Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
Given all of the technical challenges for robo taxis, communicating with emergency vehicles would appear to be the easiest problem to fix. I’ve heard of emergency vehicles communicating with traffic lights (to make them go green) in other cities, so similar work has been done. Also, many humans have hit emergency vehicles.
1 robo taxi accident gets more media attention than 100 human driver accidents.
Uvalde was police from many different agencies waiting more than an hour before someone got the balls to confront the shooter. They heard the continuous shooting yet valued their own lives over the lives of kids, totally disgusting.
The “left” wants to take our guns from us so we’ll behave like the slaces THEY are.
Uvalde & Los Vegas PROVE – to me at least – that when seconds count the police are over an hour away. Again, it’s all part and parcel of “left wing” plans for us normals.
DISCLAIMER: While I fully support the 2nd Amendment without restrictions, I personally do not own any firearms. With my personality I would be more dangerous to my loved ones than any bad guys. I know many people who DO own firearms and I feel much safer for it.
The police arrived in Uvalde within minutes.
After an hour over 378 policemen responded*.
*responded = handcuffed parents attempting to rescue their children and stood around while children screamed for help
Exactly this. The hysteria around robotic drivers is absurd given how terrible human drivers are. No one thinks they are a terrible driver, of course. Stats say otherwise. We are impulsive, emotional, easily distracted creatures who decline cognitively under stress and over time. We are bad at judging situations and ourselves. 75% of the articles on this site are schadenfreude on the impulsive judgements of others : )
If robots are safer by 1 life then there should be no argument.
The MASSIVE hypocrisy of our “Shining City on the Hill” seems to me grounds for locking most all of us up at hard and dangerous labor on road gangs, farms, grid work, solar/wind factories, etc, for “Crimes against Humanity”. (At least the top 20-30% economically….maybe even more) till they die.
Inventing shit, then inventing shit to correct (or IMPROVE!!!) that shit, and on and on and on, has generated for many…..the American Dream they now enjoy.
Seneca’s Cliff is close……no doubts whatsoever, with tortuous agonizing climate change at the bottom. By all means, save that one life with a new robot, and win the American Dream for a few more…..”entrepreneurs” ? I think that is the right word for all these leader/heroes in our society….at present.
“We are impulsive, emotional, easily distracted creatures.who decline cognitively under stress and over time”.” VERY GOOD!!! And let me add, paraphrasing Nietzsche, “MANY times worse in groups, eras, and epochs”
We probably won’t see what Evolution on this planet does with us….MANY unfortunate bastards will….but maybe?
Of course most all are “ignorant” of their “bio-sins”….but that’s just another couple words. Few can curse themselves, anyway.
If the “lifestyle” fits, WEAR it! On the African plains we co-operated with ALL and shared. It is hardwired in. We would not have survived without it.
Civilization….where did we make the wrong turns in that big change?
Traffic will move way faster once the normies no longer have a brake pedal.
The biggest issue ( in my opinion) with robotaxi’s and any self driving vehicle is one of liability. Our entire automobile ecosystem exists because of the concept of the ” accident”. How many news reports have you seen with the following verbiage, ” grandma jones lost control of the vehicle and plowed in to the preschool killing a dozen small children.” If the person is not impaired or committing a felony at the time the liability is limited because it is an ” accident”. If the world of human driven automobiles was held to the same standards of safety as an industrial workplace, in which there are no accidents, just failures of workplace safety, it would have ceased to exist years ago. That is why you see the forklift driving across the parking lot of Home Depot with a flagger in front.
But with robo taxi’s there are no accidents, just failures of programing, or equipment. So any incident leading to loss of life or property damage by the self driving vehicle will implicitly be the fault of the company operating it and the company building and maintaining the vehicle.
Unless google and their buddies change the entire world of civil liability, I think these vehicles will have the same risk profile for companies as marketing firearms for toddlers.
And guess whose insurance rates will go up to compensate for the multimillion dollar lawsuits… every other human driver in the US
Insurance premium should be responsiblity of self driving system vendors. It will likely be included in monthly lease/subscription fee of the self driving vehicles.
Wolf, human drivers make all kinds of mistakes, but they’re also accountable to those mistakes. I can also look them in the eyes and know that I’m being seen when crossing the street. A sensor doesn’t quite provide the same sense of assurance.
I don’t disagree with you that machines would be better overall drivers. But if roads should be the exclusive domain of automated traffic because humans suck at it more than a machine, then OK, let’s extend that to everything. Fully automated luxury capitalism, bring it on.
Gig workers revolting is small potatoes compared to the real test: what happens to truck drivers if automated truckers becomes the norm. All this march forward isn’t accounting for or looking after the lives affected by these changes. Just how how the owners of that technology can save on labor costs.
People must be better drivers where you live. Many of the drivers here in Texas are nothing more than criminals. They drive 60 mph on residential streets. They go through stop signs and red lights. Most of them should be in prison but the police do not hand out traffic tickets anymore. I saw a motorcycle cop handling out a ticket and I shook his hand and thanked him. That was about three years ago and I haven’t seen him since. My street at rush hour is like the Daytona 500. Bring on the robots!
Lol, come drive in Miami for a few days, if you and your car last that long. Drivers here make Texas drivers look like little old ladies
I believe the driving habits of both Miami and large Texas cities reflect a Third World mentality which has taken over. A long time ago I drove my truck down to Monterrey, Mexico and the driving was wildly reckless. I was impressed that there weren’t accidents every couple of minutes. This “style” of driving is now prevalent here. Traffic signals, speed limits, the rules of the road mean nothing. It’s every man for himself and God against all. Road rage and gun battles in the street are the new normal. Life here is cheap.
I agree with you folks. Here in Portland drivers have become much worse over the last 10 years. But one of the key skills needed to stay alive on the road is to be able to quickly identify license plates from Texas and Florida and give those drivers a wide berth. Ever since I was a teen the common term for a bad driver was a ” California Driver”, but the visitors and transplants from Texas and Florida take it to a new level.
You can always tell those “safe” drivers from Oregon and Washington. They’re the ones in the left lane going below the speed limit with their left blinker on.
I agree with Miami and south Florida in general. My god. My theory is in Miami especially. You have a very large population of elderly people and a very large percentage of young Knuckleheads. That combination is lethal.
On the west coast on 75 by Naples it’s absolutely insane. And the wrecks you see on these highways are complete interstate closedowns and a car in fifty million pieces. No fender benders on Florida interstates that’s for sure
Escierto – describing Texas traffic with a Werner Herzog movie title is new and different . I’ve driven all over Texas, Kaspar Hauser should have it so good. I’ve driven in Mexico City and all over Europe too. I respect law enforcement, but extending a hand towards an on duty police officer is never something that I would do, any more than I would pet a canine officer. They’re at work
I drive up and down the east coast, and around populated centers drivers tend to be more irrational and impolite. Even while at home, which is rural Downest Maine, people drive like lunatics. We live in a 35 MPH zone and most people drive 20-30 MPH faster! There’s a double yellow line on the road out front and many times each day people pass someone not going fast enough for them. And the sheriff’s deputies are no where to be found. Ever! Even huge trucks wiz by, in a rush to get God knows where. This is all symbolic of a collapsing civilization, IMO.
Looks like Texas issued 540,000 traffic tickets last year and another 1M stops with “warnings”. Do you think you might be exaggerating just a bit?
chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://www.dps.texas.gov/sites/default/files/documents/director_staff/public_information/2022_traffic_stop_data_report.pdf
What would happen without external forces is that the truck drivers would add to the labor pool, which would lower labor costs and allow another technology or product to flourish. Sadly we seem to talk of UBI, which is just a hidden subsidy for robotics manufacturers.
