I just don't see it. I mean these three things aren't "everything you know."
Even then I have a major issue with the concept that people are going to be reading Ulysses, or watching a movie on the drive. Car sickness is a real thing and affects a sizeable enough portion of the population that it's not going to overhaul entertainment.
I think what it is going to do is make people who drive for a living obsolete and have better flexibility in transportation for the working class and poor, making it into a better version of the bus.
There will also be people that have a good reason to own cars - like if you have a dog, or a baby (so you essentially keep baby stuff in your car nearby), or just need to lug large amounts of stuff around that you need handy.
That said, I do believe - if the technology proved successful - it will have an enormous impact.
The car sickness is good. People can read in buses sometimes, so it might not be impossible. But if every self-driving car drove like a bus, the non-self-drivers might get very frustrated.
Such a powerful sense of Deja-vu. I'm old enough to remember in 1998 when everyone was saying "The Internet will effect everything you know" and yet lots of people would say "Really? I get along fine without email, I just use a FAX machine. And everything I want to buy is right down at the mall or in one of the mail order catalogs on my shelf. Internet hah! All that so I can get a chain letter of jokes I learned in middle school. No thanks."
It's true that self-driving cars will/would change a huge amount. But the article seems to make a lot of assumptions that may or may not be warranted. Will this facilitate population spread or population density? I think there are a lot of paradoxical factors which may come up.
One change someone pointed out here a while back is self-driving-ability increases the willingness of people to idle their cars in traffic. It stands to reason that an owner would have little disincentive to send their car out to idle in traffic and pick them up a few hours later, if it was convenient. It seems the final result would be gridlock.
Roads today are a free or low priced good whose use is limited by the ability by each person's ability and willingness to drive. Driverless cars could consume this good with fewer limits creating a "tragedy of the commons" type situation.
So it seems like either you will have regulation or you will have an unregulated situation where traffic grinds to even more of a halt than it is at today in major metropolitan areas.
If you have regulation, you may find eventually everyone forced into an Uber/Lyfte situation (self-driving cars become smart-buses) since that would optimize the use of street space. This would facilitate density imo.
If you have no regulation, you might find some folks going further out despite the time and because of the not having to do their driving. But that seems hard to know.
And generally, the change would be big enough that indeed predicting all the results seems hard. What level of autonomy is achieved when is a big factor, etc.
> self-driving-ability increases the willingness of people to idle their cars in traffic
I disagree with this premise. There is a cost to idling in traffic: electricity/gas expended while idling, and the opportunity cost of recharging/refilling before going home, if your car idled too much. Reducto ad absurdum: there will always be a need for airport parking lots.
If there is income to be made, Uber-style, from idling your autonomous car, effectively decreasing the costs of idling, then it's easy enough for the municipality to tax income from services like Uber untill some equilibrium is reached, whereby most people are incentivized, like today, to take high-throughput public transit. Everything old is new again: particularly, taxi medallions.
The availability of self driving cars will significantly remove some of the biggest disincentives to car use, such as licensing and parking costs. The immediate impact of this would be a stronger incentive to drive a car over other alternatives and more cars on the road.
This change in balance between transportation options will have significant negative impacts on cities, many of which are already actively engaged in encouraging people to choose other transportation options than the car.
A collision between self driving car share companies and city governments seems likely. As self driving cars become more prevalent, we may see cities enact more anti-congestion regulations as a result.
If self-driving vehicles become common, there is no real reason to own a vehicle besides extremely heavy utilisation or aesthetics. A public grid of transport is a natural outcome of removing the transport driver bottleneck. Heavily travelled routes will see buses and mass options dynamically (why bother with routes and schedules?) and requests not efficiently served could be handled by unit transport. One might pay extra for direct per-unit transport or pay less for hub-spoke transfer routing. The options are quite intriguing.
There are already examples of routeless public transport, which is in third world countries with private, relatively poor quality mini bus based public transport. As transit systems become more developed they graduate toward fixed routes with larger capacity buses because it is the most efficient way of transporting people. This is why we've seen Lyft move toward this as well (reinventing the bus).
