Driverless Cars Could Make Transportation Free for Everyone—With a Catch

As self-driving cars inch closer to everyday reality, journalists, futurists, economists, and ethicists have weighed in with numerous predictions about autonomous vehicles’ future impact. Liquor sales will rise, the predictions go, since no one will worry about driving under the influence. Cars will have ethics knobs, with settings that vary from minimizing overall harm in a crash to saving the occupants at any cost. Steering wheels, traffic jams, and parking meters will become quaint relics, like hitching posts and watering troughs.

But these prognostications miss what will be one of the biggest developments of all: In a world full of autonomous autos, transportation will become free. Not just hands-free, or driver-free, or go-wherever-you-want free. But free as in beer: complimentary, gratis. Summon a car and travel for nothing—that is, so long as you are willing to make a stop or two en route at sponsoring locations.

Picture a not-too-distant future where a trip across town is available to anyone who will spend 15 minutes in McDonald’s on the way. Not a fast-food fan? Then for you it’s Starbucks, a bookstore, the game parlor. Rides with a child stop at the Disney store, while teenage girls are routed via next decade’s version of Zara and H&M. Unlike today’s UberPool, with its roundabout routes and multiple passenger pickups, “UberFree” features tailor-made routes and thoughtfully targeted stops.

Realtors could pay to have the cars drive slowly past featured properties for sale, past the nice new elementary school in the slightly more affluent neighborhood. At election time, a candidate’s campaign would route voters through run-down areas while a voice-over blames the opponent for this decline. And if you happen to mention at some point in the day that you are chilly, or your shoes hurt, or you have a party to go to, the friendly virtual assistant that lives on all your devices—Alexa’s granddaughter—ensures that your next trip’s stops include relevant sponsored solutions.

These rides will be popular both with passengers, who see a great bargain, and with marketers, who leap at the chance to not just show people ads, but to physically place potential customers in front of temptation. Say you and your friends want to go to the beach. For $20, a van could take you all directly there. Or, you can go for free, with stops at Amazon Food and at Target. At the beach, you’re pleased with the deal: You wanted snacks anyhow, and though you did not need a new beach towel, Target had really nice ones, so you picked up two—along with a cute hat and a phone case you saw by the register.

A generation will grow up with the expectation that transportation—convenient, door-to-door transportation—is free. But how free is free? Less enthusiastic will be environmentalists, consumer watchdogs, obesity monitors, and others fearful of the dangers of unchecked consumption. And the person traveling fare-free may find that the free ride, like the proverbial free lunch, does not really exist.

* * *

Autonomous vehicles are still about a decade away, but the potential pitfalls and inherent conflicts of this transportation arrangement are already apparent in existing precursors of the promoted ride. Take sponsor-routed cab rides. In Las Vegas, it has been common practice—made legal and taxable in 2010—for businesses, most notably strip clubs, to pay taxi drivers for steering customers to them.

Which Vegas businesses did this and how much they paid was, until recently, valuable covert info, shared within the cab-driving community. Now, there’s an app for that: Kickback helps drivers quickly find, by category—nightclub, strip club, gun range, wedding chapel, and so forth—all the businesses that will pay them, by the customer or carload. The going rate at cannabis dispensaries is $10 per drop-off. Vegas Weddings pays $50, so long as the customers purchase the $199 package. Strip clubs have tiered rates, paying up to $100 to cabbies for male customers and as little as $10 for women brought by ride-share drivers.

Even though apps like Yelp mean the local cabbie is no longer an indispensable guide, many people still seek their driver’s advice, enjoying the feeling that they are getting secret insider information.

But the system benefits drivers at the passengers’ expense. The drivers’ incentive is to take people to the places offering the biggest kickbacks, but those businesses are often the least desirable. Indeed, the amount a club pays on the Kickback app seems to have an inverse relationship to its Yelp ratings: Sapphire Gentleman’s Club offers among the highest kickbacks ($80 for male taxi riders), but has many one-star reviews, whose complaints range from watered down drinks to drugging and robbery. On top of that, some clubs present customers who arrive in a cab with high cover charges to make up for the driver’s kickback.