There is a painful transition period though, but it’s inevitable if the technology is viable and better
With UPS drivers reportedly making $170,000 per year, I am going to assume that’s were the automation technology will be focused.
I remember battling online with the SDC* Utopians about 10 years ago. I would point out this is exactly what would happen but they would insist it wouldn’t be the case.
* Self-Driving Car. Of course, it rang with STD. And that pissed a lot of them off further. :)
You cannot stop technology from developing — it has been tried, and it failed every time. But you can try to deal with it effectively.
Agree but the problem is the myth that technology ‘develops’ or ‘progresses’ towards any goal other than profit. There is no over arching goal of benefit to mankind.That’s all PR added later. We need to guide the development of technology not deal with consequences after the fact. This is the awful break things fast ask forgiveness later techbro mindset.
Correct. Of course, developing, testing, and implementing new technology requires the consumption of a tremendous amount of real resources and capital. If the latter is not available, or is being mis-allocated, new technology is DOA…
I guess I’m stupid but I always thought autonomous vehicles would be most useful on long boring 1hr+ drives down interstates.
I imagined a version of an HOV lane for autonomous cars to help long distance commuters on their morning drives by taking the wheel on the highway and more safely/quickly get them to their offramp.
Seems operating an autonomous car in dense complicated San Francisco streets is the hardest and least valuable place for them. But what do I know.
They’ve long had autonomous trains and subways, mining trucks, agricultural vehicles, etc. The hard part was to make this tech work in the traffic chaos of a dense congested city with steep hills. If they could just add some snow to the mix, they’d really have the ideal test bed.
It’s easy to program something to operate in a controlled environment. As a software engineer, I have zero interest in driving a “self-driving” car because I know that in a novel situation I will probably react correctly and the software will not.
All the greatest cities in the world are “15 minute cities”: London, Tokyo, Paris, New York, Montreal, Marrakech, Seoul, Rio, Buenes Aires, San Francisco, Berlin. In fact, up until the invention of the automobile, its how EVERY city was built.
It’s just another word for “dense and walkable”, where you can live without needing a car. It’s too bad the term has become so misconstrued by conspiracy theorists.
Dense, walkable cities are great for business. They emit less carbon (44% of a city’s emissions are from transportation). They’re more fun to live in, to visit. They are the world’s most attractive tourist destinations. They inspire. They’re melting pots for food, culture, ideas, innovations, creativity.
They are the antidote to the soul-less, disconnected suburbs we’ve built over the past 75 years.
Very true. Obviously Boston, New Orleans, and others can be mentioned. In the U.S., cities such as Chicago and Detroit started the transformation— cities with more sprawl. Then the Los Angeles and Houston’s etc. accelerated the deterioration
Very true. Texas cities are uninhabitable disasters with no redeeming qualities whatsoever. Places where you better be packing because eventually you will need it. One might think that Thomas Hobbes was imagining future Texas cities when he wrote “No arts; no letters; no society; and which is worst of all, continual fear, and danger of violent death; and the life of man, solitary, poor, nasty, brutish and short…”
You described Detroit and any city in California perfectly. Why would you single out TX? presumably because your ideology blinds you.
Yes, that’s one perspective. Here’s another.
Dense cities are crowded and stressful. The cost of living is high. The quality of public services tends to be low (in the US, at least). Social cohesion and trust is also low. Housing is cramped and expensive. Parking is a nightmare. Taxes tend to be much higher. Grocery stores tend to be expensive and smaller with fewer options on the shelves. Crime is significantly worse on almost all measures. Safety has always been an issue, but is even worse lately. While the idea of density and walkability usually implies that life is more convenient, in many ways it’s less so, especially if you’re not a young healthy person. Trips to the grocery store, hardware store, dept store, etc, often require careful planning and tradeoffs in other areas. In many cases it’s simply easier to just drive your car than to carry heavy bags for a long walk or use public transit.
Just as importantly, basically all dense cities in the US these days have become political monocultures. If you’re in independent or Republican, you routinely feel like you have no place and have no representation. The only political debate and discussion allowed is left vs far left. Expressing any other thoughts is likely to put you in hot water with neighbors, friends, and even weirdly aggressive strangers on the street.
These are many of the reasons I personally will be moving to the suburbs after 10 years of living in multiple neighborhoods of multiple large, dense cities.
Yes, fanboys of city life love to only speak of the positives, conveniently forgetting all the negatives, and then trash on the suburbs.
But I’m SO ready to live in the suburbs again. I’m ready for space, peace, quiet, trees, kind people, easy living, reasonable cost of living, parking, and perhaps above all, being able to walk around my neighborhood at night without being constantly afraid someone will stab me for my phone or my jacket.
The city would drive me crazy too, and hopefully you’ll find a “quiet, easy living, reasonable COL etc.” suburb somewhere, but doubt that would be here around Boston. I live in an affluent metro-west Boston suburb, a town with a long history of farming, literature, and the American Revolution. Quiet? Sometimes, but often not – there is the daily suburban cacophony of traffic, mowers, blowers, whackers, chippers, chain saws, home construction, and assorted heavy machinery. My neighborhood appears almost pastoral, but I’ve endured 3 major construction projects on abutting property where the noise and earth moving went on for 2-18 months. The longest one was building a new school and playing fields, and tearing down the old school, and the noise nearly drove me insane (a noise-canceling ear device probably would have helped).
Easy living? Only if you think maintaining a McMansion and expanse of chemical lawn is easy, along with spending so much time working to afford the COL and driving the kids to school/activities so they can get a good education and have a prestigious career, that most services need to be hired out and there’s no time for things like growing food or home cooking. And cost of living is high – property taxes, utilities, and services are exorbitant, although we get what we pay for to some extent with excellent schools, quality town services, and it is reasonably safe.
This suburban lifestyle is not really a good fit for me and I long to live a simpler life of hard work close to the land where there’s unspoiled nature and real peace and quiet. Hopefully in the near future.
I have lived in suburbs and a lot of other places, and now retired live in a rural area where my taxes are super low (with correspondingly low services, which I do not mind) friendly neighbors (mostly anyway) low crime (we have permit-less conceal carry, so criminals think twice before trying something stupid) a large garden, orchard, small house that is easy to clean and heat, etc., etc. During my military service in the 1960s, where we were packed together like farm animals and everything required standing in line, my dream was to get away from crowds (cities) and everything that goes with them. I have finally succeeded in doing so and will end up my days here, where I am relaxed and content.
P.S. Please stay in your cities and suburbs.
Lived in both but I yearn for a large basement for arcade/pinball hobby and large garage or workshop for building cool things. Also I don’t want connected walls as I am a noisy with lots of musical toys and hate to inflict pain on neighbors that way.
If 15 minute cities had affordable parking garages on the edges where you could park and bicycle in that would be cool. I have tried the train-into-DC thing but it’s usually just horribly time wasteful versus driving in.
Move to a big city in Japan and you will find that life is easier, better, and without most of the problems of those in the USA.
But wait, glad you can not move as there is a language barrier and visa barrier not to mention a job barrier as well.
Depends, I guess. My wife, who is from Tokyo, loves living in San Francisco and loves all it has to offer — from great hiking, skiing, and kayaking at our fingertips to all the things in the city itself — and she likes her job in Silicon Valley, and doesn’t find the commute (30 min by car, door to door) nearly as arduous as the packed rush-hour trains in Tokyo, and would never move back to Tokyo.
Well ” lots ” of information in that post regarding the differences between Tokyo and SF. (not)
As far as commuting to work that is easily solved in Tokyo by buying a decent condo or house near work that eliminates the need for those “packed rush hour travels”.
But I guess that brings up another problem…..
