Self-driving cars developed by the likes of Google and Tesla are generating a lot of buzz as their wheels are tested and designs refined. While it’s difficult to imagine driverless cars in a New York City — a city defined by its traffic jams and vocal drivers — all signs point to their eventual introduction. But what exactly would that look like? The vehicle and pedestrian congestion levels of NYC would make the adoption difficult, to say the least.

Self-driving cars universally raise questions regarding safety, legislation, mass transit, and environmental and human impact, not to mention their inevitable impact on gas prices. Add to that list the volume of foot traffic in New York City, the city’s well-established taxi fleet, and a massive public transit system, and you end up with an interesting conundrum.

Here’s a look at how self-driving cars would transform one of America’s largest metropolises, and what it could mean for urban spaces across the nation and world.

Safety

For self-driving cars to become commonplace, the cars first and foremost need to be safe. Early in 2016 Elon Musk claimed that Tesla’s Autopilot feature was “’probably better than a person right now.’” However, Sebastian Thrun, the founder of Google’s autonomous driving initiative, noted Autopilot’s limitations during test drives.

According to this New York Times article, “The Tesla performed well in freeway driving, and the company recently fixed a bug that had caused the car to unexpectedly veer off onto freeway exits. However, on city streets and country roads, Autopilot’s performance could be described as hair-raising.”

In fact, the first recorded fatality occurred in May of this year, when a Tesla car on self-driving mode failed to break as a tractor trailer ahead made a left turn. In a city, this type of incident could be even more likely — especially in New York City, where vehicle and foot traffic are perpetual problems. Still, human error being the main cause of accidents, driverless cars are likely to prevent many deaths and accidents.

In a tight city, that won’t be easy. The Midtown grid may be easy for Autopilot to navigate, but what about the twisting streets of the Financial District and the Village? Northeast weather conditions can introduce additional challenges.

Another big issue is the switch to manual control, as prompted by the Autopilot, which presumes close human supervision. According to John Leonard, an MIT mechanical engineering professor, “’The whole issue of interacting with people inside and outside the car exposes real issues in artificial intelligence.’”

Human error while operating large machinery is a known–if beleaguered–quantity. Although potentially safer, self-driving cars will probably be held to higher safety standards prior to widespread use, prompting a chilling question: will these self-driving cars and corresponding apps be susceptible to sabotage and hacking?

Despite precedent in features like cruise control, which automates some aspects of driving, people will undoubtedly want certain guarantees before giving up manual control.

Legislation

With continued funding, testing, and refinement, safety issues can certainly be improved upon, if not eliminated. However, that leaves the next hurdle of legislation.

Autonomous cars must be made legal state-by-state, which could be a challenge depending on the region and public perception. What will the inevitable transition period look like–the mix of driven and self-driving cars? How will issues of litigation be settled in the inevitable event of an accident?

If driverless cars are to become normalized, especially in big cities, all of the legal kinks will need to be worked out first.

Congestion

Beyond all these practical implementation concerns, how will these self-driving cars affect the New York cityscape? According to an article by Curbed, autonomous cars would eventually lead to a “shared vehicle ownership model, ” similar to Citi Bike or Zipcar, which would in turn lead to more efficient driving and parking practices, such as the reduction of parking spaces and roadways, which would mean the reclamation of public space.

According to Dr. Kara Klockman, “shared bikes” will also get a boost from this kind of transit model. Alain Kornhauser posits that autonomous vehicles will actually make regional transit and by extension urban public transit more accessible.

Essentially, if all cars (or even most) in NYC were to go the driverless route, it would mean less traffic and smoother roads.

Commuting

Still, one can’t help but ask if autonomous cars would be a preferred mode of transport over public transit. There’s the transfer to consider, as well as the costs of each mode of transportation (and the tally when combining them). Why not take an autonomous car all the way to work? Even with carpooling, we might see increased urban sprawl and an uptick in commute times.

As Dr. Kara Klockman notes, “’A big concern that I have for cities, states and regions is excessive travel. […] I think we’ll need a credit-based congestion pricing model.”

As the environmental costs of vehicular travel continue to be scrutinized, alternative modes of transportation are more likely to be prioritized, especially in cities like New York that have robust public transport systems already. But if driverless cars could operate sustainably, quickly and cheaply, they may play a chief role.

Human impact

The question of cost raises additional concerns. Gerry Tierney notes: “If we’re moving toward this autonomous, decentralized transit system, we need to make sure that it’s accessible to everybody, that there’s a social equity concept in the design.”

For people of lower income who cannot afford to live in the city center where they work, self-driving cars could be a real time-saver. But if these people are priced out of the service, we could end up with a very lopsided combination of public transportation systems and autonomous vehicle transit.

Another human cost to consider is the impact on labor. As this New York Post article puts it, “Chalk up another possible job victim of the Internet age — the New York City cab driver.” Mayor Bill de Blasio already signed an agreement with Google in April 2015 to add thousands of self-driving cars to New York’s taxi service.

The proposed 2016 White House budget included $4 billion for pertinent research funding. And Uber, along with its competitor Lyft, are planning to use autonomous cars, which–given the recent controversy surrounding Uber’s employer practices and the app’s controversial reception in many cities–could prove contentious.

Lastly, what of New York City’s finest, the NYPD? Traffic tickets add a consistent stream of revenue to police forces, and assuming self-driving cars limit (if not remove entirely) the possibilities of road violations, some estimates predict half of cops could be put out of work.

While this would be bad in the short-term for police officers, it could free up their time and resources to concentrate more fully on serious crimes.

At the end of the day, it’s tough to say what NYC would really look like if and when driverless cars are popularized in urban spaces. But we can say for certain they will play a part in the future, and that the impact on America’s infrastructure and the face of its most vibrant city will be drastic.

Featured image: DiAnn L’Roy via Flickr.