Every time I do a post on self-driving vehicles, readers respond that it cannot happen and won’t happen for a decade, if ever.
It’s time for a reality check.
A quick check of my calendar shows we are not quite halfway through 2017. And a quick check of headline news shows Waymo’s Self-Driving Minivans Now Offer Rides to Real People in Arizona.
Starting today, residents of the greater Phoenix metropolitan area can sign up to go for a ride in a self-driving minivan. As often as they want. For free.
Waymo, the self-driving car startup spun off from Google late last year, announced today that it’s offering its services to members of the public for the first time. Waymo is calling it an “early rider program,” intent on cataloging how on-demand, driverless cars will factor into people’s everyday lives. Interested participants can sign up on the company’s website, and Waymo will select riders depending on the the types of trips they want to take and their willingness to use the self-driving service as their primary mode of transportation.
A Waymo test driver will be behind the wheel at all times, but the company insists that the vehicle will drive without human intervention as much as possible. Rides will only be available to residents of Phoenix and the surrounding towns, like Gilbert, Tempe, and Chandler. Waymo describes the service areas as twice the size of San Francisco.
In order to accommodate what it hopes will be “hundreds” of riders, Waymo is ordering an additional 500 Chrysler Pacifica minivans from its automaker partner, Fiat Chrysler, which it will then outfit with the laser sensors it manufactures in-house. Waymo already has 100 self-driving minivans that have been driving the streets of Phoenix and Mountain View, California, since earlier this year.
“We want as many people as possible to experience our technology, and we want to bring self-driving cars to more communities sooner,” said John Krafcik, CEO of Waymo, in a Medium post published today.
Early Rider Program
What better way is there to create demand by giving away free services. That’s just what the Google spin-off did with its Early Rider Program.
Safe Predictions
Numerous readers emailed various links to this story. My favorite comes from reader Wade who wrote …
Hi Mish,
I enjoy your back-and-forth with the unbelieving Luddites about self-driving cars.
I’ll make three predictions:
- Self-driving cars will work
- They will be hugely popular
- Nobody will admit they were wrong in being a skeptic
Wade
If you think Google, Amazon, Ford, Toyota, Nikola, GM, Tesla, Apple, etc, etc, etc, will all fail with driverless and electric, you need a serious reality check.
Mike “Mish” Shedlock
I can name one group that are going to love self-driving cars. People with mobility issues. My best friend can’t drive for that reason, and for him, when they come along, a self driving car == freedom.
Lots of groups
Families
night impaired
people who just do not like to drive
People would rather relax than drive
nearly endless
Mish
There is a driver inside these vehicles and according to Uber he has to intervene every 0.8 miles to safely arrive to a destination. This is so early in development that I doubt it will happen in next 10 years. More likely in 20-30 years.
https://cdn0.vox-cdn.com/thumbor/dmF7NsyETXepZfDJuWqBqqWCNwA=/1000×0/filters:no_upscale()/cdn0.vox-cdn.com/uploads/chorus_asset/file/7772655/waymo_chrysler_pacifica_detroit_auto_show_2017_0621.jpg
still not able to be driven in snow or heavy rain.
the facts speak for themselves
so do the lawsuits – example
The lawsuit goes on to allege that Model S and Model X drivers have “become beta testers of half-baked software that renders Tesla vehicles dangerous” and that Autopilot-enabled Teslas can veer off the road and experience “lurching, slamming on the brakes for no reason, and failing to slow or stop when approaching other vehicles.”
” sensors cost $7,500 for Waymo to build itself. ”
now count the sensors on the car and tell me what
the car costs with the computer and sensors.
no one ever mentions price.
I can buy a brand new 2017 Ford Fiesta
MSRP: From $13,660
.
Sensor prices will fall with volume. Chances are, self driving cars will be cheaper than human driven ones fairly quickly, once/if they can be made to work without the accouterments required for a human to operate them. For one, they’ll never speed, so don;t need much in the way of high speed enabled hardware.
As long as manual override equipment have to be present, which may be for quite a while, you’ll “pay double,” but the electronic stuff will soon be cheap enough for that not to be too big a deal.
There is heavy rain in Singapore and they already have driverless cars there. A friend of mine posted a video to her Facebook page a few weeks back. She and other locals are excited for them because getting a taxi when it is raining can be problematic. (The taxi drivers stop taking new fares until it stops because they are afraid of accidents.)
I wouldn’t climb in that sucker if they paid me.
Putting one’s children in a vehicle with no human driver during this experimental phase is beyond the pale. What are those parent’s thinking?
Those who think a massive roll out of these futuristic ball of electrical circuits on wheels has any chance of widespread success must have been obsessed with B-rated C Sci Fi flicks as children and were unable to make the transition into adulthood..
There are 1001 individual reasons why this could never work.
I’ll give you my keys when you can pry my cold dead hands from the steering wheel.
You are FN crazy
Driverless cars way safer than human driven cars
LFOldTimer’s post is so crazy he could be a parody of a Luddite.
Who’s running the insane asylum here?
My points were perfectly rational and logical.
You remind me of the snowflake Berkeley protesters who call the conservatives names.
In the end I will be the one who was proven right. And you will be nowhere to be found. 🙂
You already admitted it won’t happen in your lifetime. You will be missed LF.
I’m the farthest thing you can imagine from a Berkley protester.
Safe until someone doesn’t like you and knows how they work…reference Brietbart
…Scalia?
