Choose Your Own Utopia: What Will We Make of Driverless Cars?

Cross-posted from the Frontier Group

A century ago, a new transportation technology burst onto the scene that threatened to disrupt everything: the car.

Thinkers of the day, along with boosters of the new technology, dreamed grand dreams of the utopia it would bring. General Motors’ Futurama exhibit at the 1939 World’s Fair (shown in the amazing 1940 promotional film, To New Horizons) envisioned a nation criss-crossed by broad highways engineered for “safety – safety with increased speed”; American cities that were “replanned around the highly developed, modern traffic system”; and a network of urban express highways with rights of way “so routed so as to displace outmoded business sections and undesirable slum areas whenever possible.”

Sound familiar? The vision of the future dreamed up by General Motors largely came to pass… but utopia did not follow. Missing from Futurama, as from most utopian visions, was a full understanding of the trade-offs involved — the gutting of city after city for the construction of urban freeways; the expenditure of trillions of dollars over the last half-century on the highway system; the loss of roughly a million lives to motor vehicle crashes in the U.S. since 1990 alone (so much for “safety with speed”); environmental degradation and public health damage from vehicle exhaust and fossil fuel production — the list goes on and on.

The utopia of Futurama was one the United States willingly strove to implement through public policy. The data presented in this story in CityLab by Richard Florida and Aria Bendix show that, while cities and countries around the world embraced the car to some degree and continue to do so, no one pushed as many chips to the center of the table in a big bet on automobility as the United States. The technology available to those countries was the same; the role it was given within society differed.

This history is important to remember as we consider the next disruptive technology to transform transportation: autonomous vehicles. David Roberts at Vox has a fantastic piece up on the “transformative potential of self-driving cars” and his is a utopia worth aspiring to – one in which autonomous, electric cars reduce the amount of space in our cities devoted to cars and parking, play nicely with bicycles and pedestrians, and allow such travel as does occur to take place with a minimum of environmental impact.

But it is critical to understand that neither Roberts’ utopian vision, nor the various dystopian visions about the future of driverless cars, are foreordained. The form autonomous vehicles and the cities that surround them will take in the future is likely to be driven as much by public policy as by technology.

Autonomous vehicles can be many things. They can be safe for their users, safe for other road users (including pedestrians and cyclists), comfortable, fast, efficient, affordable, and environmentally sensitive. It is unlikely that they can be all of those things to everyone all of the time. A Google car that stops safely for every pedestrian in harm’s way is unlikely to get through a city very fast – and if it is not very fast, who will choose to use it? Nor is a vehicle that starts and stops a great deal likely to be a place where everyone will feel comfortable watching Netflix or catching up on texts, as the emerging literature on autonomous vehicles and motion sickness suggests.

There is also no consensus as to which of these values – safety, speed, comfort, environmental impact — is likely to predominate when the time comes for trade-offs to be made. On one hand, you have Roberts’ vision of low-speed autonomous vehicles traveling in harmony with pedestrians and cyclists on shape-shifting streets that perfectly balance everyone’s needs while preserving safety. At the other extreme, you have visions like this one in which autonomous vehicles whip through city centers at high rates of speed with nary a cyclist or pedestrian in sight.

Who will strike the balance among these values and imperatives? Often, the question is raised as one of ethics — if an autonomous vehicle is forced to choose between an action that will harm its occupant or one that will harm, say, a child wandering into the street, what will it choose to do? But really, it is a question of law, and questions of law are ultimately settled — for better or worse — through the political system.

At a forum Monday on autonomous vehicles at the Shared Use Mobility Summit in Chicago, Lauren Isaac of Parsons Brinckerhoff noted that the difference between an autonomous vehicle utopia and an autonomous vehicle nightmare is one of smart public policy. But the same is true of our current, non-autonomous vehicle system. We already, for example, have many proven “technologies” for keeping pedestrians and cyclists safe in our cities – low, aggressively enforced speed limits, traffic calming techniques, protected bike and walking paths, etc. We just simply choose not to use them all that much.