It will be interesting to see how autonomous vehicles change public transit as the technology develops, but I would not be surprised if the utopian vision of public transit that delivers the user to their exact destination never becomes a reality.
It's sadly likely to be true that, unless someone finds some way to implement corruption features in such a system, that it is unlikely to replace taxi, private and mass transport. However, one can dream.
Ah, but why is it more efficient? Sure, there's some energy savings, but not that much. The main reason is having fewer drivers per passenger. But when there are no drivers, that doesn't apply.
Who cares if it takes time, if car time becomes cheap? That's the whole point. Time is not a cost in itself, the cost comes from the resources being spent during that time. If the resources/time ratio drops, the total cost remains the same even as you expand the time used.
Self-driving cars will have to pass through the stages of rare, occasional, common and universal and each of these is going to have an impact.
Especially, Americans have billions of dollars invested in their cars and many them cannot afford to buy a new car at a moments notice. I hate driving but I have a car because I live an area where it's necessary. Once you own a car, each drive is fairly cheap (cheaper than public transit, sadly) because you're effectively expanding money you've spent on your car. So a lot of people, especially poorer people, will keep the cars they have unless Lyft/Uber/something gets a whole cheaper.
Maybe the final outcome you describe will happen but we'll have to survive the process of transition first, process that may a change a lot before it finishes.
David Dudley (CityLab) offers a unique perspective worth considering:
"A decade hence, we might find ourselves even more beholden to automotive infrastructure, as city streets swell with the dread “zero-occupancy” vehicles—empty cars roaming about running errands for their affluent owners. And those left behind by the first car revolution—those whose paychecks are already consumed by automotive expenses or who rely on increasingly shaky public transit networks—will be even more hopelessly marooned in the mobility deserts that remain." [0]
I'm growing more concerned that self-driving cars will actually increase traffic. There is an equilibrium found based on a driver's willingness to drive. Once we eliminate that need for a driver, that equilibrium disappears.
For example, if I have a sudden urge for ice cream and I see that my freezer is empty, I have to make a decision whether it's worth driving all the way to the grocery store just to buy a single item. Suppose I decide, no.. it's not worth the hassle.
Now introduce an app that allows me to send an autonomous car to the store, have the employee fulfill my order, and the car brings back my ice cream. What you'll have is a horrible gridlock of people using these cars to buy single items at a time, out of impulse. Before, the hassle of me having to drive to the store stifled that impulse. Without having to drive to get the items I want, I no longer have stake in that matter. Who cares if traffic has gotten worse? I'm not in it.
[Edit - add another scenario:]
Consider another scenario. My friend wants to borrow an item X from me. He doesn't need it immediately, so I may decide to drop it off at his house next week when I drive over to visit. Now with a self-driving car, heck... why not just send the item X over to him now? It's no hassle to me since I don't have to drive, and he gets it sooner. But now I've just placed more burden on our roads and our environment.
This also spells a bad scenario for climate change, if it increases the amount of vehicle-miles traveled and CO2 emissions.
While it's possible for autonomous vehicles to reduce sprawl by eliminating vast parking lots and road space, I think a more plausible scenario is that people use self-driving cars as their own personal butler, picking up items here and there from stores, increasing the strain on our roads. After all, if you had an app on your phone that allowed you to order a vehicle to pick up items from a store, all in the comfort of your own home, wouldn't you make a butler out of this car?
[Edit]
Perhaps what self-driving cars will enable is a plethora of what amounts to useless and wasteful voyages.
Smart consolidation will happen. E.g. People with a similar route will share the vehicle retrieving the item, or a single designated car will pick up multiple items from that store and intelligently deliver..
That's how things currently work with deliveries, so personally I'd guess that would continue and delivery companies probably have a greater push to optimise costs than you do personally.
I agree that at least in the short term traffic may increase a lot. Long commutes may become more acceptable too. I hope that sending your car somewhere will not be too cheap or people may send it on all kinds of not really needed trips.