A decent sized condo in a decent building even with the yen falling against the US dollar is going to cost you a couple of million US.
And of course then there are the monthly fees on a decent place which could run anywhere from reasonable to ridiculous.
Most of those decent buildings have fees that run around the 200, 000 yen a month area.
You bet!
And no crime either!
Just returned from London. Melting pot my rear quarter. 15 minutes? LoL. One wild truck (say a programmable driverless one) from succeeding terrorist attacks. At least these large cities demarc more clearly the wealthy from the servant classes!
Decadence. Decay. Wealth destruction. Violence. More violence.
Yawn.
It’s not technology – true to a point. It’s the people themselves. People who understand how to bend technology to their own purposes.
It’s the “other” stuff. I work for a large city that Wolf may or may not live in. These cars are having lots of incidents at construction sites, well beyond the famous once recently where it went into wet concrete. We’ve have these cars driving through coned off areas weekly. A couple weeks ago one went through the cones and stop just before hitting a chipper….imagine how bad that could go if a ground man was feed trees limbs.
They are frequent topic at safety meetings.
They simply can’t handle anything out of ordinary and AI is not even closed to being able to process that.
Every single problem you describe has been perfected by humans. Humans drive through cones all the time, they drive through wet concrete, they drive into cul-de-sacs and then don’t stop and instead roll the car down the embankment, right here in SF, they hit construction workers and construction equipment, they run over pedestrians and bicyclists, because humans are horrible drivers and can handle neither the ordinary nor “anything out of ordinary.” The only question is how much better than humans do robotaxis have to be for you to accept them. 1,000x better? Or just 10x better?
Listen to the podcast. I gave you a bunch of examples of how bad human drivers are, seen with my eyes at our intersection here.
I listened.
People can be held criminally accountable.
Who goes to jail when a self driving taxi kills a construction worker?
Sending people to jail for things like drunk driving is meant as a deterrent so people don’t do it. But deterrence doesn’t work very well for humans. Robotaxis don’t drive drunk. They’re unlikely to hit a construction worker because they’re not speeding in a construction zone, and they’re not drunk, and they’re not distracted by the kids or the dog, or the smartphone — that’s how shit happens with humans, and robotaxis don’t do that. The deterrent for causing accidents by robotaxis is higher costs of insurance, repairs, etc. and fines for the companies that operate them and that design them. And higher costs IS a functional deterrent for companies.
If it makes you happy, they can give the AV the death penalty.
But I think a fine and suspension of service would be a more effective accountability for the company operating that AV.
Just happened to see the headline way down on the front page of the L.A. Times:
“3 Uber riders killed, driver and fourth passenger hospitalized after South L.A. car crash.”
In addition, a person in the other car was hospitalized — so 3 dead, 3 hospitalized.
“According to local media reports, a northbound white Mercedes sedan entered the intersection at a noticeably high speed when it crashed head-on into a black Honda operating as an Uber rideshare.”
Something like this — which has never happened in a robotaxi and was never caused by a robotaxi — would make global headlines and a HUGE brouhaha if it happened in a robotaxi. With human drivers, it’s just routine, 3 of 40,000 that got killed in a year, and 3 of a couple of million that got severely injured in a year. Who cares?
The Mercedes seemed to have been speeding. Robotaxis don’t speed.
Human drivers are TERRIBLE drivers.
Just read this headline in the SF Chronicle (top left column):
“Two killed, one hurt in apparently wrong-way cash on Highway 101 in S.F.”
But who cares! 2 out of 40,000 killed a year, 1 out of millions injured. It just never ends. I’m so sick of this shit. Human drivers are TERRIBLE drivers.
And if a robotaxis gets stuck in wet concrete, it makes the national news, and is held up as proof that the technology doesn’t work? Come on give me break.
Few weeks ago a buddy of mine showed me a pic of a human driven car which made it through multiple lines of cones and signs, entered a tunnel, and when the concrete ended dropped down and scraped along the tracks until reaching the first subway station. Not very long before that I was coming home from work, while crossing the bridge on the highway I spotted a car on the other side of the concrete barrier in the pedestrian/bike lane, to get there they would have had to enter from side streets across the river, shimmeyed past or sidestepped the posts and onto a bike lane and then driven through a park and then turned onto the bridge before getting stuck on the other side. On our local subreddit we have pictures and video of this kind of human driving weekly and sometimes multiple times per week. People do lots of very stupid stuff and every now and then a road worker gets killed by them.
Im sure these robo taxis are a problem, I’ve no doubt at all, but given the huge sums of money to be made by perfecting the tech I’m also sure there’s a sea of engineers pulling monster hours working on improving it. No matter how difficult, the issues will be worked out and in time I’m confident there will be stats to show it. I say that as someone who loves driving and has no vested interest in self driving tech.
The owner/driver can and will be criminally liable….again…a robot taxi the CEO doesn’t go to jail and the apparently they can just game theory the price of the death and destruction into their business model and the author is ok with that.
You put 1 person in jail and another will replace them as nobody think they are the bad driver or criminally negligent at the time of incident (otherwise “accidents” would never happen)
Financially punish a company or legislate a change and you affect all vehicles.
Your “accountability” model for driving safety is a failure and always has been. Why do you stick to it?
Uber and Lyft drivers are now concerned for their jobs. Where were they on this issue when they took the jobs of the taxi drivers?
I’m not advocating for the loss of jobs. I’m just making a comment.
Yes, no one cared about the taxi drivers. I know an old guy here who bought a SF medallion to drive a taxi himself, and also to let others drive. He couldn’t even sell the medallion anymore, lost his business. No one cared.
The taxi medallions were a monopoly. The monopoly got busted.
Now do the NAR/Realtors.
The concept of having anything other than a human being driving any sort of vehicle is patently stupid and should be prohibited entirely.
“The concept of having a human driving any sort of vehicle is patently stupid and should be prohibited entirely.” So there, fixed it for you, LOL
Humans are NEARLY PERFECT as compared to bogus brainless machines driving vehicles.
LOL. Let me fix that for you: “Robotaxis are NEARLY PERFECT as compared to bogus brainless humans driving vehicles.”
I just posted a couple of examples, from today’s major California papers the L.A. Times and the SF Chronicle, which I just happened to see as I’m browsing through CA news this afternoon. So I’ll repeat them here for you:
Just happened to see the headline way down on the front page of the L.A. Times:
“3 Uber riders killed, driver and fourth passenger hospitalized after South L.A. car crash.”
In addition, a person in the other car was hospitalized — so 3 dead, 3 hospitalized.
“According to local media reports, a northbound white Mercedes sedan entered the intersection at a noticeably high speed when it crashed head-on into a black Honda operating as an Uber rideshare.”
Something like this — which has never happened in a robotaxi and was never caused by a robotaxi — would make global headlines and a HUGE brouhaha if it happened in a robotaxi. With human drivers, it’s just routine, 3 of 40,000 that got killed in a year, and 3 of a couple of million that got severely injured in a year. Who cares?
The Mercedes seemed to have been speeding. Robotaxis don’t speed.
Human drivers are TERRIBLE drivers.
……
Just read this headline in the SF Chronicle (top left column):
“Two killed, one hurt in apparently wrong-way cash on Highway 101 in S.F.”
But who cares! 2 out of 40,000 killed a year, 1 out of millions injured. It just never ends. I’m so sick of this shit. Human drivers are TERRIBLE drivers.
And if a robotaxis gets stuck in wet concrete, it makes the national news, and is held up as proof that the technology doesn’t work? Come on give me break.
let’s see a human drive 150 mph down a narrow track carrying hundreds of passengers and tons of cargo over narrow bridges, tunnels, and curving hills.
