Sometimes a driverless car is not just a driverless car: Thoughts on widgets and systems
Driverless cars are coming to a street near you, and soon. “I think that genuine self-driving cars will be available within a decade and that they’ll be big game changers,” writes Kevin Drum, who’s been on a driverless-car kick lately. Or take it from Dan Gage, spokesman for the Alliance of Automobile Manufacturers: “That fully autonomous car of the future is not that far away.” Or just watch this:
Google has been testing driverless cars extensively. Nevada and California have passed legislation permitting them on the streets. Other states will soon follow suit. This is a near-future thing, not a sci-fi dream.
I happen to share Drum’s enthusiasm for driverless cars, and want to use it to make a larger point about social change.
Widgets and systems
Let’s think about the distinction between widgets (pieces of technology) and systems (the cultural, economic, and infrastructural systems in which technologies are embedded).
When it comes to sustainability, as with many other things, we find it easiest to think in terms of widgets. The first step is always to swap them out for better widgets. If we want to use less electricity, we swap our dishwashers and furnaces for more efficient ones. If we want to transition from coal, we swap out coal plants for natural gas plants or wind farms. If we want our homes to consume fewer resources, we stick solar panels on the roof and chicken coops out back. And if we want to use less oil for transportation, we swap out gasoline cars for cars that run on something else.
The virtue of widget-swapping is that widgets can be compared on an apples-to-apples basis. They share a common set of parameters, are embedded in the same systems, so they can be put on the same scales, as it were. We can compare power plants in terms of dollars per kilowatt hour, cars in terms of miles per gallon, appliances in terms of sticker prices.
These kinds of comparisons usually show that the more sustainable widget (electric cars, renewable energy, low-flow toilets) is more expensive. One reason is that the benefits of more sustainable widgets — less CO2, or cleaner air, or more efficient resource use — often “don’t count” in market competition. Carbon emissions, for instance, are free right now, so if you spend more money for power generated with less carbon, you are, within the context of our current power system, wasting money.
It is much more difficult to compare alternative systems. It’s difficult to envision different systems — different patterns of habit and behavior, different markets, different infrastructure. And if we can imagine them, it’s difficult to compare their net costs and benefits to that of the status quo. They represent different ways of valuing things, so it’s difficult to weigh them against a common metric. (I’m trying to avoid the term “paradigm shift,” as I don’t want to be part of debasing Thomas Kuhn’s useful notion.) Suffice to say, system change can be nonlinear.
Most of all, it’s difficult to see a path from here to there, from the systems in place to new systems. We know how to swap out a widget for another widget. But how do we drive intentional systemic change?
Cars vs. transportation systems
Take, for instance, transportation systems. What will happen if regular cars are switched out for driverless cars?
In the beginning, driverless…ness will likely be an option in some cars, sold as a safety measure or a way for older or visually impaired people (like this dude) to be able to drive. That’s what Alex Tabarrok says here and I mostly agree. That will be the foot in the door.
But adopting driverless cars could mean much more than just substituting a new, better-performing widget in place of an old one, especially if we act with some foresight and thoughtfulness. Let’s ponder what sorts of systemic changes might be possible (emphasis on might).
• When cars can rely on foolproof sensing and robot precision, they can drive much closer to one another, and at steadier speeds. That alone will double or even triple the capacity of existing roads and help reduce traffic jams. If existing roads are able to accommodate more traffic, it reduces the need for new roads.
• With human error out of the picture, cars won’t need to be so heavy. Right now we drive around in huge steel tanks, up-armored to protect against high-speed collision from any angle. With collisions minimized or eliminated, cars can be made from far lighter materials, reducing stress on roads and resource consumption.
• Felix Salmon speculates that, with a combination of increased road capacity and lighter vehicles, it may become more resource-efficient for people to travel by car than by rail, eliminating the need for public rail transport. I don’t quite buy that, but it’s intriguing.
• When cars are much lighter, it will be easier to move them using electricity. And when you switch out an internal combustion (or hybrid) engine for simple electric motors on each wheel, you save another huge chunk of weight, further increasing electric range.
• Once most cars are electric, large-scale electric-charging infrastructure will become more cost-effective. It is already possible to charge an electric car without a plug, through induction. Imagine if induction-based charging were ubiquitous, in parking lots, on curbs, even in the roads themselves. Smart electric cars would be charged continuously, ambiently, just by being in an urban area or at a rest stop. Goodbye, range anxiety.
• Once most cars are electric, the top cause of urban air pollution would be eliminated. Urban air quality — and thus urban public health — will increase exponentially. There will be fewer work and school days lost to illness.
• If there’s a fleet of shared driverless vehicles available, and a car can drive itself to you whenever you need one, there will be no need to own a car. Why lug it around? Now, think about the last time you drove around in an American suburb. Imagine if all those houses didn’t need garages. What could all the extra space be used for? Think about all the gasoline and insurance and maintenance money involved in car ownership replaced by a simple, cheap subscription fee to a car service. What would all the saved money be used for?
• Privately owned vehicles are in use about 10 percent of the time and spend about 90 percent of their time parked, useless, taking up space. Cars that are shared could be in use almost continuously. Imagine if every adult American owned 10 percent of a car rather than 100 percent of one. That’s 90 percent less demand for cars! (Not exactly, but you know what I mean.) More sharing = less resource use.