I can say this: if my car is now going to be throwing ads at me as well, I'm going off grid. I have enough information pollution that I have to deal with.
> I can say this: if my car is now going to be throwing ads at me as well, I'm going off grid
You are already the frog slowly boiling in the pot, and you have not noticed it has already gotten extremely hot.
I live in the Yukon, and spend weeks at a time in the wilderness near the Article circle in Yukon and Alaska. I do that for a year or two, then maybe take a flight to Vancouver. Before I even leave the airport I am utterly dumbfounded at the amount of advertising people put up with daily. And phone use. And restaurants, and well, basically just everything.
Sometimes I think it's healthy to come back to "the world" occasionally, other times I say good riddance.
There is nothing inherently wrong with the modern world, including smartphone usage. The fact that you don't choose to use one (or as often, or whatever) doesn't make it wrong that I do.
My phone is an iPhone, and I make great effort to only use apps that are paid and without ads. You get a ton of utility usage out of these things (and yes, they help you kill time waiting for your doctor or your train too) and the only way it's a billboard constantly shouting bullshit into your face is if you allow it.
I do agree though, the amount of advertising people are subject to is egregious. I have a feeling that's a bubble that will be bursting soon.
"It’s estimated that there are as much as 2 billion parking spaces in the US alone and a third of them are in parking lots! That’s over 42 billion square feet of space..."
That seems to be off by about a factor of 5 - or it is too late for me to do math. The article seems to be using 21 sqft per space. A typical space might be 8'x14' or 112 sqft each.
If we can eliminate some parking spots that will free up alot of space for something else.
That's roughly correct for on-street parking. In a parking lo or garage, the parking space + associated room for lanes and turnarounds works adds up to about 330 sqft per space.
> In an ideal world, where no one owns a car, everyone requests a ride through their phone. This means that the ride-sharing companies would own the cars themselve
Who said people want such an disruption? They want electric cars. They want self-driving cars. They want to own OR lease OR rent cars. Have you missed the "own car" portion? I for one want to own my car, as I can do the math, it's cheaper and I can sleep in peace. This smearing campaigns to lobby for change society to take them away cars and promote "cars as a service" (aka rent) as only option is very disgusting. I know it's their wet dream - just go away.
Not sure who are "they" but in the end its all about cost, if renting cars is cheaper and its good enough for most of your work then people will go with the cheaper option.
The reason you have such extreme position is because you grew up with having your own car, in the future there will be a generation who will grow up with the renting option and they will ask the question why do you need the hassles of owing your own car.
You're basing your math on the state of cars today. This will change drastically. A self driving fleet will be very cheap to order compared to current taxis or ride sharing services.
> it won’t make sense for people to own cars in the future because ride-sharing apps like Lyft and Uber will be our go-to when it comes to moving from point A to point B.
Is there anyone left who hasn't heard this Silicon Valley cliche ?
Even outside the SF, most people own bicycle and cars to get around. I lived in Oakland for 5 years and rarely took Lyft (never used Uber for known reasons). I can count a lot of reasons to own a car, for example, you would never call Lyft to go to a waterfall or visiting Yosemite.
If the case is owning car only for navigating in the city, bicycle combined with good public transportation solves that problem already in a much better and healthier way.
Economics will drive this. The car is the most, or second most, expensive investment most people make, yet it has anutilizatio rate of only a few percent. And big corporation s can borrow at a much more preferential rate (prime, basically).
You could say the say about renting vs buying houses, and yet many people still buy. The only reason a rental service is funded is because investors think they can make more money than by selling mortgages. This will continue to be true for cars.
> You could say the say about renting vs buying houses, and yet many people still buy
Cars and houses are excludable. And both are, within reasonable bounds, rival [1] when in use.
But people customise their houses in ways that are difficult to transmit. That makes houses more rival when out of use than a car. That is to say, people are less comfortable staying in others' houses than being in others' cars. Similarly, people are generally more okay with others being in their car than in their house. A rough estimate of this delta can be gleaned from comparing ride-sharing and Airbnb adoption rates.