Now, quite a while since the elevator operator was replaced by automation. 😉
As far as robotaxis go familiarity will either breed comfort or contempt. One issue is that we are aware of the types of errors that human drivers make and those of us who have managed ro survive into middle age have learned to predict them and act accordingly. Errors made by self driving vehicles will likely be more random and therfore harder to predict and compensate for.
This is patently nonsense. Of the 40,000 people killed by human drivers in a year, there are lots of people who never saw what hit them. It wasn’t their fault that they got killed. Human drivers make random stupid decisions that are totally unpredictable, which is what kills people.
You have merely established that my comment is partially nonsense. I grant you that some people do act like malfunctioning robots and act randomly. But every day I see drivers telegraph their stupid moves and I avoid them. And I also recognize responsible drivers who pose a low threat. This is a sixth sense that I fear will fail me when there are robot drivers on the road. Perhaps I will develop new skills to deal with them.
Whenever I drive I assume that there are people trying to kill me. Soon I will need to assume that there are also murderbots on the roads.
Will adding one murderbot remove one human driver from the road? I’d be ok with that.
This might be indeed be good positioning for Robo taxis. Read the headlines – from what you are saying literally everyday (can be done by AI) – position the danger of human driving vis-a-vis Robo taxis. Facts and Figures – No fake news. Genuine lives lost.
Are robocops next hahaha
I refer you to an article from the May 1 2016 issue of Robotics Law Journal “Unpredictability of Driverless Cars”. While the article may be a bit dated it comports with my own interactions with driverless prototype vehicles being tested in the streets of Austin. The actions of driverless cars are unlike those of human driven cars and thus harder to drive with in harmony. They act like stereotypical bad drivers and I do my best to avoid them.
Jeeeesus, if you have to dig up an article from 7 years ago when there were no AVs on the road at all, to prove your point, LOL, you already lost. You’ve been obviated by events. End of discussion.
I think the big “if” with the robotaxis and robotrucks is “if” the courts will allow then to operate. I’m thinking that government may not want all the unemployed taxi and ridesare drivers. They really won’t want most of the current truck drivers unelployed (see how big the truck driving profession is on the link below).
The robotaxis are ALREADY operating. There is nothing for courts to decide. No one can stop technological innovation. It has been tried, and it failed. When something is working, it’s coming. Now the question is: how do we deal with it.
Look at manufacturing in the US. It’s a huge sector, the US is the second largest manufacturing country by output, and larger than the next three combined, but it’s highly automated, and the number of jobs in the sector has shriveled. But those jobs that are still in the sector have much hire qualifications than those 50 years ago.
I trust the Self Driving features of Tesla or other vehicles in any decent open highway. But in a city environment is very difficult. What is needed is roads gadgets that will help those vehicles drive better. Instead what we got are nitwits doing their best to force every city into being Amsterdam.
City driving was always a challenge. Rush hour driving is actually not the problem, as those drivers do the same thing daily and know their roads and problem areas. 4 things are making thing worse for all city drivers. Part of is the push to make people use mass transit, but all it does is slow down traffic and make travel worse, no one is going to leave their cars to use mass transit.
1- The sillyness of putting bike and bus lanes, speed bumps and other junk to make roads car unfriendly and constantly changing those markings.
2- The bike lanes are now lawless lands of motorized 2 wheeled vehicles with no enforcement of legality to drive or legality of vehicle. The bus lanes have become defacto double parking lanes, so buses use the car lanes.
3-A lot of people are driving impaired, aka stone.
4- A lot of people drive paying attention only to their GPS and not paying attention. GPS has problems – take the GWB the exits for the bottom level are an exact mirror image of the to level of the bridge, but GPS can’t tell you are in the top or bottom until the road deviates to the exit ramps, so always a few cars that are scrambling and moving across 3 lanes.
My qualifications:
-Drive in New York City – Bronx/Manhattan and use the George Washington Bridge – (busiest bridge in the country)
-Old paramedic -Drove an ambulance in NYC in the late 80’s & 90’s before GPS – EVOC Certified (Emergency Vehicle Operator Certification – trained taking the vehicles at high speed in wet closed road and feeling the failure/lost of control points and active recovery)
-Tesla Y owner and user of full self drive in appropiate environment and weather.
The issue with Tesla is it relies too much on camera. Needs more LiDAR point clouds.
Now if we can just robotify those vocations of more mortal consequence — politicians, bankers, C-suite snakes — boy, we’ll have it made.
Do you really think so? Or were you wasting everyone’s time by trying to be sarcastic?
Try with the left hand — I hear it has a slightly different feel.
Probably the wisest thing you have ever posted here.
I wonder if putting an orange cone on their heds will stop them?
There are many things that will technologically work. But the challenges and economics of the real world often foil their widespread adoption.
My favorite is the Pizza Vending Machine. For 25 years now I have had different entrepreneurs come through my shop wanting to buy parts for their latest breakthrough vending machine that will prepare, cook and dispense a pizza from scratch. Many have tried, and there are even you-tube videos of one here or there in action.
But to my knowledge they have not made it in to common use. The basic tech is there, the concepts are proven but the economics, the challenges of cleaning and spoilage always ruin the party.
I think self driving cars are like that. A technology that works for a while, in a few places until its contradictions catch up with it.
U read that humans fail to apply brakes in something like 1/2 of all accidents. If ADVs can communicate with each other the driver will no longer have to guess whether the oncoming vehicle will turn left or not, and if it does, brakes or acceleration will occur. It’s called machine learning and should improve by orders of magnitude every week or so. Does the computing power and lack of latency exist? I thought 5G was supposed to help address the latter…
Automatic vehicle-to-vehicle communication/coordination would eliminate a lot of collisions – like TCAS in aircraft. I’ve seen a number of cars ‘submarined’ under the car in front of them, after rear-ending at 30-40 MPH. Distracted/impaired drivers slam into cars from behind without ever braking. By observation, the driver ‘report card’ in Seattle metro area is also poor since lockdowns lifted.
You don’t even need something so fancy. I just rented a car here in Ireland — a Toyota Yaris Hybrid (my first hybrid — it was pretty cool — and NOTHING like the tiny Yaris of old). As I was on the motorway heading back to Dublin to return it, I had it on on cruise control when I approached a slower car in front of me. I was fully aware of it and was about to slow down when I felt the car do it for me. I didn’t know it had that capability! Just for the heck of it I stayed behind the car, and my car maintained a 4-second gap behind the car in front. After a bit, I changed lanes and the car immediately started to speed up until it reached the speed I had set, all without any input on my end (except for the steering part!). I don’t think I’d want to depend on it, but it was a nice safety feature.
The problem with robotaxis is we’ve jumped the starter pistol in the race for ever-improving technology. There’s no urgent need to rush to put these things in place. Yet, we push for the earliest possible introduction.
I say iron out all the bugs in the universal system, and then introduce it smoothly, at once. A good time might be around the year 2035. It’s a conservative date, and it gives plenty of time for the bugs to be ironed out.
“There’s no urgent need to rush to put these things in place.”
There is urgent need: 40,000 people killed and a couple of million severely injured per year in the US by human drivers. Human drivers are terrible drivers.
The urgent need to reduce those 40,000 traffic deaths and milllions of people injured each year in the USA can be accomplished much easier than by implementing a complete make over of the driving method to an AI system.
The first and simplest fix is to get impaired drivers off the road. About 1/3 of those that died were as a result of drunk drivers or those with a BAC of 0.08. Another 2000 or so were killed with a BAC of 0.05.
A BAC of 0.08 is too high. A BAC of 0.05 is too high It should be the same as In Japan. In Japan it is 0.03.
You should also face higher penalties for DWI/DUI, but the USA being the USA can not even enforce laws regarding shoplifting so that is highly unlikely.