@LFOldTimer: It it is any solace the next Waymo advert will be directed by Tarantino and feature Willis, Stallone and Schwarzenegger blasting terrorists with ak47s and bazookas ….
human drivers are the leading cause of death for children today
You would have to be insane to allow your children to be driven around by a human driver
My daughter was on the school bus when schoolmate Karla Campos stepped off and immediately run over, right in front of Karla’s waiting mother.
The 82 year old driver of the vehicle was apparently startled upon reaching the crest of the hill seeing vehicles stopped in front of her behind the bus. The speed limit is 35 MPH on this two lane road.
The driver stomped the accelerator instead of the brake swerving around the right of the stopped vehicles and the bus striking Karla and coming to a stop a hundred or more yards later.
Fortunately, my daughter had no idea what had happened.
A horrific incident.
Google “karla campos mountain view elementary school”
That’s horrific. 100 years from now, the period of human-driven cars will be written about much like the Black Death. Future societies will marvel at how primitive and barbaric we were.
“..human drivers are the leading cause of death for children today…”
I know! Canine drivers ones don’t cause nearly as many kids to die in traffic accidents. So, to appear all progressively sciency and statistic’ly, we the gommiment should ban those horrible humans from driving, and let the dogs take over…..
And in another piece of startling news: Guns are a leading cause of gunshot wounds for children. So the gommiment should go ban something and somoene!
Sometimes that kind of sarcasm is warranted, even obligatory. But the statement “human drivers are the leading cause of death for children,” assuming it is true, is perfectly reasonable. Sure the word “human” might be superfluous, but what the hey, the statement is still relevant.
Humans are the only even remotely viable drivers available. Swap them out one for one with any other form of driver: Robot, dog, what have you, and it’s not as if that makes children any safer, as long as cars are still being operated the way they currently are.
Almost anything that kills children anymore, has humans involved at one level or another. From human built houses burning down, to human transmitted diseases. Doesn’t mean one is insane for not keeping ones kids away from any and all contact with anything human nor human created.
Of course driverless is safer
The risks for kids are different but most of the problems are easily solved. Some will insist their kids not get in driverless cars.
So what?
But insinuating one must be insane for letting a kid ride in a car driven by a human, on account of some pointless “statistic”, stretches “pretty reasonable” quite far……
Right, it just means “More children are killed in car accidents then by any other cause.” If computers are better at driving then humans, then fewer kids will be killed. Unfortunately 100 million jobs world wide will be lost when cars, trucks and taxis drive themselves. The leading cause of child death will then be starvation.
leading cause is already starvation, poverty, malaria etc.
the leading cause of death
more than guns, drownings, cancer, falling out of trees,
drivers kill more children than any other
This will be a happy happy story until a Waymo gets T-Boned and totalled during a high speed police chase killing everyone on board … one incident like this could stop the adoption for some time …
Mish I am one who has generally agreed with you on your assertions – the electrified and fully automated Dubai Metro system has been in place for some years now it is clear that the technology exists and works on a commercial and industrial scale. The Dubai Metro transported around 8 million passengers in 2009 and today carries about 90 million – to my knowledge on time and without incident.
Adoption of these new systems is more of a human problem than a technical one – sort of a parallel to the adoption of the modern bathtub (1880s) which took about a generation to be implemented widely in American households …
Human interest will gravitate to the scofflaw human that T-Boned the driver-less van. His record will become the issue. The conclusion promoted by insurance companies? Outlaw human drivers if you want senseless deaths to stop.
Yep… add in a few studies that driverless cars are safer, and the poster above wont have to worry about fighting to have his keys pried away. Cars just simply will no longer have keys and whatever he is driving now won’t be eligible to pass inspection, or at best will be grandfathered in and be the last manual car he can ever purchase. All in the name of public safety. That police chase will just add one more reason to take ’em away. No more need for police chases.
I actually don’t think it will be this involuntary. You’ll likely be able to operate a human driven car for some time as long as it is upgraded with level 4 sensors, but your auto insurance will likely charge you per mile driven with your hands on the wheel. The cost of which will be very high because with every self-driving car on the road, that is one less person paying into the auto insurer’s pool, thereby increasing claims payouts that are spread to the remaining human drivers. At a certain point, a human drive car will become a luxury.
Dubai Metro is on rails and thus not relevant to this. Vancouver had driverless metro (Skytrain) since 1985. That’s 32 years ago. Still no driverless cars there.
It is relevant – growth of 8 million to 80 million in 6 years illustrates that humans will adapt to automated travel rather quickly – at least within the cities. This is a clear example of Mish’s expectation that it will be adopted quickly.
The global growth potential of this new system is simply astounding. It will be another gold rush.
Autonomous autos are light years more difficult than autonomous trains. Not to say it won’t happen, but it is comparing apples and jetliners.
The technology is a separate issue, I agree, but this is simply a matter of time and testing, of course.
Segregating cars onto tightly controlled rail, pretty much solves all problems related to autonomous driving in one fell swoop. Plus most issues related to short hop air travel. The most densely populated parts of the US, which is where the banksters funding the whole “autonomous” fad lives, may be too effed up to be able to take much advantage of that.
But in less compromised parts of the world, its a much higher upside path than forcing robodrivers mix it up with stray dogs in Tijuana, Cows in India and road raging drunks in Los Angeles.
Whilst I really appreciate Mish’s articles; this subject seems to come up a little too often and frankly, I do not understand why. I am sure there will be some room for driverless cars. However, I don’t believe we will switch en-masse.