There is a big difference between cars and houses, which is that houses tend to gain value over time, whereas cars drastically decrease in value. Driving is also time consuming and stressful. Dealing with maintenance, licensing, refueling, etc. are also annoying burdens.
That's funny to say. I had a Subaru Outback (2004) that I paid $2000, and it was clean and strong, worked perfectly until I left US.
If you take Lyft or Uber to get around, you'll spend $2000 in max. 3 months. May be it'll get cheaper in the future, and may be some people will prefer within the city, but it's still not a better option than good quality public transportation.
Zipcar is the worst choice for roadtrips. Have you ever tried in your own ? First, it charges per hour, second, it charges extra after 200 miles outside the city.
For going to a long weekend trip, you'll end up paying about $500 for three days. This is only considerable choice in cities like San Francisco and NYC.
About five years I thought of buying up some parking lots in manhattan figuring I could make a killing when AVs made it unnecessary. The economics were unfortunately out of my reach (and in many cases the air rights had been sold off).
I looked at some other places like St. Louis but the long term economic prospects werent encouraging.
It gets used for something more valuable than parking. (One might question that premise; unless there were zoning reasons the land had to be used for parking, you'd expect the free market to have already found the most valuable use for it.)
I've stopped thinking of them as cars. They're "thing movers".
Consider how much of the speed, size and load constraints of a lorry are driven by the needs of the tiny human at the front. Self-driving trucks make possible a slow-moving vehicle trucking super-heavy stuff across the country at cheap rates because it only operates at the most-efficient speed its engine drives at. Or very-small trucks. Or cars that seat configurations and numbers of people that don't make sense given the constraints imposed by a human driver (who must be entertained or paid for their time).
This person barely touches on what seems like the biggest problem - autonomous cars (especially electric ones) could make 100-mile commutes the norm, massively speeding the conversion of agricultural and wild land in to housing and asphalt, and further destroying this already rather strained planet. All of the harm wrought with automobiles will have the two things tempering it (monetary and time cost) removed.
Oh, and they're hell on Earth for walking, cycling, motorcycling, or vibrant, human-scale neighborhoods.
I'll first state my opinion that I too believe that (to use a different cliche) self-driving cars will be transformative.
That being said, I just don't think they're as close as many think.
What is coming and will continue coming is assisted driving. These cars will be designed for human drivers but the car will just do more of the driving. This will have an incremental increase in safety. I imagine too that bugs will cause deaths and even though this will pale in comparison to the number of motor vehicle deaths we have now, alarmist concerns will stymy automation here.
The holy grail here however isn't assisted driving. It's self-driving and you shouldn't confuse the two. A self-driving car in my mind will not be designed to be driven by a human at all.
There is a lot of low hanging fruit for this but the low hanging fruit isn't that low and there is a long tail of situations a car will need to deal with. I actually wonder if we can design a self driving car without first creating what I can only describe as a "general AI". Anticipating what the other meat bags will do is a huge part of this.
Once this happens (and, again, I think we're talking a long time still) I don't see private ownership of vehicles going away. People like to own things. What you'll see is car ownership disappear in highly urbanized places helped I'm sure by punitive taxing of car ownership ("punitive" here means not subsidizing car ownership like we do now).
I suspect such cars will allow people to live even further from cities when you can work in your car. More people doing it still means a lot of cars on the road.
Self-driving cars just aren't the end of private car ownership. If anything they'll likely be a death blow to taxis. It might be a huge problem for mass transit too.
IMO, what has even more transformative potential is human-carrying drones. Flying, self-driving taxis. That could really shorten commute time, and allow people to live very far from work. You could easily live 100KM away, and make that commute in 30 minutes. It would completely change urbanization and population density. It's maybe not that far away either. Self-driving air vehicles is in many ways simpler than making something that can understand driving on the road and interacting with human drivers. The main limitation right now is that we aren't quite there yet on battery tech, we maybe need a factor of 2 or 3 improvement in power density for this technology to become inevitable.
