Exclusive: Lyft Cofounder Lays Out His Vision of the Driverless Future

Correction appended, Sept. 18

Lyft, the nation’s second-largest ride-hailing service, expects autonomous vehicles to account for a majority of its rides within five years, cofounder and president John Zimmer told TIME in an exclusive interview. Days after its chief rival Uber’s self-driving cars began ferrying passengers in Pittsburgh, Zimmer also said he expects car ownership will “all but end” in major U.S. cities in less than 10 years.

“There are already specific trips—whether it’s just on this street or just at this time in this perfect weather condition—that an autonomous vehicle could do today,” Zimmer tells TIME. And he believes that this is how the self-driving revolution will come to the masses: not by consumers swapping out their old cars for fully autonomous personal vehicles but by consumers paying for rides in self-driving cars they don’t own, with the type of trip restricted heavily at first and then growing more complicated as technology and regulations advance.

That is also the logic on which his five-year prediction is built. Because Lyft has data about where people’s trips tend to start and end, the company can estimate that a certain percentage of Lyft rides already lend themselves to self-driving cars. Such data also helps the company know where to focus their efforts for work like detailed street mapping, so that they can cover the biggest percentage of trips over the fewest miles. “We know in what order autonomous technology should be built, such that it covers more and more of those trip types,” Zimmer says. As of August 2016, Lyft was doing 14.6 million rides per month, triple their volume one year before.

This news comes as Uber is breaking a barrier in the self-driving race: Google may have been logging test miles longer than Uber and Volvo, their partner in Pittsburgh, but Uber just became the first company to bring self-driving cars to market (even if there are still backup humans in the drivers’ seats). Earlier this year, Lyft and GM formed a partnership and announced plans to develop autonomous vehicles together, with GM investing $500 million in Lyft. They have since been testing self-driving cars in Phoenix and have promised to deliver their first autonomous rides to customers next year.

In the interview and in a post set to publish on Medium, Zimmer details the ways he believes self-driving cars can potentially revolutionize our cities and our lives. In the not-so-distant future, Zimmer imagines people getting rid of the cars they own—or deciding not to own a car in the first place—and instead buying subscriptions for self-driving services from companies like Lyft and Uber. Think of it like a cell phone plan, with people signing up for monthly pay-per-mile plans in a certain class of car or a flat fee for unlimited miles in another.

“It’s our belief that these cars should run on a network,” he says, and in the extended smartphone metaphor, Lyft is the equivalent of the phone carrier running that network. Partners like GM and Cruise, a self-driving tech startup that GM acquired earlier this year, are responsible for the other two key components: the smartphone hardware (the car) and the operating system (the car’s autonomous brains).

The combined team working on autonomous vehicles from Lyft, GM and Cruise is in the hundreds. While GM and Cruise toil away on tangibles like sensors and algorithms that make sure the car can detect unexpected events in the road, Lyft’s team is working on problems like how to connect users with self-driving cars, how to convey to riders what the car is thinking and doing, and how to personalize the experience for riders (with, perhaps, known preferences for temperatures, music, lighting and routes). The latter is key, Zimmer believes. When customers are faced with the choice of pressing a button to get a self-driving car from different companies at roughly the same prices, they will choose Lyft like they’d choose a hotel chain or airline based on the “interior experience” or quality of service, Zimmer says.

Zimmer declined to comment on whether Lyft will exclusively work with GM in reaching this five-year milestone, nor would he say how many self-driving cars he believes will be required to cover the majority of Lyft trips at that point. “In the U.S. alone, two trillion dollars is being spent on car ownership, and cars are being used only four percent of the time. This makes no sense,” Zimmer says. “So what’s happening here is that there’s no bigger opportunity to create economic value in the country than changing that fact. And that means there are a lot of people that want to take part in that.”

Beyond the potentially massive economic payoff of self-driving cars, Zimmer says he is as taken with their potential to transform the way we live. By building our cities and lives around cars, he argues, space that could be devoted to housing or small businesses or public spaces is instead taken up by parking spaces, garages and wide roads to accommodate the congestion that is choking so many American cities. If networks of self-driving cars become robust enough for people to see them as a cheaper and simpler option than car ownership (which averages an individual about $9,000 per year), efficiencies that come with autonomous vehicles could allow cities to reclaim that space for better uses, he says.

With far less congestion, people could live further outside cities, with more space, less cost and a less taxing, self-driven commute. “If you optimize and eliminate traffic, all of a sudden, the circle around what is valuable near a city expands,” Zimmer says. He also sees Lyft playing a role in the policy debates about how to remake infrastructure that would inevitably define this version of the future. Arguing against parking garages, for instance, works for Lyft on two levels: “It’s both beneficial for us, because it makes it more likely that people will use these services but also we can make a strong argument that it’s beneficial to the cities,” Zimmer says.

The Lyft leader wants to start selling the public and government officials on his vision long before it’s time to make those on-the-ground decisions. “You just have this blank canvas of real estate that was used for empty vehicles,” Zimmer says. “It’s not inevitable that we use that real estate correctly, because we didn’t. But now we have a technology opportunity to redo that.”

Correction: The original version of this story misstated John Zimmer’s title. He is the Lyft cofounder and president.