We have lived with the automobile so long that we forget that the rules surrounding its use in society were determined by people motivated by a particular set of values and self-interests. We live with the legacy of their decisions every day, as Roberts’ concerns about the danger of autonomous vehicles falling victim to path dependence illustrate.

I am an optimist in that I believe that a vision for autonomous vehicles like the one Roberts describes is possible – a future in which autonomous vehicles can be used to serve the goals of health, safety, environmental protection, efficiency with regard to the use of public funds and natural resources, and sustainable economic prosperity. But, that future is not inevitable. To get there, we first must articulate a compelling vision (Roberts’ is a good start), be clear about our values, and then lay out a pathway for how those values can be supported through smart and judicious use of public policy.

The arrival of the car in the early 20th century led to the creation of countless laws to regulate, encourage and (in some cases) virtually require its use. The policy debate around autonomous vehicles provides an opportunity to revisit many of those decisions. If the AV revolution achieves nothing more than that, it may still be of great value. But there is potential to achieve more – much more. Articulating our vision of what utopia might look like is a first step toward making it a reality.

88 thoughts on Choose Your Own Utopia: What Will We Make of Driverless Cars?

  1. I think it’s probably a lot more complicated than that. Yes, some morons may purchase an F-150 for the reasons you say but it’s likely also because they need a vehicle anyway, so they might as well purchase one which compensates for the small size of a part of their anatomy. Now if the need to purchase a vehicle vanishes, then whatever vehicle they may be riding in isn’t theirs, hence it’s not associated with their “personality”. In that case, it’s not really unmanly being seen in a minivan. It just happened to be the first vehicle which answered your electronic hail.

    Yes, people my rent robotrucks, but I’d say most of the time it will be because they just need the capabilities. The hire price per mile will more or less ensure that. So will the lower speeds. I’ve repeated advocated much higher highway speed limits mainly because it would help get rid of SUVs. They wouldn’t be cost effective to drive at 125 mph, if indeed they could even reach those speeds. It will be even more so with robocars where a person renting a truck to show off will pay for it twice—once financially, and again by taking much longer to get where they’re going simply because the truck is unstable at speeds small, aerodynamic robocars can easily reach. Given a choice of, say, 20 cent per mile 150 mph transportation, or 50 cent per mile 70 mph transportation, most people will choose the former.

  2. Why would the end user be required to pay for something beyond their control? For hire vehicles get dents and dings from rocks all the time. Fixing them is part of the cost of running that business. Any cost is amortized among all the passengers, not just the one who happened to be in the vehicle when the dent occurred.

  3. A driverless system would probably be a lot worse that the best drivers but better than the worst, even in the city.

    The good news is they are adding accident prevention technology as a supplement first. A speed governor would also be helpful.

    On 60 Minutes last night they said that Google has given up on the idea of having a car that operates driverlessly most of the time but expects the person to step in and hit brake when something unexpected happens, like a pedestrian or bicycle. Too dangerous. I hope others will follow that example.

    The fear scenario is that others promoting the technology will see the liability cost of a big upturn in pedestrian and bicycle deaths as worth it to reduce the number of pedestrians and bicycles and scare everyone back into cars, whether then can afford them or not.

  4. Personally I think everyone is WAY too hyped (on both sides) over it. Driverless cars probably won’t ever be fully driverless due to safety/insurance reasons, and will be in effect a glorified cruise control. Commercial companies have zero incentive to use automation tech when AB60 exists.

    At most, it’ll be an expensive addon that will appeal mostly to suburbanites and older people.

  5. Modern videos are the same thing, they just use a more modern vocabulary. In 20 years people will consider all of our “modern” predictions as a joke just like we do with hilarious predictions from the 80s and 90s.

  6. Are you sure about that? I think deployment of autonomous vehicles is still decades away. I was watching a 60 Minutes segment last evening and one of the comments from the reporter (Bill Whitaker) remarking on the deployment of these mobile devices was that one of the AV builders’ (I forget which one) vehicles was incapable of operating in rainy conditions. What about operating in fog or blowing dust, the latter of which is common in California’s inland empire region?