> The main limitation right now is that we aren't quite there yet on battery tech
That might be the main tech limitation (don't agree with that either because flight coordination is actually a bigger problem.
The actual MAIN limitation is that in an urban area (where 95% of jobs are) no resident is going to stand for helicopter noise by thousands of commuters at all hours.
And creating a drone/copter that is quiet while transporting people-sized payloads is a much bigger challenge than creating one which is fuel efficient.
Once we have the technology for level 5 automation we'll probably sensibility arrive at the conclusion that it makes little sense to ferry people across geographies so they can share information in the same room when we already have pretty good technology for doing that without everyone being in the same room.
It took until I became a private tutor before I realised just how difficult some things are to do remotely. There are many subtle cues I pick up from a tutee in person which are either not conveyed, or are really hard to pick up, when the person is on the other side of a screen. Probably mitigated by larger and higher-fidelity telepresence rooms, admittedly, which might have been where you were going with this.
When I read this, I can't help but think of the enormous promises that came with the computerization of medical records versus the reality that we have today.
55 comments
[ 1.8 ms ] story [ 124 ms ] threadEven then I have a major issue with the concept that people are going to be reading Ulysses, or watching a movie on the drive. Car sickness is a real thing and affects a sizeable enough portion of the population that it's not going to overhaul entertainment.
I think what it is going to do is make people who drive for a living obsolete and have better flexibility in transportation for the working class and poor, making it into a better version of the bus.
That said, I do believe - if the technology proved successful - it will have an enormous impact.
There are a lot of huge question marks.
It's true that self-driving cars will/would change a huge amount. But the article seems to make a lot of assumptions that may or may not be warranted. Will this facilitate population spread or population density? I think there are a lot of paradoxical factors which may come up.
One change someone pointed out here a while back is self-driving-ability increases the willingness of people to idle their cars in traffic. It stands to reason that an owner would have little disincentive to send their car out to idle in traffic and pick them up a few hours later, if it was convenient. It seems the final result would be gridlock.
Roads today are a free or low priced good whose use is limited by the ability by each person's ability and willingness to drive. Driverless cars could consume this good with fewer limits creating a "tragedy of the commons" type situation.
So it seems like either you will have regulation or you will have an unregulated situation where traffic grinds to even more of a halt than it is at today in major metropolitan areas.
If you have regulation, you may find eventually everyone forced into an Uber/Lyfte situation (self-driving cars become smart-buses) since that would optimize the use of street space. This would facilitate density imo.
If you have no regulation, you might find some folks going further out despite the time and because of the not having to do their driving. But that seems hard to know.
And generally, the change would be big enough that indeed predicting all the results seems hard. What level of autonomy is achieved when is a big factor, etc.
I disagree with this premise. There is a cost to idling in traffic: electricity/gas expended while idling, and the opportunity cost of recharging/refilling before going home, if your car idled too much. Reducto ad absurdum: there will always be a need for airport parking lots.
If there is income to be made, Uber-style, from idling your autonomous car, effectively decreasing the costs of idling, then it's easy enough for the municipality to tax income from services like Uber untill some equilibrium is reached, whereby most people are incentivized, like today, to take high-throughput public transit. Everything old is new again: particularly, taxi medallions.
This change in balance between transportation options will have significant negative impacts on cities, many of which are already actively engaged in encouraging people to choose other transportation options than the car.
A collision between self driving car share companies and city governments seems likely. As self driving cars become more prevalent, we may see cities enact more anti-congestion regulations as a result.
It will be interesting to see how autonomous vehicles change public transit as the technology develops, but I would not be surprised if the utopian vision of public transit that delivers the user to their exact destination never becomes a reality.
People seem to think algorithms and computers will change this fact, but this hasn't appeared yet and is unlikely to.
Self-driving cars will have to pass through the stages of rare, occasional, common and universal and each of these is going to have an impact.