Do you really believe we really want:
a) To be tracked at all times?
b) Lose control to car makers/governments/etc – take your pick?
c) Slowly morph into unemployed sacks of biological matter devoid of any purpose?
d) Not be able to decide where we go on the fly; look out the window and decide we will take a right turn because we like what we see, or something pops into our mind which requires a change of route.
e) Be removed from the experience of travelling and chimp over our mobile devices whilst life passes us by?
I find the whole concept absurd-what is the point of living if we stop doing anything!
I understand there will be some valid uses-restricted mobility – deliveries etc, but I sincerely hope we have not reached the point where people see this as a good thing.
With regards to electric, I am all for the gradual adoption, but not using battery technology of today. There are far too many losses along the way of supply and environmentally it is still not sound (including manufacture and replacement/scrapping of batteries. Fuel Cell and other technologies might eventually have a place.)
I am certain that “safety” and other supposed advantages will be used to push us in this direction, much as “terrorism” and “Child abuse” has furthered the increase of regulations and surveillance.
I think and hope that this is not a future people want.
This will be most attractive in densely populated cities where residents use mass transit or taxis anyway. For the rest of us, not so much.
Most likely first implemented outside the US, in places like Singapore and Tokyo – the USA will be a late adopter in comparison.
likely most useful in the ex urbs where people spend 1/3 of their waking hours driving and every adult needs to own a private auto.
SD taxis will replace exurban households 2nd, 3rd, and 4th cars.
Dense urban environments don’t use cars much SD or Human. They don’t need them
I certainly abhor more surveillance. Unfortunately, the incremental gain from your car, pales to what is already gleaned from your phone… even with your GPS off and no malware (big assumptions). Cell towers, texts, receipts…they all cross reference with your friends and family to provide an amazing composite. Realize how many use cash, and you will know how many value privacy.
What does it matter what you want?
Driverless is 100% guaranteeed
just like the deposit in in my bank.LOL
seriously though, do you really believe there will be 100% adoption, or specific use?
Yes, same concept: you want your deposit physically in the bank…
but it doesn’t matter what you want…
you get a guarantee instead.
The only “people” that have a chance of getting what they truly want are Corporations.
Do you seriously think you are not tracked 100% of the time now? Do you own a cell phone? Do you drive a car? Do you pay bills? Do you use the internet? (we already know that’s a yes) Do you use any electronic payment methods? Do you have any devices (including your car) connected to the internet or which transmit data via radio signal?
Do you have ANY idea what data “you” transmit and where it’s transmitted? Judging by your first statement, the answer to that is clearly no.
a) To be tracked at all times?
If .gov wants to track you, they already can.
b) Lose control to car makers/governments/etc – take your pick?
You see it as losing control, I see it as being liberated from a system in which we are forced to choose between living in a city without reliance on cars, or living in the suburbs, with heavy reliance on cars. And what is the obsession with this idea that the government is going to be the provider of all these cars? Is this communist Russia? Is .Gov working on a self driving car concept we don’t know about, and planning to keep all the private companies out?
c) Slowly morph into unemployed sacks of biological matter devoid of any purpose?
Sounds like what happens to the 1 million people who die in auto wrecks every year. Or the hundreds of thousands seriously injured who can no longer work because someone had road rage that day and caused a wreck.
d) Not be able to decide where we go on the fly; look out the window and decide we will take a right turn because we like what we see, or something pops into our mind which requires a change of route.
Voice commands to the car to alter the route? Have you been living under a rock?
e) Be removed from the experience of travelling and chimp over our mobile devices whilst life passes us by?
Or take a nap, or talk to the family and play games, or read a nice book, or catch up on some work, or eat, or not die in a wreck because some chimp is looking at his mobile device?
did I say lose control to Government? NO I mentioned a number of alternatives. Not to mention all the stakeholders in the control systems/grid etc that make this work. It makes us more dependent. I rather strive to be more independent.
If you think that this will prevent all future accidents then good luck with your utopian dream.
Quality time with your family whilst driving? give me a break.
Anyway judging by so many fans, it will come but universal? no way not worldwide or even nationwide.
I have yet to hear some reasons why this is such a great thing. the safety thing is a red herring. Some lives may be saved but do we really need to let go of all our liberties thinking we may save lives. For all we know that extra napping and mindless entertainment during our rides will lead to shorter lives.
Boring subject-made a mistake even reacting to this post but as Mish brings it up so often, I thought I would throw in my 2 cents. Far too opinionated and entrenched.
Some more groups:
Lids get very heavy with a mundane commute
Would rather be talking to friends w/o distractions
Like to meet some friends at a pub for a few beers
Like to travel overnight without loosing sleep
It can cause problems for vendors of single occupant Motel rooms when people can sleep on the move. Service stations offering tended, clean showers and breakfast could benefit on stops after a night moving. Beds on wheels.
Sounds like the people who used to ride the Southern Pacific Lark and Owl trains between the Bay Area and LA. Board the train in the evening, have drinks and a nice dinner with friends or coworkers, sleep, have a morning shower and breakfast, and get off the train ready for a day of work.
Doesn’t matter what we think, it’s not going away and the only question is rate of adoption.
slightly disagree. I think it matters what we think and we should make sure we are heard by acting accordingly whatever the issue is. If there is anything we want to see differently, we should ensure opinion is heard. If we don’t, we will become a heard of sheep.