Will replacing the human driver with an autonomous car shift the equation in the passenger’s favor? Not necessarily. It will depend on how the systems are designed, who is designing them, and how aware the users are of the potential for manipulation.

How we get from point A to point B is a process determined by numerous and often-competing interests. Some of these tensions can be seen today in the rise of algorithmic navigation apps such as Waze and Google Maps. These apps are ad-supported, so their incentive is to retain the human drivers who want to get to their destinations as easily as possible. Thus the apps reroute, for instance, from congested highways onto obscure side streets. While this speeds the users along, it also disrupts life in areas intended to be quiet, residential neighborhoods. Currently, neither the irate neighbors nor city planners have much recourse: The roads are public and the cars driven by individuals, making intricate congestion pricing and driving zones infeasible.

Cities may have more control once these algorithms are routing autonomous cars. The “driver” would now be the algorithm—or more accurately, the company that controls the algorithm. At this scale, planners could create zones of permissions and pricing for traveling on different roads, regulations that the algorithmic “driver” must obey to continue to be licensed to drive these streets. Ideally, their goal would be to fairly balance the competing needs of rich and poor, people and businesses, passengers and residents. The question of what is fair, however, is likely to be contentious. Wealthy neighborhoods could be made off-limits to all cars without resident or guest permits. Speedy scenic routes would become the business class of car travel, and the slow routes, lined with McDonald’s, dry cleaners, and other strip-mall stalwarts, the urban economy class.

The relationship among businesses, passengers, and drivers is different. Payments (whether kickbacks or sponsorships) from a business to a driver (whether human or algorithmic) in return for redirecting people are a way for businesses to align the drivers’ incentives with their own. In the case of the Las Vegas taxis, the passengers still have recourse: The drivers operate independently, and passengers are capable of directing them to a preferred location. But once independent drivers are replaced with autonomous vehicles under the control of a monolithic routing algorithm, if the company that controls the algorithm has special relationships with businesses, it can wield far more influence on where people shop and eat, on what they see—and where they do not go.

Without the equivalent of net-neutrality regulation to protect humans traveling in cars, the flip side of sponsored rides could be ride surcharges: barriers to visiting places the algorithm sees as rivals.

* * *

There are a few reasons why we do not have widespread sponsored rides today. One is that the cost of the driver’s time is too high (the Las Vegas cabbies take tourists to destinations, not to additional stops on the way). But once the taxi equivalents of the future are autonomous, the cost of idle time should plummet. Another factor may be psychological: Suggestions about where to where to go and how to get there will come from a familiar, trusted virtual agent.

Indeed, while the behind-the-scenes brainpower of an autonomous fleet is a vastly complex algorithm, passengers will only interact with a friendly persona, something like Alexa, Siri, Cortana, but with all the advances—subtler vocal inflections, more accurate perception and manipulation of human emotional states—that a few years of concentrated technical development will bring.

Autonomous cars will be part of an ecosystem of intelligent agents and personal-data vendors. The information they are able to base your route on—and how they present an itinerary to you—will not be limited to where you say you want to go, but on all the data they have about you. Note that companies with immense personal-data collections, including Amazon, Baidu, Google, and Uber, are in the race to develop autonomous cars. (Uber has recently launched UberEats, a delivery service that collects data about customer habits for participating restaurants.) The same system that one day provides your ride may have access to, if not control over, your calendar, contacts, medical records, and holiday shopping lists.

With her deep knowledge of your life, the virtual agent that plans your trips will be great at her job. Unlike the insinuating cabbie or the awkwardly inexperienced Uber driver, her flawless persuasive banter will never make you feel uncomfortable or coerced. When she suggests you leave a bit early for your meeting across town so that you have time to stop on the way for a gift for your sister’s birthday, and didn’t you also want to get your hair cut, and oh here’s an excellent new place to try—she does so as your friend, your confidant. Or so she will be designed to make you feel.

This future might seem, at first glance, like a techno-utopian scenario. Yet little will have really changed. Like the Las Vegas cabbie, the goal of these systems will be to make profit for themselves. They’ll just be much better at persuading you otherwise.