• Today the average American spends about 100 hours a year commuting, and there is nothing more soul-crushing than driving in traffic. Imagine having all those hours back. The car is driving, so you’re reading, sending email, calling clients, surfing the internet. You’re working. What would it mean to you to have 100 more hours a year? What would it mean to U.S. economic productivity to have billions of additional work hours every year? What would be the net health and psychological effects of all that reduced stress?
• Driverless cars are equipped with sensors to avoid other cars, bikes, and pedestrians. So, as Drum says, there will be fewer bike and pedestrian road deaths — eventually, perhaps, close to none. If cars are simple and light and unobtrusive (just a plastic shell and motors for each wheel), and they’re smart/aware, they will be much more amenable to moving around amidst human beings. That means that many city surfaces will be multiple-use, with pedestrians, bikes, and electric vehicles mingling freely. It would be a much more dynamic, organic mode of interaction than the current, rigid streets vs. bike lanes vs. sidewalks set-up.
• If cars are smart — i.e., filled with sensors and connected to the internet — they will all be aware of one another, aware of traffic conditions, aware of available parking spots. That means no more driving around aimlessly, looking for parking, an activity that accounts for around 30 percent of traffic in U.S. business districts.
• If cars are smart, they will be able to customize to each user. Wave your subscription card and the carshare system remembers you — it starts the music you like, establishes the interior temperature you favor, calls up your email and favorite websites on its screens, suggests nearby restaurants you might like. It is your car, nay, your rolling office/den, albeit only temporarily.
• If cars are smart, they will be able to create carpools, on the fly, for those willing to save a little money by sharing rides. All cars will know where all other cars and all car requests are at any given time; they will be able to communicate and coordinate to consolidate as many riders as possible as efficiently as possible. Carpooling will become the default: a dynamic, just-in-time form of public transportation.
• If cars are accessed via subscription, it will be trivially easy to charge user fees that can replace gas taxes as a source of funding for infrastructure maintenance. It will be perfectly equitable: everyone will pay for roads exactly to the extent they use them, automatically.
Anyway, that’s just scratching the surface. (I wrote about all this stuff in an old American Prospect piece I still remember fondly.) It leaves us with two big questions.
How much to pay, how fast to drive
First, how much would it cost a city to build a transportation system like that — fleets of smart, lightweight electric vehicles, available on demand to subscribers — and would it be worth it? Questions like this are incredibly difficult to answer. How to quantify all those changes in behavior, public health, road capacity, productivity, land use … any attempt to predict all of that and tally it up with a definitive number is just this side of a wild guess. Worse, the costs are usually easier to predict and quantify than the benefits, especially the second- and third-order benefits, so conventional cost-benefit analysis is almost always biased against ambitious, long-term systemic change.
Ultimately, in my humble opinion anyway, the question of whether to pursue a new system, whether it’s transportation, electricity, water, whatever, cannot be settled by economic models. We just don’t know enough about the future, or about systemic change, to achieve that kind of precision — we end up with outcomes that reflect the assumptions we feed in. In the end, it comes down to values: what kind of society and culture we want, what kind of people we want to be, what we owe future generations.
Those are the kinds of questions that people who benefit from the status quo don’t want you to ask. They can tolerate some limited widget-swapping, but systemic change scares them to their bones. So they would very much like to restrict everyone to widget comparisons: Solar electricity is more per kWh than coal electricity. Electric cars are more expensive than gasoline cars. An acre of regenerative agriculture produces lower yield than an acre of industrial ag. They want narrow, intra-system comparisons, not broad comparisons of other possible systems.
The second big question is, assuming we want a new system, how do we get from here to there? How do we build a system like that? System change often appears impossible because of the complex, nested chicken-and-egg problems: Which comes first, the electric cars or the charging infrastructure? You can make smart electric cars much lighter if they’re sharing the road with other smart electric cars, but what about now, when they’re sharing the road with armored behemoths? How do you transition from a gas tax to a user fee when half the cars are smart electrics and half dumb internal combustions? How do you get the public to demand things they don’t yet know could exist? And how do you justify the costs, when the larger systemic benefits of the new order are not yet manifest?
It can seem intractable. And yet we desperately need systemic change. It won’t be enough to plop wind farms down in the same system coal plants operate in, or electric cars in the same system gas cars operate in, or rooftop solar panels in the same suburbs we’re building today. We need to figure out how to use widgets — and policy, and activism, and behavioral science, and every other tool in the toolbox — to accelerate social change in the direction of sustainability.
I don’t have any simple answer or grand conclusions here, but I do think it’s worth keeping in mind the tension between widget thinking and systems thinking. It may be that, sometimes, widgets that are more sustainable on some metric or other actually have the effect of further entrenching unsustainable systems. (Lots of people are going to say just that about electric cars, and yell at me for enthusing about them.) It may be that, sometimes, widgets that are less sustainable on some metric do more to drive systemic change. (Arguably natural gas plants are doing this by killing coal in America.)
I think driverless cars could be such a lever for systemic change in transportation systems, just as I think distributed renewable energy and smart grids will eventually be such a lever for power systems. But how can that system change be accelerated? That’s the question of the 21st century.