Especially, Americans have billions of dollars invested in their cars and many them cannot afford to buy a new car at a moments notice. I hate driving but I have a car because I live an area where it's necessary. Once you own a car, each drive is fairly cheap (cheaper than public transit, sadly) because you're effectively expanding money you've spent on your car. So a lot of people, especially poorer people, will keep the cars they have unless Lyft/Uber/something gets a whole cheaper.
Maybe the final outcome you describe will happen but we'll have to survive the process of transition first, process that may a change a lot before it finishes.
"A decade hence, we might find ourselves even more beholden to automotive infrastructure, as city streets swell with the dread “zero-occupancy” vehicles—empty cars roaming about running errands for their affluent owners. And those left behind by the first car revolution—those whose paychecks are already consumed by automotive expenses or who rely on increasingly shaky public transit networks—will be even more hopelessly marooned in the mobility deserts that remain." [0]
I'm growing more concerned that self-driving cars will actually increase traffic. There is an equilibrium found based on a driver's willingness to drive. Once we eliminate that need for a driver, that equilibrium disappears.
For example, if I have a sudden urge for ice cream and I see that my freezer is empty, I have to make a decision whether it's worth driving all the way to the grocery store just to buy a single item. Suppose I decide, no.. it's not worth the hassle.
Now introduce an app that allows me to send an autonomous car to the store, have the employee fulfill my order, and the car brings back my ice cream. What you'll have is a horrible gridlock of people using these cars to buy single items at a time, out of impulse. Before, the hassle of me having to drive to the store stifled that impulse. Without having to drive to get the items I want, I no longer have stake in that matter. Who cares if traffic has gotten worse? I'm not in it.
[Edit - add another scenario:] Consider another scenario. My friend wants to borrow an item X from me. He doesn't need it immediately, so I may decide to drop it off at his house next week when I drive over to visit. Now with a self-driving car, heck... why not just send the item X over to him now? It's no hassle to me since I don't have to drive, and he gets it sooner. But now I've just placed more burden on our roads and our environment.
This also spells a bad scenario for climate change, if it increases the amount of vehicle-miles traveled and CO2 emissions.
While it's possible for autonomous vehicles to reduce sprawl by eliminating vast parking lots and road space, I think a more plausible scenario is that people use self-driving cars as their own personal butler, picking up items here and there from stores, increasing the strain on our roads. After all, if you had an app on your phone that allowed you to order a vehicle to pick up items from a store, all in the comfort of your own home, wouldn't you make a butler out of this car?
[Edit] Perhaps what self-driving cars will enable is a plethora of what amounts to useless and wasteful voyages.
[0] http://www.politico.com/magazine/story/2017/06/30/biggest-ch... (the section titled "Self-driving cars")
You are already the frog slowly boiling in the pot, and you have not noticed it has already gotten extremely hot.
I live in the Yukon, and spend weeks at a time in the wilderness near the Article circle in Yukon and Alaska. I do that for a year or two, then maybe take a flight to Vancouver. Before I even leave the airport I am utterly dumbfounded at the amount of advertising people put up with daily. And phone use. And restaurants, and well, basically just everything.
Sometimes I think it's healthy to come back to "the world" occasionally, other times I say good riddance.
My phone is an iPhone, and I make great effort to only use apps that are paid and without ads. You get a ton of utility usage out of these things (and yes, they help you kill time waiting for your doctor or your train too) and the only way it's a billboard constantly shouting bullshit into your face is if you allow it.
I do agree though, the amount of advertising people are subject to is egregious. I have a feeling that's a bubble that will be bursting soon.
That seems to be off by about a factor of 5 - or it is too late for me to do math. The article seems to be using 21 sqft per space. A typical space might be 8'x14' or 112 sqft each.
If we can eliminate some parking spots that will free up alot of space for something else.
It's vaporware. People are getting themselves tied up into knots over vaporware.
Who said people want such an disruption? They want electric cars. They want self-driving cars. They want to own OR lease OR rent cars. Have you missed the "own car" portion? I for one want to own my car, as I can do the math, it's cheaper and I can sleep in peace. This smearing campaigns to lobby for change society to take them away cars and promote "cars as a service" (aka rent) as only option is very disgusting. I know it's their wet dream - just go away.