I do agree it won’t go away though.
bring it on
the only way to know if it works is increasingly larger scale pilots
even if there are minor flaws in driverless systems it might still be cost effective to pay for the little old granny or the baby stroller that gets crushed on the path to tech progress. Local insurance, licensing and bronx juries are pitfalls to widescale adoption.
granny… nothing is scarier than driving in Florida and seeing two hands on a wheel and a grey head of hair barely above the steering wheel. There is a reason people joke about someone driving into the community lake every day. They do.
LOL there are bike lanes almost as wide as regular traffic lanes on some major roads in the Naples area, it’s suicidal to use them, really confusing lane markings at big busy intersections
We call ’em “blue hair and knuckles” drivers.
Rain, snow or dust in the forecast and these things basically pack up and go home.
Any sort of minor mechanical or electrical fault and they have to be taken out of service.
Any need to travel outside of areas that Google has fully mapped, and they don’t get to be used.
Google engineers and mechanics will be coddling these vehicles extensively pretty much round the clock.
Nobody denies that Self-driving cars can be made to work under very tightly controlled conditions. The technology exists. But its a whole different ball-game making them work extremely reliably, and economically.
“Any need to travel outside of areas that Google has fully mapped, and they don’t get to be used.”
But all you need to do is drive a vehicle with the same sensor package as the autonomous vehicles down the unmapped road and it instantly becomes mapped. If you had said 10 years ago that no one would have a comprehensive picture of almost every street in the industrialized world (and then some) due to the shear volume of roads to be traveled an information involved you would have been wrong. Google, pretty quickly and quietly, went about driving those roads and you can now see a vast majority of the roads in the US as well as many other countries around the world as though you were standing there.
Per the article Mish cites: “A Waymo test driver will be behind the wheel at all times…”
If you liked Segway you’ll love self driving cars!
Segway is now made in China and sells for about 1/10 the cost of the original, the chinese owner (ninebot) appears to have reverse engineered segway and undercut segway price before Donald Trump entered office!
Segway has changed hands several times in recent years. Kamen sold it to U.K. businessman Jimi Heseldon in 2009 after raising about $176 million from investors. Kamen didn’t respond to a message seeking comment for this story.
Tragedy struck Segway the next year after Heseldon was killed after he lost control of one of the scooters and fell off a cliff while walking his dog. Experts didn’t find any mechanical defects with his Segway. Two years later, Summit Strategic Investments acquired the company and subsequently sold it to Ninebot.
Segway: the perfect device for people who are too lazy to walk and enjoy looking like a dork at the same time.
segway a brilliant solution searching for a problem
Bring it on. Robot cars would be a blessing for the elderly.
and drunks
no need for drivers licenses
Pets too.
I guess what’s not clear to me is what degree of Adoption does Mish see for driverless vehicles (Trucks and Cars) and when. Does he see 100 percent adoption within the next couple of decades? If so, I see that as unrealistic. Does Mish see Emergency vehicles being self driven as well?
For cars in an urban/suburban environment could driverless vehicles have a large role? Could segments of the population benefit from the adoption? Answers are clearly yes. But outside cities in flyover country (where there are still many gravel and dirt roads) I don’t see it. The population density just doesn’t support it even in semi rural areas.
For Trucks, could long haul trucking be automated yes, but flatbed trucking, last mile and van delivery? no. there is always going to be human intervention to handle the difficult maneuvering in a driveway, crowded city street or, a crowded parking lot.
I see automated vehicles playing a significant role but only in niche areas.
It will be interesting to see how well these vehicles do once they are out of the pilot stage and components are made by the lowest bidders. I am also a huge skeptic that the nation can afford to both upgrade and then maintain our infrastructure to the level to support the mass adoption of self driving vehicles. We can’t even afford to maintain what we have today.
I think we could afford to maintain what we have today. We just choose not to.
“But outside cities in flyover country (where there are still many gravel and dirt roads) I don’t see it. The population density just doesn’t support it even in semi rural areas.”
I grew up in flyover country and outside of driveways I rarely traveled on dirt or gravel roads. So because some exist we can’t use self-driving cars for the 99% of roads that are paved? Maybe the solution is paving the gravel roads?
here is my crude adoption rate ( Mish is slightly more conservative than me )
2017 – Alpha testing of level 4/5 ( thousands of units )
2018 – Widespread beta testing level 4/5 ( tens of thousands of units )
2019 – Tesla retail full level 4 plus early adopters of others makes ( a few hundred thousand units )
2020 – level 4 retail 2% market share, level 5 beta testing ( 500,000 units )
2021 – 5% market share ( 1,000,000 units )
2022 – 10% market share ( 2,000,000 units )
2023 – 15% market share (3 mill)
2024 – 25% market share (5 mill)
2025 – 30% market share ( 6 mill)
2026 – 35% (7 mill)
once SD reaches 35% market share, I am guessing there will be a significant reduction in growth rate with a upper limit of perhaps 75% share until 2040 or so. The grumpy old farts will need to die off before exceeding 75% market share.
“A Waymo test driver will be behind the wheel at all times, ”
What’s that? Virtual reality for self driving cars?
If the goal is to save money on drivers, replacing the average Uber driver with a Waymo engineer, probably isn’t the optimal way to go about it in the long run 🙂
As I still say, this has metropolitan uses only. It will be used similar to buses, taxis and uber in Americas top twenty cities. Look for uses to fluctuate with cost. I have still not seen any system offering affordable transportation that turns a profit. Uber is case in point. This will happen with LIMITED use.