Just fix that problem and almost 43% of the people would be removed from the death toll.
Next is the issue of drugs. Yeah you people in the States are a bunch of dopers when it comes to driving too.
The NHTSA guesses that about 56 per cent of all accidents involving serious injury and crashes involved at least one drug.
And even though it has been hammered into people to wear seat belts a lot of idiots still do not wear them. Again the NHTSA guess that a couple of thousand people would still be alaive if they had been wearing seat belts.
So fix that and with the above two problems fixed you have eliminated almost 50% of the deaths already without doing anything new.
Finally, the roads in the USA are poorly designed and lead to lots of pedestrians being killed. The roads are probably in a poor state of repair too which also leads to many more deaths.
About 15% of the deaths are pedestrians. Maybe money should be speant improving roads and pedestrian safey instead of going hog wild on new technology.
So get drunk and drugged people off the road, get people to wear seat belts, and fix the roads and you have solved over 65% of the problem right away.
It also appears that the USA has a lot of deaths on the roads anyway comapred to other countries which in and of itself shows that it is an outlier and not a *human* problem as other countries have much lower rates.
All this might help a little maybe, but not much. The problem is humans. For example, they already don’t adhere to the existing blood-alcohol limits. So tightening the limits isn’t going to change anything. You could put the death penalty on drivers that kill someone while drunk, and these humans would still drive drunk and kill people. That’s just how humans are. The death penalty has not stopped homicides in the US either. These humans are immune to deterrence. They’re emotional, they like to drink and be distracted and get road rage, etc. And that’s human behavior that kills.
If you make a (so called) “secure” (nothing is truly secure) way of first responders to order these “self driving” vehicles about, this will create wide a surface area for hackers to attack. As there are a plethora of these “first responders” with a huge array of requirements, securing such communications is very difficult. I’m not saying that it’s impossible but recent events shows us that nothing is immune from hacking. I’d hate to try and spec such a comms standard, which would have to be adopted by EVERY EV manufacturer and first responder agency. It also brings up “big brother” issues and thus has broad implications. Of course, if the politicians try to solve this we will end up with a completely dystopian dogs breakfast, as they have well proven that they are technical luddites.
1. Never listened to this podcast. Never read an article either.
2. Too many ME’s running around. I counted two apart from our guy.
3. It seems to me this technology is not perfected yet. DARPA ( who wastes money on stupid stuff like computers in 50s, who also created google) has a competition for autonomous vehicles to drive in real world simulation (not virtual). No vehicle from any company or an individual own it yet.
4. Autonomous driving is some 5th generation technology. We are not ready for it yet or the roads are not or the vehicles are not.
5. AI companies like google are visionaries. Yes, I agree. But releasing this vehicle in wild where no real natural predators exist is a backwards move. Wait and perfect the technology. If its not ready for LA, think about a creative person (with all the time and energy) in rural MO.
6. I love computers but humans are too stupid to live peacefully with technology. Think miami or DC? I can’t.
7. Sorry for my long absence. I know no one missed me.
I think the robo-taxi’s have a great future in San Francisco. From what I read and see on the internet everyone is leaving the city by the bay. If the you- tubers are to be believed, everyone will be gone soon ( except for Wolf). To keep up appearances google will turn its attention to making human style robots to take up all the space left by the humans moving out. Then eventually they will reach Nirvana with nothing but robots walking around, and robots riding in robs-taxi’s. At that point the whole system will work super good because there will be no pesky humans. It will be kind of the real world equivalent of the Fed printing to feed and economy based on asset appreciation and crypto. 😊
What’s the allure of a self-driving car? Is everyone soused?
Not having human drivers kill 40,000 people a year, no?
I just posted a couple of examples, from today’s major California papers the L.A. Times and the SF Chronicle, which I just happened to see as I’m browsing through CA news this afternoon. So I’ll repeat them here for you:
Just happened to see the headline way down on the front page of the L.A. Times:
“3 Uber riders killed, driver and fourth passenger hospitalized after South L.A. car crash.”
In addition, a person in the other car was hospitalized — so 3 dead, 3 hospitalized.
“According to local media reports, a northbound white Mercedes sedan entered the intersection at a noticeably high speed when it crashed head-on into a black Honda operating as an Uber rideshare.”
Something like this — which has never happened in a robotaxi and was never caused by a robotaxi — would make global headlines and a HUGE brouhaha if it happened in a robotaxi. With human drivers, it’s just routine, 3 of 40,000 that got killed in a year, and 3 of a couple of million that got severely injured in a year. Who cares?
The Mercedes seemed to have been speeding. Robotaxis don’t speed.
Human drivers are TERRIBLE drivers.
……
Just read this headline in the SF Chronicle (top left column):
“Two killed, one hurt in apparently wrong-way cash on Highway 101 in S.F.”
But who cares! 2 out of 40,000 killed a year, 1 out of millions injured. It just never ends. I’m so sick of this shit. Human drivers are TERRIBLE drivers.
And if a robotaxis gets stuck in wet concrete, it makes the national news, and is held up as proof that the technology doesn’t work? Come on give me break.
In addition to being terrible drivers, humans are also expensive drivers. So if you could cut out the human driver, you’d have lower costs of operating taxi enterprises, which should lead to lower fares. Automation has worked that way from get go. Hence the huge automation in manufacturing.
So what do we do now with car enthusiasts, such as myself? Are we told to go pound sand, find another hobby and embrace robot cars?
Drive something that’s fun to drive when it’s fun to drive, say, on a curvy country road or on a dirt track or whatever on Sundays, and let machines do the day-to-day traffic-fighting while you relax in the back. That’s how I see it. I want a chauffer without the chauffer.
It will be fun to use them as rolling pylons or to spook them into slamming on their brakes by doing a little wiggle into their lane. Fun can take many forms, and it would be funny to see the techno fetishists spill their Sugarbucks all over the laptop.
I can’t wait for my robodoc and robonurse. Humans are terrible doctors and nurses with five times as many preventable deaths in hospitals as there are deaths on the road.
The best drivers have zero accidents. That’s the standard robocars need to meet.
cnchal,
YOU are Exhibit A why humans are shitty driver, terrible drivers, reckless drivers, and dangerous drivers. YOU should be banned from driving.
YOUR emotional reaction to other vehicles on the road is part of the reason why 40,000 humans are killed by human drivers.
I was in Washinton, DC yesterday to do an appraisal. I counted 4 separate incidents of dangerous driving and reckless behavior by pedestrians and bikers all of which almost took me out. By the time I got home I was a complete wreck having to deal with evading these 4 near accidents. The drivers here are the worst in the continental USA and getting worse every day, especially since the end of the pandemic.
Try Atlanta on the 285 Loop (totally under construction and loaded with BIG trucks). Any robo anything will be squeezed into a thin pencil or just flattened. Craziest place I have ever driven. And I have lived and drove in LA, Detroit, Chicago, Houston, Boston, NYC, and a few other big ones.
Like you say, DC is bad too,
How many humans did human drivers kills??? That’s the question you need to answer first.
A pedestrian almost took you out? What do you drive, a crystal vase filled with volatile explosives?
Gaston
Had to jam on the brakes to avoid hitting the moron who crossed right in the middle of traffic. Almost got rear ended in the process.
Howdy Folks. Watched the Jetsons as a child. Always knew it was a cartoon though. Love 15 mile cities and am glad there are still lots of places without so many people. Gotta go, Jim Bobs bull got out again and into the road….
That’s called “free range”.
My amish neighbors raise free range chickens for these city dwellers who insist there eggs come from a chicken that lives more like a suburbanite.