    The real test of AV operation is the capability to perform flawlessly under all circumstances; in any situation – the so-called weakest link in the equation. Also, mentioned in the 60 Minutes piece was that an occupant would still need to be completely alert as to possible dangers that may arise forcing immediate intervention (driver assistance) should there be a need for such, such as in dealing with an unexpected encounter with wind, rain, fog, snow, hail, dust, you name it.

    PRT systems could very well become commonplace before the first commercial AV ever hits the road.

  7. Or you can just use temporal separation. ie, have all delivery vehicles operate on roads from 8pm-8am while the rest of the day only mass transit, taxis and emergency vehicles can. But even then, most streets would not need it.

  8. Most gearheads hate robocars because gearheads are gearheads because they themselves can work on a car (which uses proprietary software) and drive it. You’re thinking of cagers and soccermom types. There’s a difference between hobbyists and casual users.

    Also, the NRA is the biggest opponent of smart-guns (ie ones that have an electronic lock on them) which is why they are not made. This is due to fears that the government could just press a button and turn off all the smart guns.

    It’s important to recognize that car and gun enthusiasts tend to be much more staunchly and ideologically opposed to smart cars and smart guns than anyone else. Whether or not this is a good or bad thing, I won’t comment on.

  9. The train or bus rapid transit line could be self-driving as well, significantly reducing the costs of operating a transit system.

  10. The issue with kids transportation is there must be a lot of trusts and supervision. Lets say that if you were to start a carpool program, there needs to be a community where the parents know and trust each other. A random system isn’t going to work for parents, and lots of low capacity vehicles are not going to solve the traffic jam. If you need a professional and high capacity that might as well be a school bus.

  11. Appealing “mostly to suburbanites and older people” may undermine your argument as (depending on how you measure) more than half of America lives in suburbs and older people will be 22% of the population by 2040.

  12. More like a On-star system where you will need a regular subscription to enable or continue that feature. These types of system require regular if not frequent updates and software companies have been selling software as services. The cars however continue to sell like a regular product and must continue to work if the user does not want that service.

  13. If they did that they’d probably just take the self-driving car all the way to their destination instead of using a suburban train or bus. Case in point, Britain eliminated a whole bunch of their branch line train service in the 1960s with the idea that people would just drive to local “hub” stations, and then get off at destination “hub” stations and then presumably take local public transportation, instead, people just started driving the whole trip rather than doing so many mode transfers.

  14. Part of the problem is kids don’t go to neighborhood schools typically, but to more distant ‘central’ schools. School districts long ago figured out how to cut down some administrative costs by having kids go to fewer and bigger schools, typically located out of the way on “greenfield” sites. This is one of the reasons it’s very uncommon for American schoolchildren to walk or bike to school.

  15. “there needs to be a community where the parents know and trust each other”

    Do you have children? I know every single parent of every single child in my son’s classroom. How is that for a “community”.

    In this day and age the ability of an app to connect riders going to the same school and to even be able to select for specific co-riders is trivial. Parents can decline to accept a carpool with specific students that perhaps their child does not get along with. The vehicles can be trivially installed. with webcams so parents can monitor the trip.

    Huge numbers of parents would make this decision given the chance. I once had a school bus driver fired for laying out a string of expletives at me on the roadway. Frankly I might just prefer the robot.


  16. I certainly agree that IF they had their way, the gearheads would ban robocars. But since there’s no Constitutional support for anything even remotely resembling such a proscription, they’ll use the existence of the cars as a tool to restrict access to roadways by other forms of transportation.

  17. It already is in Vancouver, BC. And they run trains on three minute headways all day long. It’s the way one would like all transit to run.