Any early (next 10 years) traction likely due to sinkhole for public $$s.
And it 10 years Mish will be railing against public $$s used here.
The Circle will be complete.
“If you think Google, Amazon, Ford, Toyota, Nikola, GM, Tesla, Apple, etc, etc, etc, will all fail with driverless and electric, you need a serious reality check.”
Industrial cabal : Google, Amazon, Ford, Toyota, Nikola, GM, etc.
Military cabal : General Dynamics, Lockheed, Boeing, Northrop Grumman, etc.
The same Deep State operatives that keep the US forever on track for more military conflict no matter how badly the public yearns for peace also has an industrial cousin. They will keep the driverless car mission on track no matter how much nobody really wants it. They’ve got the ball rolling. Who cares how many passengers/motorists get killed? That can be covered up as easily as a service member’s remains arriving at Dover AFB in the wee hours of the night.
The MIC wants it, it will happen. Never mind timelines, in fact an ambiguous timetable serves their desires even better. It’s about safety. Safety from the terrorists (military cavaliers gig) or else safety from your hapless neighbor who really doesn’t know how to drive (industrial xabalust gig).
Public transportation is designed to make a profit for itself. It’s job is to deliver workers and customers who use them to make a profit. A more traditional style of crony capitalism.
I don’t know why you believe that. Long freeway stretches through nowhere is by far the easiest parts to automate. Letting the driver have a few hours of shuteye, so he is fresh to take over once things get less predictable.
Like every other development in America, for those who are into that kind of speculation; extreme exurban (XXurban….. reserve the last X for what people may be doing in their self drivers during the commute…..) real estate development will be a beneficiary. The first hour or two of commuting in the morning can be done asleep. And ditto for an after work nap, on the way back to ones white picked fence fantasy….
Next up – the external uterus. No need to for the mess and fuss of raising your own kid. Heck, this can be provided as a government service. After all, I’m sure the leading cause of death of children is parents and the poor parenting decisions. Imagine the lives that could be saved AND your increased productivity because you don’t have some whiny little kid saying “play with me” and “I’m hungry” and “wipe my butt”. Just go to work and hang out in the evenings with your orgasmatron (no dangers form intimacy) – pure bliss.
Nothing’s debuting, it’s an experiment for chrissakes: there’s a driver sitting at the wheel in every car. This is hardly and effective business model. I get the feeling Mish got bored of endlessly predicting the next recession.
business model?
“can sign up to go for a ride in a self-driving minivan. As often as they want. For free.”
must be one of those internet business things – $billions and $billions in valuation with no earnings … but, hey, at least the internet things have some REVENUE.
And the recession? It is coming. And will be worth the wait.
“A Waymo test driver will be behind the wheel at all times,”
…
Where I quit reading.
Put me down for 12+ years before driverless makes a sizable dent in overall driving.
Unfortunately I don’t know how to insert images… but why would anyone be against this future. The driverless minivan.
https://s-media-cache-ak0.pinimg.com/736x/95/cb/65/95cb6505ca097398f903bab8fc424b5d.jpg
well spank my ass and call me sally. It worked!
Sally.
I believe that self drive is coming, my skepticism is when.
I have a saying;
Optimism is born of blissful ignorance
Pessimism is the product of bitter experience.
I’m a Control Systems Engineer.
I’d love to create fully automated systems. When I get to control the parameters, automation is easily achievable. Unfortunately, roads and driving are incredibly complex activities.
As I tell the youthful optimists, 99% in University makes you a legend, but 99% in a control system in the real world is a piece of trash. To get to be a legend, you will need 99.9999% and even then, there will be those who find fault with your abysmal achievements.
I suspect that reliable self driving systems are 5 years away. In controlled environments, it’ll be closer, but in suburbia, 5 years is optimistic!
Cheers from Straya and New Ziland
Mish, I’m pretty certain you have big investments in this driverless technology segment, just the way you so vehemently defend it. It’s obvious you don’t care that millions of Americans will loose their livelihood from this. I’m also convinced that you are a fraud. You have mistakenly identified yourself as an elitist/globalist. I won’t be visiting your site anymore, you piece of shit.
Classy.
Well that escalated quickly
Hitler !
Everyone has big investments in driverless. As well as in any bankster hyped scheme/scam. The Fed makes sure of that.
The only difference being the class of share one receives in return for ones investment. The banksters receive regular shares. The rest, Z-shares; which have the specific property that they let you parttake in the cost, but not in any possible upside. Now shut up, go wave a flag, and be proud that you get to live in a country where banksters can pay advertiser supported media enough to be designated as “visionaries.”
Bye Todd.
“A Waymo test driver will be behind the wheel at all times”
So actually they are doing exactly the same as they have been doing for the last couple of times – except with some poor crash test dummy in the back.
The problem with self-driving cars isn’t dealing with ‘normal’ situations, it’s dealing with the unusual situation, as these are actually quite frequent.
Actually, much as I dislike Tesla (a small US electric car company), I have to grudging admit that they do have *almost* the right strategy for “self-driving” cars. This is because they are actually putting their technology out there on real vehicles, so it has to work at some level. It seems that their automatic emergency breaking may be saving lives, and their predictive warnings/actions about imminent crashes are useful. This is the sweet spot for “self-driving” technology, not replacing the driver completely. Where Tesla falls down is their repeated insistence on fully autonomous modes – despite obvious flaws.
