber are on the way. Two separate, well-placed sources at the company told the BBC that Mr Kalanick could possibly step down as chief executive soon after the new COO is in place - a move that might reassure investors ahead of a long-anticipated potential initial public offering.posted by octothorpe at 4:53 AM on March 20, 2017 [1 favorite]
A spokesperson for Uber would not comment on the suggestion. However, shortly after this story was published, another source, who also did not want to be named, said there was "zero chance" of Mr Kalanick stepping down when the new COO is announced.
octothorpe: "Uber president quits after six months on the job."
The headline kind of undersells the story. The outgoing president didn't just submit his resignation to the management & board. He left with a public statement reading:
posted by mhum at 10:43 AM on March 21, 2017 [2 favorites]
The headline kind of undersells the story. The outgoing president didn't just submit his resignation to the management & board. He left with a public statement reading:
I joined Uber because of its Mission, and the challenge to build global capabilities that would help the company mature and thrive long-term.From what I understand of corporate America, this kind of statement is the equivalent of the Angela Bassett lighting up and walking away from a burning car in Waiting to Exhale gif.
It is now clear, however, that the beliefs and approach to leadership that have guided my career are inconsistent with what I saw and experienced at Uber, and I can no longer continue as president of the ride sharing business.
There are thousands of amazing people at the company, and I truly wish everyone well.
posted by mhum at 10:43 AM on March 21, 2017 [2 favorites]
« Older Project Under-Horse | Internet of shit indeed... Newer »
This thread has been archived and is closed to new comments
posted by schadenfrau at 12:48 PM on March 13, 2017 [20 favorites]