Here’s the thing I can’t get out of my head: i keep seeing that scene in the original Total Recall where the ’Johnny Cab’ goes berzerk and Arnold has to tear it apart. So, I guess as long as none of them are named Johnny Cab, I’ll be okay.
DM: Federal investigators have probed 42 Tesla crashes suspected of involving self-driving software – with 23 killed including two fatal collisions last month in California and Virginia
The Tesla crashes under special investigation by the National Highway Traffic Safety Administration (NHTSA) involve 23 fatalities since 2016, including two pedestrians and two motorcyclists.
Tesla doesn’t have autonomous vehicles. It has an autopilot which is a diver-assist thing, and it has the promise of FSD, and a beta FSD, for testing, but even that isn’t autonomous driving, but a driver assist thing. Both are shitty.
Quit mixing up driver-assist systems with fully autonomous vehicles. That’s a really ignorant thing to do.
Except musk and Tesla keep trying really hard to fool people into confusing driver assist vehicles with fully autonomous vehicles. People are only responding to marketing.
Wolf:
> The robotaxis are ALREADY operating.
> There is nothing for courts to decide.
In the real estate business there are ALREADY people doing work without permits and licensed contractors but once someone starts doing a LOT of it the government shuts it down.
In the mid 70’s (when I was a kid working for $1/hour I would personally take cans from CA on trips to OR to cash them in at $0.05/each (160 Coke and 7up cans from CA was equal to a days pay when I cashed them in at an OR store).
I probably made a total of $20 (close to $200 in 2023 money) sneaking cans in to OR and didn’t get busted but I just read that someone that was doing the same thing today on a large scale did get busted:
Unless the politicians (and the teamsters union) figure out how to make MORE money from robo truck drivers than they make now there will be a LOT of pressure to stop a major expansion (just like with the few cans I took to OR in the 70’s nobody cares if 1 in ten MILLION cars on the road has a robot driver).
P.S. I agee that self driving cars and trucks will be safer
Spending tons of money to put autonomous vehicles in cities has always seemed like a bizarre idea to me.
This technology I think was first commercially adopted in mining which kind of made sense. It’s an industry where they run a limited number of trucks in circles, round the clock, in order to perpetually move rocks across open spaces from one side of a mountain to another. Little traffic and lots of rocks to move.
But spending hundreds of billions to put autonomous vehicles in urban areas is something that seems less useful to me. It involves putting automation in the same cars and trucks that we’ve had for decades so they can drive across the same roads and highways that we’ve had for decades. There are only so many cars and trucks that you can put on these transportation arteries before traffic flows become unstable and break down into congestion or traffic jams. The point I’m getting at here is autonomous vehicles will not deliver anything faster, and it will not deliver more of anything over a given period of time – traffic will always restrict this.
What’s its really doing, particularly in an urban context, is replacing human drivers when there isn’t currently a shortage of them and the cost benefits are not substantial. Uber might want to replace its human drivers so it can collect the full fare for every trip, but then they will have to pay for all the transportation overhead that is currently covered by its drivers (vehicle lease payments, vehicle insurance, vehicle maintenance and repair, gasoline, road tolls, traffic tickets, and mobile phone bills). I’m not sure how much of a benefit they will get from implementing this technology – and that’s if it works.
Other transportation technologies, like railroads and steam ships and the internal combustion engine, all provided massive boosts in transport efficiency for their time. Railroads replaced horse drawn carts because they could transfer more freight over longer distances over shorter periods – a lot more. Autonomous vehicles on current roads and highways, however, cannot increase the frequency of trips, or shorten delivery periods, or deliver over longer distances. They can only replace the driver in what is an inherently restricted network.
There might be a way to slightly increase this efficiency, but it is unclear if the cost associated with doing this would be prohibitively expensive. It would involve building autonomous vehicle only highways, and assigning a networking router like computer to manage them. The computer would control all the autonomous vehicles on this highway, simultaneously, and determine the most efficient way for every vehicle to collectively travel from their entry to exit ramp. It would function in the same way as computer routers that are manufactured by companies like Cisco, which determine the most efficient way for data packets to make they way through a telephone or computer network. How these vehicles get to the highway on ramp or to their final destination from the highway exit ramp might be managed by drivers at a remote location in the same way that unmanned aerial drones operate. The point here is safety and some efficiency – but in a way that may not be cost effective.
So while autonomous cars or trucks are a nice novelty and a curiosity, this technology will probably not be able to provide the same types of efficiencies as our transportation innovations from previous centuries.
I fully embrace robo’s throughout the transportation complex – cars, trucks, trains, subways, even airplanes. Human error accounts for nearly all accidents, the sooner that’s eliminated the better. Robo’s also don’t have road rage, which is an increasing problem as our societal fabric continues to disintegrate. It will be curious to see how airfares on a robo plane compare to a human one during the transition.
It’s all good until the cancel mob or any other mob gets control. Who will influence this tech? Will you be stopped out of areas, turned off?
No thanks, we actually need a good old carbureted engine car with keys and roll up windows.
LOL. I want the steam engine back.
Nah.
Thing is, an old school simple car would sell gangbusters. I was thinking more along the lines of 60’s Mustangs, Falcons, 6 cylinders are fine.
That’s just nonsense. It’s said by people who’d never buy a car like that. Americans love big fancy vehicles with all the options and extras. That’s the first thing you have to understand about Americans. This is not Italy.
Back in around 1990, Ford tried to sell the Ford Festiva in the US. This was a Korean-made 2-door hatchback, with crank windows, 5-speed stick, no A/C (dealer-add-on), and manual everything. They got 50 mpg and held up well, tough little cars. We advertised them new at $4,999 and sold about 1 per month and lost money on each one. We sold something like 150 pickups a month, for well over three times that price, most of them nicely decked out (though nothing like trucks are today), and we made a ton of money on them.
People who say that Americans would buy “an old school simple car … more along the lines of 60’s Mustangs, Falcons, 6 cylinders are fine” don’t have a clue about Americans.
In the late 60s, even though I did some simple mechanics on my cars, and did some “experimental” driving as a teen, driving in San Francisco was almost always a stressful nightmare (going to rock concerts, big city stores, etc.). Human drivers would seem to enjoy positioning their car a foot or so off my car’s rear bumper when they saw I was dealing with a manual transmission at the top of a steep street waiting for light to change (car started sliding quickly backward when foot taken off the brake and put on the clutch). Seems like a well-programmed robotaxi could be more considerate.
I would have loved to take the train up to SF (from silicon valley), and hopped on a robotaxi to get to my destination.
You should have learned how to “heel and toe” your manual transmission equipped car. Your problem would have been solved.
I think much of this conversation about autonomous vehicles is people talking past each other because they are thinking of it as a binary decision (are autonomous drivers better that human drivers?) rather than a spectrum of instances and figuring out when autonomous car are better than human drivers.
I would definitely agree that for most basic driving tasks that only deal with the road and other traffic that autonomous drivers are far better. I think autonomous vehicles struggle where there is something abnormal about the road/environment or when driving requires a deeper understanding of human nature. Don’t get me wrong, human drivers struggle in those situations as well, but not as much as autonomous vehicles do.
Some examples:
1. I remember driving in downtown Chicago during rush hour where the sidewalks are completely packed with pedestrians from building to curb. The issue is the curb drops off from the sidewalk straight into the first lane of car traffic. This means that if a car was driving in the right lane (say planning to turn right in the near future), that car was driving parallel to a wall of moving pedestrians only a couple of feet away. Any rationally safe autonomous car would immediately stop with pedestrians so close. A human driver can continue to drive past (at slow speeds) because a human driver can read the body language of the nearby pedestrians and see they are not going to step out in front of the car. Again, this really isn’t a driving problem, it is a reading human body language problem.