Another repetitive, monotonous task that machines will eventually excel at. It will be huge. I am not looking forward to what the regulators and insurance companies do with it, though.
In an excess of 50 comments, not one about the ability of government to control your sphere of movement–or stasis–in an autonomous vehicle secondary to the current enviromental policy du jour…Ozone action day? Pleases use the bicycle provided in your trunk…Carbon footprint exceeded for the week? Please use the walking shoes provided in your former glove box…Somehow my gut tells me you technocracy types are secret shills for agenda 2030’s smart cities with controls on even how many sheets of T/P are used on a daily basis….for virtually any “environmental” reason your vehicle will not leave the drivway unless government says, for a price of course, that it can…
actually, it was one of the points I was attempting to make but too much hate going on in the discussion. In addition to your very valid point, what about hacking. Hacking will always happen if the target is big or important enough.
Is Uber still going to be valued at 70 billion dollars? Once this technology is common place, car companies will have their own brand of driverless taxis and bypass Uber all together.
Wanna bet against Uber drivers sharing tips about how to make life maximally difficult for the poor robots, that dare infringe on their right to earn enough to afford smoking spleefs on the job?
@JareheadJohn — please read my post again. “I am not looking forward to what the regulators and insurance companies do with it…” That was criticism aimed at continuously expanding government control.
10-4. Good Buddy…
I’ve been driving for 40 years and have had 2 car accidents. The second one was 7 years ago. I was driving south in interstate 95 to Vero Beach with my wife and 12 year old son. I was in the right lane. A car came up on the left and moved into my lane hitting me on the side slightly forward of the driver’s side front door.
I couldn’t go off to the right because of a ditch. And I couldn’t slam on the brakes because my son had taken his seatbelt off to reach for something in the front of the car. So I made a conscious decision to be hit.
Makes me wonder how a autonomous vehicle would have reacted. Would it know a ditch was off to the right? Would it know a passenger would be injured by slamming on the brakes? Real world stuff is very difficult for a machine. Especially a human programmed one.
Shame on you for bringing up a real life example of why the self-driving cars won’t work.
You’re ruining their narrative.
Your comments should be moderated.
Nah – will be used as an example of why all cars should be automated, idea being that the incident would not have occurred if they were.
Bureaucracy, here to make the world what it isn’t.
“Your comments should be moderated.”
Nah. Patton loved it when the Germans shot at him, they told him where they were.
Had the allies lost the war Patton would have been just another war criminal.
The self-driving fanatics living in their fantasy world should keep that in mind.
LFOT: “real life example of why the self-driving cars won’t work”
As someone vaguely familiar with both sides of this (computers and driving), Jon’s scenario is an example of why self-driving cars can’t happen too quickly.
The hard thing for a computer driver in Jon’s scenario is intuiting the intention and likely actions of the guy on the left ahead of time. (Jon’s son, how to brake, and where to steer are handled by something called “planning”. Chess playing logic is used for planning. Old tech, though new optimizations are always welcome.)
When I imagined myself in Jon’s scenario two thoughts arose instantly:
1) Was the guy on the left squeezed in to Jon, say by someone further left, maybe coming the other way? If so, then Jon should have seen the squeeze coming. To do so may have required eyes in the back of his head. Like computer eyes.
2) Did the guy on the left just bozo out? If so, Jon was a normal driver and got squeezed. But, lo, those many years ago, when I took driving very, very seriously, I considered it base-line paying attention to sense a cop car ahead of me within in mile or more. How? Body language and nature of the traffic coming my way. Knowing the guy on the left would veer in to me would have been at duh level. Guys like that are (“were”, not “are”, to be honest) easy to spot.
Now, that level of road awareness I’d not expect in 1st generation computers. Reading human faces from low resolution images ain’t easy. Reading the “body” language of the human’s exoskeleton is probably easier, but … well. Yep. Version 2. For an effective version 1, though, computers don’t have to be road-aware at the level of an ultra-focused kid, working his tail off learning and doing that job.
OK, the left guy’s squeeze play could have come out of nowhere. A triggered jack-in-a-box in a passenger’s lap. So, the computer driver’s planner – the logic that lets it find its way somewhere – needs to be responsive in real time. True, that.
Yes, I was going to make a similar comment but yours is close enough. Also consider what if these cars made decisions based on data about the passengers in other cars around them? Would your car make a more dangerous maneuver in a pending accident to avoid contact with a car full of children next to you? How many people would accept a computer making this decision to put their lives in greater danger even if it the “logical” choice?
And what would a stupid human do, especially when drunk, tired, eating a sandwich, or talking on the phone?
“Car decision” arguments are pathetic, show a complete lack of understanding as to the primary cause of accidents: stupid humans.
Mish
There’s a real dilemma there and assuming the human is drunk or not paying attention doesn’t answer the question.
I were driving and a car was coming at me head on, I would likely side swipe the car full of children next to me to avoid certain death, not knowing for sure who I am putting in danger. But I’m a stupid human. I am just wondering how many people would accept a machine making a decision to put them in a greater danger than necessary because of other factors.
Interesting question: would you rather be killed by a human or a machine?
A: A distinction without a difference?
Your comment gets to the real core of the driverless problem: Will the robots be allowed to play the game of chicken as aggressively and unpredictably as humans? And if not, if the robots are boringly predictable, cautious and conservative; what is to stop human drivers from learning how the robots will respond with a high degree of certainty, and use that cut in front of them with abandon?