  18. Totally autonomous cars are really far off; but there’s a bunch of technology that could be released fairly soon that could make driving less wasteful and less of a pain:
    1) ad-hoc mesh networking between cars, from which the idea of “prevailing speed” could actually be determined
    2) accurate sensing arrays that can inform actual distance between cars, instead of relying on eyeball measurements and tricks to estimate
    I think the way forward will be some sort of big-brother-ish opt-in program pushed by insurance companies, who will absorb the costs in exchange for the right to craft better risk-models for their plan holders. Sucks, but that’s driving for you…

  19. Herbert S. White, dean emeritus of Indiana University’s library school, used to have a column in American Libraries in which he remarked more than once that organizations always find money to do what they really want to do. (The context was the long tradition in most universities and local governments to cry poor when it came to funding their libraries. Anyone surprised?) I often have occasion to think of that observation, especially when we talk about governments being broke in what is arguably still, in the aggregate, the richest nation on Earth.

    Now when you bring up the true cost of far-edge building and living, you’re preaching to the choir as far as I’m concerned, or at least to the oldest member of the bass section. Suburban sprawl, especially the exurban explosion of the last twenty or thirty years, has been a disastrous waste of resources and is unsustainable by any rational standard. But then, so is maintaining a military budget larger than the next seven countries combined (down from the next ten, though, according to!), perpetuating a large and permanent underclass in a wealthy country, or having one of the highest rates of gun violence in the world for decades without even attempting to find solutions. But we do all those things, and will continue doing them, because at some level they are things we, as a country, want to do. (Presumably not most of us here, but then we’re such a tiny minority that we’d hardly be calling the shots even if our country were still a democracy.)

    Similarly, I don’t think the U.S. has really changed its attitudes about sprawl; not fundamentally, not deeply. Certainly, super-wealthy individuals like the Kochs, most corporations, and most of the politicians who take their orders, are fine with the status quo; but I don’t think that it occurs yet even to most regular Americans that a good life could be possible without a McMansion on a giant isolated lot and a minimum of two gigantic vehicles to get wherever one wants or needs to go. I don’t know if that thought will ever percolate far in our society. Basically, I think that as a country, overall, we still love sprawl and will do whatever it takes to keep it going.

    I don’t look forward to self-driving vehicles with much enthusiasm. I do see how they could be very useful in some applications: they could do a lot to enhance mobility and reduce isolation for rural residents who aren’t comfortable driving and for people with disabilities that affect their ability to drive; and as several have commented in this discussion, they could be an invaluable “last-mile” adjunct in communities not dense enough to be fully served by conventional transit. They may end up serving such functions in countries where a modicum of reason and a basic awareness of the concept of the common good still influence public policy. Here in the U.S., to the extent they are developed and adopted, they will usher in the end of public transportation; the end, to a large extent, of our nascent experiments in returning to human-scaled, walkable, and bike-friendly communities; and a new Age of Sprawl that will make the last sixty years look like a timid opening act.

    As usual when I think about the likely future, I fervently hope I’m wrong. It’s not the future I want for my children and grandchildren, or for myself for that matter. But I don’t see realistic prospects of a less dystopian alternative.

  20. “A driverless system would probably be a lot worse that the best drivers but better than the worst, even in the city.”


    I expect it to replace taxis in crowded cities first. Taxi drivers are mostly pretty terrible, but the taxis go really slow anyway…

    The really hard problems for driverless cars are actually in the countryside at high speeds on rural roads.

  21. Depends how congested it is downtown. Nobody’s going to switch from taking PATH or NJ Transit to driving though the Holland Tunnel…. it would be so much slower.

    In the UK, people drive the whole trip out in the “counties”, but if they’re going to London, why then often they do park at the train station and take the train, because driving in London is yeeech.

  22. Actually, the 1950s stuff has dated really badly, but the stuff from the 1920s holds up pretty well. So does the stuff from the 1930s and 1960s. It’s interesting — there was something *weird* about the 1950s which is very out of sync with other time periods.

  23. Schoolbuses have always been the most out of date rattletrap buses on the road — always decades behind city buses and charter buses. They suck.