2. Las Vegas (off strip in the residential areas) has a lot of dangerous pedestrian crossings. These are on streets that are 35 or even 45 mph. There is no lights or stop signs for vehicle traffic. Cars are allowed to drive by at normal speeds if no pedestrians are present, but are expected to stop/give right away to any pedestrians that are present. Yes, it is stupid and dangerous. Pedestrians get hit far too often. Yes, autonomous vehicles are probably already better at these types of interactions than human drivers. Here is the thing, as a careful pedestrian I don’t cross until I am sure vehicle drivers are aware of me. When the cars slow down, I can look at the human drivers in the eye and see what they are looking at. So even if cars do not completely stop, I can feel safer that they see me. I have no idea if an autonomous vehicle “sees” me or is just slowing because other cars on the road are slowing. I would have to wait for an autonomous vehicle to fully stop before proceeding. Like I said, not a driving problem, but more of a human nature problem.
3. Seeing dangerous situations before they happen. I can think of numerous situations in recent years where when I was driving where I saw something that caused me to slow before an autonomous vehicle would have. The first was when a car driving in the lane had its front right tire blow. It literally blew up, shedding lots of rubber chunks all over the road and was quickly driving on a rim an sidewall remnants. It took a second or two, but it eventually swerved into my lane due to the blowout. Point is, I started slowing when I saw the blowout. An autonomous vehicle wouldn’t have slowed until the car started swerving. Only a couple of seconds, but it made a difference. The other instance was when I was passing a pickup truck crammed full of junk. I saw that some of it had come loose and was moving around. I was able to slow BEFORE it started flying out of the truck bed and hindering traffic. Autonomous vehicles would have seen both vehicles, but they wouldn’t have recognized the problems until their path was blocked. Again, it is just a human knowing how the world worked better than a computer with limited knowledge.
4. One time driving westbound on interstate 20 in west Texas, literally middle of nowhere. Sandy, rocky ground covered with sparse grasses and cacti as far as the eye could see. An interstate with 2 lanes in each direction. A car going eastbound had crossed the median and hit a westbound car with deaths likely in both cars. Anyway both westbound lanes were blocked by the accident so westbound traffic came to a dead stop. Emergency vehicles were literally just arriving on the scene. Human drivers were able to recognize that the rancher whose land butted up against the highway had a dirt track running parallel to the interstate about 50 -100 feet away from the highway. Human drivers were able to carefully drive off road for the 50 -100 feet, get on the dirt track, then drive down it a few hundred yards, get past the accident, and then get back ofln the highway and continue on. Even sedans were able to drive on the areas just off of the highway (if they were careful) in order to reach the dirt track. It would never occur to an autonomous car to drive off road and use an unmarked road to get around the accident. As it was, there was barely a traffic jam waiting on the accident as most went around it. If most of the cars were autonomous, the traffic jam would have stretched for miles while they all waited for the accident to clear. Again, not so much driving ability as having more worldly knowledge of what is possible.
5. I recently attended an event that was held in a big field with thousands of attendees The rideshare pickup area was in a driveway area adjacent to the field. After the event was over, the driveway area quickly became clogged with rideshare drivers and passengers trying to find each other. It was a mess. However since the rideshare drivers were human, they realized they could go off of the driveway into a flat area of the nearby field. This increased the rideshare pickup area by 3 or 4 times the initial space and allowed vehiclesto get by waiting vehicles. Things ran much smoother. Autonomous vehicles never would have left the driveway. Again, nothing to do with driving, more to do with worldly knowledge than driving ability. It should be noted however that this was a situation that was designed for autonomous vehicles (dropping off and picking up passengers from a concert/sporting event) yet human drivers did it better.
6. A road with 2 lanes in each direction. A vehicle had become disabled in the left lane going northbound. It was raining pretty hard so there was water accross all 4 lanes. The disabled vehicle was leaking gasoline. The gas could be seen mixing with the water in both northbound lanes. So even though the rightmost northbound lane was empty, a civilian was directing northbound traffic to go into the rightmost southbound lane. A police officer was on the opposite side of the road directing the two southbound lanes to merge into the farthest lane. A 4 lane road effectively become a 2 lane road in the southbound lanes. All of this was done without cones and any signs. Just with the hand motions of a couple of humans directing traffic. I don’t think an autonomous vehicle would be able to recognize the signals from a human and cross over into oncoming traffic to safely get around the disabled vehicle.
My point in all of this is that initially I see autonomous vehicles being used in a limited capacity. Geofenced into a specific area. For example, in a downtown area, most cab rides begin and end within a set area, no reason these could not be done autonomously. Or even repetitive long haul truck routes between major western cities to specific hub areas. Furthermore, I think the faster we (at least initially) give up on the fully autonomous dream and just accept that there will be weird exceptions that will arise that require human decision making, the faster autonomous driving will be accepted and come about. There is no reason a few hundred autonomous cabs cannot be running around a downtown area doing their thing, and then when they hit a unique situation that is beyond their programming, they alert a human who takes control of them (even remotely) and gets them through the exception. That is how current automated manufacturing works. Robots do their thing and alert humans when there is a problem.
TDLR, computers are already generally better at actual driving than human drivers are, but occasionally driving requires skills that go beyond just driving.
Put a different way, these autonomous vehicles are already packed with electronics. There is not reason they cannot be packed with cellular equipment to call a central office when confused. Stories of autonomous vehicles being stopped by pedestrians putting orange cones on their hoods should be stories about autonomous vehicles pausing for a few seconds when a pedestrian puts an orange cone on its hood then starting up when a human operator remotely assesses the situation and takes control.
Yeah… maybe equip them with a robotic arm that extends from the trunk and removes the cone.
Go-go gadget arm!
/s
Jim L: excellent post.
Robocars are not in the mid-term future (5-8 years) in most environments for the following reasons (not in any particular order).
1) Insurance issues. Unless a government authority waives their financial responsibility, insurance companies will be paying out enormous sums for a robotcar’s mistakes. So, they will not be enthusiastic about insuring them. Rates for humans are largely based on demographics and driving record, IOW actuarial factors.
2) What does the robot car do when faced with killing the little girl chasing her bouncing ball versus killing the 3 adults in a head-on collision?
3) Hacking. There are plenty of bad actors in this country. Whether for intent to hurt or just for kicks.
4) There are many roads in the area where I live (small city/suburban/and country) where a speed limit of 25 mph is largely disregarded by 99% of drivers when conditions allow–daytime, dry road, 1/4 mile visibility. I can’t imagine a situation where many or all robocars are programmed never to exceed the speed limit. Maybe fine for grid-type cities as you cite. A less efficient form of good mass transit, (lack of which is the real problem), but not practical from a behavioral point of view. Note: Wolf constantly cites the 40K annual deaths by our imperfect drivers, but wouldn’t an excellent way to reduce this be simply requiring individuals to re-test periodically. Like other commenters, I consider our local drivers to be the worst in the world, and I’ve driven in challenging conditions on four continents. After a near miss, I sometimes shake my head and wonder if the driver got his/her license from a Cracker Jax box. (This dates me, I know).
5) I feel that there is a limit to what our biped bodies and minds will accept from tech. Again, maybe it’s my age, but I will NOT sit as a passenger in a robotcar over which I have no discretion to intervene. Driverless programmed trains on rails, yes, programmed buses on fixed routes, also. But not a robotcar.
Good comment about the speed limits.
In metro Boston, the unofficial speed limit is 10mph above whatever is posted, and its considered rude to go exactly the posted speed limit.
This contrasts with rural NH and Maine, where the posted speed limit is the actual speed limit, since its dangerous to go faster.
Horse poop! I live in Maine and know that what you say is NOT the case. People here drive FAST!