Modelling gridlock traffic as some sort of cooperative venture, makes about as much sense as pretending the dinosaurs went extinct after having a debate with the mice, realizing the latter were more efficient utilizers of scarce resources, and having been thusly “educated,” deciding that it was best for life on Gaja, if they simply laid down and went extinct.
According to the 2014 stats on Waymo’s website, it would probably look like this:
Drunk – bad
Eating a sandwich, talking on the phone – not good
Tired – best chance for a good outcome
Speeding and drunk drivers account for more than half of the 2014 fatalities. I would correlate these two categories with stupid, as they are both related to judgement and are 100% avoidable. Tired is less than five percent and does not necessarily indicate stupid.
Car decision arguments are not “pathetic” since we won’t be banning humans from driving cars any time soon. So for a long time there will be both humans and self driving cars sharing the road. This perfect world with no human drivers (and not software bugs… not likely) will not exist for a very long time so these questions are valid.
Jon:
“Would it know a ditch was off to the right?”
One certainly hopes so. It’s the driver, after all!
“Would it know a passenger would be injured by slamming on the brakes?”
Modern cars monitor seat belt clicked-ness. Can you think of a good reason to hide this information from the driver? I can’t. And the driver knows the speed, and will know your son’s weight roughly. I’m guessing the driver and its programmers know Newtonian physics, too.
Reading comments that list “I can’t imagine how a machine can handle …” scenarios is like watching slapstick comedy. Amusing, but painful. Like, the commenter can’t imagine how a piano will fall on him in a moment, so therefore a piano won’t fall on him. We, the audience, laugh despite ourselves.
“Real world stuff is very difficult for a machine.”
Yes! That’s the nub of it, isn’t it? The big jump we’re looking at is computers suddenly becoming free-range – able to handle real world stuff. It’s happening. All around us. Oddly enough, the enormous amount of work put in to self-driving vehicles mutes the effects in other areas. A lot of talent is tied up by the “car” companies. So, in the mean time, your washing machine can tell your phone the clothes are clean (an easy engineering job). But there’s no gizmo in your utility room moving your washed clothes to the dryer and cleaning the dryer’s filter. And your garbage can doesn’t walk itself to the curb on the right day. And, a roving Jiffy Lube doesn’t do your car in the middle of the night when you’re asleep. And no gizmo comes by to clean your gutters or paint your house. And, and, and. 🙂
By the way, any city people here know whether garbage trucks, themselves, are picking up the cans now? Speaking of messy, real world stuff.
As I have mentioned several times before, ethics is an area still under development for autonomous vehicles. When a situation arises where an accident is unavoidable, the vehicle AI may have to make an ethical decision as to who or what to hit: the animal?, the crowd on the sidewalk?, or drive into that wall and possibly injure or kill the vehicle passengers?.
These situations will be pretty rare, but they must be planned for.
In spite of this complication, autonomous vehicles will save millions of lives, and countless accident related injuries.
It will take time for the AI to get better.
you didn’t have an accident you CRASHED
Taking care of family members in their eighty’s it was clear that driving is a problem even if they are otherwise healthy. Vision problems and reaction times are not what they once were. Combined with less ability to hike from the bus stop to their destination means that they either get a ride from someone, get a taxi or uber ,or just stay home.
I can see large increases for some kind of low cost door to door transportation services as baby boomer demographic enters their eighty’s only around 15 years from now. Which technology will be be more affordable an uber or taxi or the self drive car. These beta tests happening now should show us.
This technology, like all technology, will not be perfect. However driverless cars will be replacing human drivers who are not perfect. Impaired drivers, sleep deprived drivers, angry drivers, inattentive drivers, etc cause accidents. So if the technology is only 99% effective it will still be safer than human drivers.
The problem with adoption is human perception of risk. Just as people are more afraid of plane crashes than long distance driving when the former is many times safer, the population will react negatively to every accident where self driving vehicles are involved no matter the cause. I see this as the biggest obstacle to adoption.
Self driving cars are inevitable, and will be first be adopted by taxi companies, uber, and delivery companies. Ironically, seniors who are too old to drive, and drunks who have had their license taken away, will also likely be early adopters.
Mr. Lawson…you are correct that air travel is the safest mode of transportation…That still requires by statute a pilot AND copilot…Good luck putting your trust in a computerized flivver with no back-up safety net…Who needs steering wheels and brake pedals…
.
Hope these things are good at dodging potholes. Some of them will take a wheel off and fling the passenger through the moon roof.
I live in a major metro area and there is one pothole on the various routes I take to work, grocery, church, etc. Do you live in Baghdad?
It will happen and it will happen twice as fast as most people predict. However, it will be extremely frustrating until we get to Level 4.
Just consider all the things that could go wrong with these self-driving vehicles. Look at all the intricate sensors, automated functions and programming that must work in PERFECT sync with one another to get you safely from Point A to Point B. A MINOR malfunction could cost you your very life. You’re car would be in the repair shop more often than on the road!
This whole concept is so crazy. Even under the BEST of circumstances it would take at least half a century to roll a program like this out in a major city. Those who claim we’ll see a massive roll out within 10 years are truly laughable. Either they will go silent on the matter or claim that they’re timetable was off by a little and to be patient.
My view is that it is an attempt to jam a foot in the door.
I can accept Tesla and co. gradually introducing features that help the driver as long as it is shown they don’t encourage the driver to be less responsible. That gives a nice real world lead in over decades to refine the concepts, judge public acceptance, and so on.