  24. The first robocar which runs someone over by going at high speed on a residential street will result in the complete banning of robocars.

    People are really, really paranoid about getting killed by robots — much more paranoid than they are about getting killed by humans. This has been true for *centuries* — it’s not changing.

    What I see happening is “nannyware” features on cars — devices which prohibit people from making dangerous maneuvers by preventing the accelerator from working if you’re trying to follow too closely, preventing the steering from working if you’re trying to crash into someone, etc. These will give insurance discounts and so will become ubiquitous.

  25. Well, here are some optimistic points.

    I think people are less and less tolerant of the massacres happening on our streets — Streetsblog is only one of the media outlets which is amplifying this.

    As a result, I think that driverless cars will be required to actually drive safely — maintain safe following distances, etc.

    About 2/3 of current drivers on the road DON’T drive safely. They tailgate, for one thing.

    Driverless cars, by not tailgating, will allow a *lot fewer* cars per hour on the expressway. (It’ll be about the same number of cars per hour in the city.)

    This means that the congestion problem will actually get *worse*. And congestion is one of the primary reasons why people start looking for alternatives to automobiles. If the train is speeding past the long queue of driverless cars, people will take the train.

    People don’t take the subway in NYC because it’s nice; they take it because it’s faster than driving. A LOT faster.

  26. I would go even further and say it’s not the subsidies OR the public policy in general — it’s *specifically* the zoning laws which BANNED dense construction, which therefore forced sprawl. This was done in the social-engineering-heavy 1950s, when there was an active political movement trying to FORCE everyone to live in a detached single-family house with a lawn with a nuclear family only.

    Relax those restrictive social-engineering zoning laws, and developers instantly start building duplexes and rowhouses, and people start moving into them.

  27. Joe: everything I see indicates that rowhouses are super-super-super popular, arguably more popular than fully detached houses. Unfortunately they’re very uncommon due to current zoning codes.

    Most people just don’t give a damn about lawns, it turns out.

    Lower and Midtown Manhattan is not normal development; it has to do with the rivers restricting movement. Bronx/Queens/Brooklyn is a more “natural” level of density. Much of LA actually achieves similar levels of density in a different way.

  28. If car storage is a profitable business, no reason not to have it. Even if it’s a “loss leader” for a business but is profitable overall, it seems fine to me.

    What’s not OK is the city subsidizing a grossly excessive supply of parking, way more than is normally used. That’s just waste.

  29. I don’t think they have a chance in hell of doing that.

    For one thing, there is actually a Constitutional right to walk down the street, and across the street. It’s worth reminding people of that, repeatedly.

    Basically if they try to turn a boulevard into a non-grade-separated freeway by excluding pedestrians, we should be able to force them to put in full sidewalks on both sides and frequent signalized crosswalks as a *legal requirement* of doing so.

  30. One major difference is that if a human-driven car kills someone, most people including the cops say “Oh, it was just an accident, let the human keep driving and keep killing”.

    If a robot-driven car kills someone, the reaction will be TOTALLY different. “BAN THE KILLER ROBOTS!!!”

  31. Interesting take on the possible effect on congestion. I hope you’re right. And I hope the alternatives (starting with transit) will still be there when people get sick of congestion. That touches on one reason why I’d rather see Chicago try to expand rail service* rather than staying with the current configuration of a sparse rail skeleton and nothing but buses in most of the city. Buses, as we know, at least as they operate here, will always be slower than driving.

    *Ha ha … I know. And I’d like a magic unicorn and a supersized order of world peace with that, please.

Leave a Reply

Your email address will not be published. Required fields are marked *


Q&A With Peter Norton: History Is on the Side of Vision Zero

Last week, a bunch of bigwigs gathered to talk infrastructure in one of Washington’s most historic and prestigious sites, the Hay-Adams Hotel across the street from the White House. I was offered an opportunity to interview former Transportation Secretary Ray LaHood, former Los Angeles Mayor Antonio Villaraigosa, and a host of other VIPs. But — […]