For comparison, I winter in rural Florida where you better not drive over the speed limit. It is a pleasure watching an ignorant person ‘from away’ get pulled over for doing what they do back home.
FYI: robocars are already here, whether you want them or not.
Yeah, sure they are. In limited local environments. My point was/is that, beyond these relatively small scale experiments, the adoption of robot cars will be very very slow in most areas of the USA. Maybe a decade or two. Aside from the points that I made above, the infrastructure and local government regulations are not going to happen very quickly. And, while not comparable in scope or kind, the adoptive experience of Segways and E-scooters is instructive.
Your main argument is that robotcars will reduce the horrible number of deaths that are due to poor drivers. I agree that most drivers, even excluding the drunk/drugged and iPhone-addicted, are not very good. (Some surveys show that something like 75%+ of drivers consider themselves “above average”, an impossibility). There are, however, other, more conventional ways of reducing the number of deaths. Look at what happened long ago with the introduction of seatbelts. First simply available, then universal, then mandatory. Same with airbags.
Finally, I come back to the idea that there are built-in limits as to what our bodies/minds can adapt to. Yes, younger people have had a personal interface with new tech for all their short lives. But, even among this cohort there are the beginnings of resistance to “take care of everything” tech.
There is one other situation that I will be very interested in seeing how it plays out. That is autonomous cargo vehicles. It will be trivially easy to stop an autonomous cargo vehicle and quickly rob it. Yes, that can be done against a cargo vehicle driven by a human, but there are two significant differences:
1. Autonomous vehicles are easier to fool. If they get confused they stop. A human driver might intuitively see the trap and avoid it or at least be able to figure a way out or around the block.
2. It is human nature that makes it easier to steal from a faceless automated truck than it does from a human driven vehicle which might require the thieves to have to hurt another human.
I am not saying this will be a deal breaker for autonomous driven cargo vehicles I am just thinking it is an area than will be interesting to see how it plays out and how it is handled.
1, Have you seen the James Bond movie where the car blow up when someone try to break into the car? Robbery will be a race between measures and countermeasures.
2. In some cases I doubt this, and the flip side. Easier for the operator to push the countermeasure switch against targets at the screen…
JimL- excellent examples. It’ll be functional on a limited basis eventually, be a decade or more before it’s seen in tertiary markets. But we shall soldier on through the carnage…..
TCAS was mentioned earlier, it’s imposed major expense on general aviation. But there’s plenty of GA mishaps still occurring. Maybe it’s a viable, effective collision avoidance system. And maybe some avionics companies lobbying the FAA had some effect? They needed the work. And bureaucrats will always need regulatory work. Technology will oblige them, for good or ill.
JimL says:
> It will be trivially easy to stop an autonomous cargo
> vehicle and quickly rob it.
Maybe the robotrucks will have a “robocop” option that includes a shotgun to defend the truck like the human driver in the link below that prevented a truck robbery
So many things— robo fire trucks, robocops, tax stupid humans to a point where stupid people can’t afford to drive on public streets. That brings up the public utility of making cities safer through robo dominance.
Forget the 15 minute city, people shouldn’t be wasting time lingering in grocery stores or thinking they need the luxury of being served by stupid humans. Why waste vast amounts of land pretending that humans need ineffectual urban cores, when that land can be converted to the higher purpose of producing robots.
This obsession with human needs has been out of control for too long and obviously these sprawling garbage spaces need robo bulldozers and AI progress!
Autonomous passenger cars sound like a wonderful idea given the number of human error related auto fatalities each year, however deployment on a large scale is still decades away and that’s an optimistic estimate. Its easy to forget in the midst of all the excitement surrounding this technology that we just simply do not have the infrastructure in place on a wide scale and to implement such would require staggering amounts of capital. We can’t maintain the existing infrastructure in this country. As countless roads and bridges fall further into disrepair, it’s laughable to believe AVs as far as passenger vehicles could be deployed on wide scale to the extent of impacting ride sharing services. It’s akin to California’s idea of banning the sale gasoline cars despite not having the infrastructure in place to support the enormous demand placed on the power grid from EVs. Cart before horse mentality…
If you believe AI can eventually operate and think like a human at least in most scenarios, there is zero reason to believe you need anymore infrastructure than you have now in which humans are currently driving.
And I think that is the ultimate safety goal: that your robocar isn’t reliant on a gps map, signage m, etc. it’ll be able to think and realize “oh yeah, I should take a turn into the Kona’s harbor because cars don’t belong in water”
Ironically at least 1 human didn’t pass that test, and I think there are more humans that have done similar
Thanks Wolf, that was fun listening to your descriptions of the evolution of robots.
Since I am a full time cyclist (no 4 wheelers) I am looking forward to robos. 99% of the time cycling is the better solution for me but every once in a while it would be handy to hire a robo.
‘ROBO’ is my nomination for the generic name of these.
Long as I can pay extra to have that robo taxi “prioritize my life” over everyone else incase of emergency. Otherwise no thanks :)
The argument if accountability that seems to pervade these comments is illogical.
I’m a cyclist. When I get killed due to some distracted driver hitting me, the knowledge they are criminally liable doesn’t matter. Not to mention most people get a slap of the wrist via the excuse “I didn’t see them”….like it’s acceptable to drive when you can’t see.
I fully trust AI cars not to hit me (at least eventually). I fully trust humans to get worse especially as phone mfg’s seem like they won’t deactivate phones used by a driver (and that phone definitely can know it’s being used by a driver)
What happens to traffic stops with an autonomous vehicle? How would law enforcement pull one over?
Why would they even pull one over? Just get their tag and discuss the issue with the owner.
It doesn’t take a whole lot of imagination to think up reasons law enforcement would want to immediately stop an autonomous vehicle. In fact, it wouldn’t even require any imagination. One could mostly use the same reasons they would want to pull over a non-autonomous vehicle:
1. The vehicle might be a hazard if it keeps operating.
2. The vehicle is suspected of being used in the commission of a crime.
3. The vehicle is carrying people who are suspected of committing a crime.
Point is, autonomous vehicles need to follow the driving laws and stopping for law enforcement is one of those laws.
All three reasons are BS that you dragged by their hairs into this discussion. Think about it!!! I don’t know why I have to waste my time with this BS.
1. Police doesn’t know when a vehicle is a “hazard,” such as the brakes not working. If one of the taillights is out, then can just sent the warning/ticket to the owner.
2. and 3. Are the stupidest comments I have read all day (well, I just started reading comments, so there will be more to come to unseat yours). NO ONE is going to use a robotaxi in a hot crime where police might pursue the vehicle, LOL, because you cannot control the vehicle. DUH
I for one am tired of the idiotic, aggressive, distracted, and entitled drivers around me. Worst of all the is complete lack of traffic law enforcement, in person or electronically. I welcome AI and full 5g low latency mobile edge computing with open arms vs absolute morons piloting 2 ton vehicles.
I agree. Problem here is the vast majority of drivers do not obey traffic laws, and if the sheriff decided to enforce those laws he would not get re-elected.
We have met the enemy ….
Howdy Folks. Maybe off topic and sorry. Didn t a large group just purchase large amounts of land near San Fran near a military airport? Maybe they are gonna create that 15 minute city from scratch?
The Bay Area already has some 15-minute cities, biggest one being San Francisco. I don’t think that’s their purpose, though it may be structured that way. They may want to make a gazillion selling this thing in bits and pieces to home buyers, people who want to set up shop there, etc. Maybe they can sell the whole project to the public via huge IPO, get richer quick, and then let the whole scheme collapse?
These Silicon Valley types always angle for the next big thing. But they chose Solano County, and the locals are not happy, and some of the land is family farms, and this city is in the flight path of an air force base. So this will be interesting.