However to put it in people’s thoughts and pockets that they should commit NOW to a preconceived reality, and where they stand to lose if they pull out because it does not meet expectations, is not right.
Mish is on the verge of that, but as his blog and views are open to criticism, I take his stance and viewpoint here as a proposition for debate, and it is quite a good debate to have, seeing as this is all going one way or another no matter whether Mish is right or not.
To put it another way – Mish is projecting his beliefs, but in the real world you are going to be pressured by politics and salesmen using hardsell and devious method. If we have not discussed it here, thought about it, we are the more likely to be fooled… not you or I maybe , but those who have never heard the idea questioned before.
So your guess is 50 years at the best case. My best case is widespread use of level five autonomous vehicles both in commercial trucking and private autos within 5 years. My worst case is widespread use in 10 years.
Good thing is we’ll be able to figure out if I’m wrong very soon 🙂
For a reasonable price per trip, in a city, it would be a blessing not own a car. Maybe people might even walk sometimes.
A sampling of human driving from the last several hours or so in Orange County, CA:
http://www.ocregister.com/2017/04/25/one-dead-one-in-critical-condition-after-crash-on-5-freeway-in-orange/
http://www.ocregister.com/2017/04/22/19-year-old-motorcyclist-killed-in-anaheim-crash/
http://www.ocregister.com/2017/04/20/bicyclist-hurt-after-collision-with-newport-beach-police-car/
http://www.ocregister.com/2017/04/18/mother-and-son-escape-serious-injury-after-vehicle-overturns-on-5-freeway/
http://www.ocregister.com/2017/04/20/truck-drags-car-wont-stop-and-its-all-on-video-5-2/
http://www.ocregister.com/2017/04/19/1-killed-in-ortega-highway-crash-with-garbage-truck-versus-prius/
OK, that’s enough. It’s amazing how diverse these “accidents” are. This is one county in America over the course of one week, and only those “accidents” that were serious enough to make the newspaper.
I have concluded that, while perhaps not evil, Luddites definitely hate other people. Why else would they want the present homicidal regime to perpetuate?
1 million deaths a year worldwide. 30,000 deaths in the US alone and tens of thousands of life altering accident injuries. Thousands of hours of legal work dealing with settlements for those injuries.
Couple the above with the millions of people who are sitting at home right now with no way to get where they want to go, or those paying thousands yearly for transportation services because they are unable. Elderly and disabled people who would like more time with friends and relatives but are unable to conveniently travel to them.
But really, there is no demand at all for self-driving cars. Nobody in their right mind would want to improve if not solve the problems above.
Glad to be in agreement with you RF. Thanks for bringing sanity here.
James – glad to have a like mind commenting as well! It’s a topic I’ve unexpectedly become very interested in and passionate about.
America can save 30000 lives per year and many hundreds of thousands of injuries with autonomous vehicles. Yet some on this site say that is a small price to pay to keep their freedom. What does that mean? You want the freedom to kill or maim someone while you are driving?
The new self-driving (Auto auto) cars are here. It’s proliferation time now. But they’re not for the biological “me,” though by voluntary proxy “I” will be transported if I have my credits in order — if we consent to the parameters of our fiction of “freedom.” There’s no love or altruism here. There is no sympathy for you and me — only the devil, and, the master is here to more efficiently manage in the technocratic/technotronic age your legal “person” in an yet to be integrated age of cashless payment systems, carbon credits, and electronic surveillance trough the internet of things.
I saw one of these Waymo taxis on interstate 10 just yesterday
Would I prefer a Ferrari or a chauffeur driven limousine? No contest.
Would I prefer crawling in rush hour traffic or reading on a rushing train? No contest.
My guess is that these self-driving fanatics living in their fantasy world all have bidet equipped toilets in their homes.
Maybe in the future we can design toilets to come to us as opposed to us going to them. It could save us time and prevent falls in the bathroom. Clap once and turn on your lights. Clap twice and your motorized toilet is at your side for service.
I love this brave new world. How ’bout you?
It will be fun to read the comments (if any) from the optimistic generation x kids (and their adult cousins) in 10 years when it all becomes clear that this funny venture of the computer nerds away from their profession has failed.
Proposals for safer traffic:
Alco-locks in every car.
Speed limiter programmed into every car by law.
Increased police controls, lazer speed controls,
even automate THIS.
Currently the penalties for speeding and DUI are monetary in most areas of the US. If we were serious about reducing the carnage from these causes until autonomous vehicles are prime time, a different penalty should be employed – loss of the vehicle involved in the infraction. In case of a fatality, the guilty party’s loss would be permanent. For lesser infractions, varying times could be used related to the severity.
If they are forced on us as they will be there will be no fun driving anymore, driving licences will be banned humans too dangerous to be behind the wheel. Welcome to the boring world of the future
All that free time we will have to watch Fake news on its internal screen, or use the interweb to use Fakebook and Fake Google, cant wait
I’ll get a LOT more excited about self-driving cars the day they announce this demo is being run without check-drivers in the front seat. Until then, it’s just another demo.
That’s when everyone else will realize they have been dead wrong for a long time
Ha. I laugh at them. Even I’m on autonomous vehicle project. People don’t trust. Haha. I don’t care. I did care if they would have invested in my project. Just kidding. Keep going. Keep writing new posts on self driving technology.
And the most important thing is, live well and fully.