Not sure who are "they" but in the end its all about cost, if renting cars is cheaper and its good enough for most of your work then people will go with the cheaper option.
The reason you have such extreme position is because you grew up with having your own car, in the future there will be a generation who will grow up with the renting option and they will ask the question why do you need the hassles of owing your own car.
You're basing your math on the state of cars today. This will change drastically. A self driving fleet will be very cheap to order compared to current taxis or ride sharing services.
Is there anyone left who hasn't heard this Silicon Valley cliche ?
Even outside the SF, most people own bicycle and cars to get around. I lived in Oakland for 5 years and rarely took Lyft (never used Uber for known reasons). I can count a lot of reasons to own a car, for example, you would never call Lyft to go to a waterfall or visiting Yosemite.
If the case is owning car only for navigating in the city, bicycle combined with good public transportation solves that problem already in a much better and healthier way.
Cars and houses are excludable. And both are, within reasonable bounds, rival [1] when in use.
But people customise their houses in ways that are difficult to transmit. That makes houses more rival when out of use than a car. That is to say, people are less comfortable staying in others' houses than being in others' cars. Similarly, people are generally more okay with others being in their car than in their house. A rough estimate of this delta can be gleaned from comparing ride-sharing and Airbnb adoption rates.
[1] https://en.m.wikipedia.org/wiki/Rivalry_(economics)
If you take Lyft or Uber to get around, you'll spend $2000 in max. 3 months. May be it'll get cheaper in the future, and may be some people will prefer within the city, but it's still not a better option than good quality public transportation.
This it's exactly the type of infrequent trip many of my friends in NYC use Zipcar for.
For going to a long weekend trip, you'll end up paying about $500 for three days. This is only considerable choice in cities like San Francisco and NYC.
I looked at some other places like St. Louis but the long term economic prospects werent encouraging.
Consider how much of the speed, size and load constraints of a lorry are driven by the needs of the tiny human at the front. Self-driving trucks make possible a slow-moving vehicle trucking super-heavy stuff across the country at cheap rates because it only operates at the most-efficient speed its engine drives at. Or very-small trucks. Or cars that seat configurations and numbers of people that don't make sense given the constraints imposed by a human driver (who must be entertained or paid for their time).
Oh, and they're hell on Earth for walking, cycling, motorcycling, or vibrant, human-scale neighborhoods.
That being said, I just don't think they're as close as many think.
What is coming and will continue coming is assisted driving. These cars will be designed for human drivers but the car will just do more of the driving. This will have an incremental increase in safety. I imagine too that bugs will cause deaths and even though this will pale in comparison to the number of motor vehicle deaths we have now, alarmist concerns will stymy automation here.
The holy grail here however isn't assisted driving. It's self-driving and you shouldn't confuse the two. A self-driving car in my mind will not be designed to be driven by a human at all.
There is a lot of low hanging fruit for this but the low hanging fruit isn't that low and there is a long tail of situations a car will need to deal with. I actually wonder if we can design a self driving car without first creating what I can only describe as a "general AI". Anticipating what the other meat bags will do is a huge part of this.
Once this happens (and, again, I think we're talking a long time still) I don't see private ownership of vehicles going away. People like to own things. What you'll see is car ownership disappear in highly urbanized places helped I'm sure by punitive taxing of car ownership ("punitive" here means not subsidizing car ownership like we do now).
I suspect such cars will allow people to live even further from cities when you can work in your car. More people doing it still means a lot of cars on the road.
Self-driving cars just aren't the end of private car ownership. If anything they'll likely be a death blow to taxis. It might be a huge problem for mass transit too.
That might be the main tech limitation (don't agree with that either because flight coordination is actually a bigger problem.
The actual MAIN limitation is that in an urban area (where 95% of jobs are) no resident is going to stand for helicopter noise by thousands of commuters at all hours. And creating a drone/copter that is quiet while transporting people-sized payloads is a much bigger challenge than creating one which is fuel efficient.