The Paradox at the Heart of Elon Musk’s Self-Driving Vision

A sleek, gold car pulls up to a bustling corner market, and a middle-aged couple alights. A woman eases a suitcase into the same vehicle's spacious trunk. Later, a doodle and its master watch rocket videos in the front seat as the car eases around the neighborhood. No driver, no steering wheel, no pedals, no waiting, no traffic, no worries: This Tesla Cybercab drives itself.

That's the vision shown off by Tesla CEO Elon Musk last week during a presentation broadcast from a set at Warner Bros. Studio, outside of Los Angeles. Some 20 prototypes cruised the movie lot as a series of mocked-up images showed scenes of the idyllic tomorrow these sleek people-movers could usher us into. But experts say Tesla's brave, new city of the future will need more than a few robotaxis to transform this hi-def rendering into reality.

While mostly sidestepping the technical challenges of building self-driving technology, Musk chiefly focused on what an autonomous taxi service might mean. Starting next year, he said, Tesla owners should be able to share their personal cars by putting them into self-driving mode while they're not using them. It would be a sort of Uber-cum-Airbnb, the car off hustling for a paycheck while its owner hustles for their own. A vehicle constantly on the move could obviate the need for parking: "You're taking the '-ing lots' out of parking lots," Musk quipped, as a presentation showed the asphalt expanses around LA's notoriously trafficky Dodger and SoFi Stadiums transformed into green spaces.

In short, Musk and Tesla argued that autonomy means more pleasant lives for all. "A car in an autonomous world is like a little lounge," Musk said, noting a ride in a self-driving taxi would cost less than even a bus trip. "You can do whatever you want ... and when you get out, you'll be at your destination. So yeah, it's going to be awesome."

But make personal self-driving cars too inexpensive, and too pleasant, and you've got a city-sized problem on your hands. Cheaper, comfortable rides could lead to even more traffic and even more driving, says Adam Millard-Ball, a professor of urban planning and the director of the UCLA Institute of Transportation Studies. For proof, check out the studies of Uber 's and Lyft 's effects on US cities; research suggests that, despite marketing promises about the death of private car ownership , their introduction brought more urban traffic, not less.

In this way, cheap robot taxis are a sort of double-edged sword, ending in more urban sprawl. "That's going backward for the environment and for other urban goals—whether it's being physically active or socially inclusive," Millard-Ball says.

Taking the '-Ing Lot' Out of Parking Lot? Parks instead of parking lots could be a nice upside to self-driving. (Apartments instead of parking lots could also be really cool.) But it'll take more than just the switch to self-driving to get there. Anyone running a self-driving car service hoping to use as little parking space as possible will have to make a super efficient network. That's going to require people to share vehicles. And people don't love to share.

"People love to move in a safe and comfortable way," says Andreas Nienhaus, who heads up the consultancy Oliver Wyman's Mobility Forum. "Whenever people have the choice and they don't have the guidance, they will opt into a personal car."

Car owners are also unlikely to share their private vehicles with others. Peer-to-peer car-sharing services , which let users rent out their cars to others when they're not using them, have struggled to scale, Nienhaus says. "The car is still an emotional product," he says. "It's mine, and I'm annoyed when it comes back dirty. People are a bit hesitant to give away their car."

One way to convince people to share self-driving vehicles would be to create public policies that make it more attractive for them to do so. Congestion pricing—which has worked in cities such as London and Singapore but failed to launch in New York City this year—would make it more expensive for people to enter some city roads at certain times of day. The prospect of splitting those costs could provide the kick in the pants needed to get people into the same self-driving car.

Problem is, this isn't really in Musk's control. "An auto manufacturer is not a policy entity," says Marlon Boarnet, a professor of public policy at USC and the director of the Metrans Transportation Consortium, a transportation research center. It'll take much more than Tesla churning out Cybercabs to surround Dodger Stadium with green.

Sharing would require a sophisticated matching system, Boarnet points out, the type even Uber and Lyft don't seem to have cracked. Both companies once offered more robust and cheap "pooled" options, but those mostly died away during the pandemic, as those companies began to focus on turning profits. There simply aren't enough people going to exactly the same places at the same times of day to create a perfect shared ride service.

Cybercab City For Tesla, and for cities, the stakes are high. Autonomy without sharing could be exactly the opposite of what Musk wants—a city that's choked with traffic and a pain to travel in rather than one that is beautiful, with smooth roads.

The other hitch: Even if Tesla meets its incredibly aggressive self-driving goals—which include full autonomy in Texas and California next year—building a city around self-driving is going to take some time. City infrastructure, with its roads and parking lots and buildings, "is quite static," says Brian Jencek, the director of planning at HOK, a design and architecture firm. Build a road, and form follows function. Someone will probably drive on it. If cities are redesigned to better accommodate self-driving Teslas, the implications could stick around for a long, long time. Or, as Jencek puts it: "Anytime we alter mobility, we change the very nature of our cities."