Waze Hijacked L.A. in the Name of Convenience. Can Anyone Put the Genie Back in the Bottle?

Level 1 is where you start. The higher echelons of control come only with obsession. You jack in for 45 minutes on your lunch hour. Afternoon coffee break. At night when your spouse imagines you’re cranking away on that office project. They have no idea. But when your fingers brush the keys, out there on the very real streets of Los Angeles, traffic shifts and undulates, like rainwater cutting a path. Homeowners wake up to find themselves trapped in a pop-up freeway hell that makes it nearly impossible to exit their driveways. The transportation officials and the council members and the whining neighborhood associations are mere spectators. The gamers won. The algorithm is God. Technology has spoken, and you know your supporting role. The unheralded superhero in this movie, giving every Angeleno the power to take back the streets.

Los Angeles went wild over Waze when it first hit the city in 2011. It was the brash counterpart to the Google and Apple traffic apps that took a more conservative tack to turn-by-turn directions, generally providing two or three traditional highway and thoroughfare routes. Not Waze. One definitive route. The wormhole, the ultimate shortcut, the secret trick to avoid getting screwed on the 405 or the 110. In September 2014 this magazine published “The Surprising Wisdom of Waze,” an ode to what seemed algorithmic serendipity, with honeyed lines like, “To me, Waze is the Voice of God.”

Why the adulation? Waze grabbed you—the harried, downtrodden L.A. commuter—with the subjective, hard-to-counter feeling that this crazy route over hill and dale just had to be faster. Speeding through residential neighborhoods gave hundreds of thousands of Angelenos something they’d scarcely known in this famously gridlocked city—the sense that they were in control. Early adopters suddenly had a technological edge on those same dummies who pay retail, don’t rewind the odometer when they return their leased Porsche, and don’t know the tricks to getting a table at Maude. The Waze founders were geniuses. Waze did save you precious minutes…for a while.

James Marshall

But Waze’s algorithmic promise did not last. Today Los Angeles residents could be forgiven for feeling stuck in the film Groundhog Day, waking up over and over again to a tech-triggered, traffic-nightmare time loop now going on five years, with no end in sight. As early as December 2014, L.A. Councilman Paul Krekorian, whose district includes North Hollywood and Studio City, began hearing from the front lines of the nearly 100,000 households he oversees. “Neighborhood streets that had been quiet all of a sudden during the morning commute were filled bumper to bumper,” he recalls. “Residents were forcefully complaining to my office.” It came on like a tornado. “There was a sudden change. Mostly in the hillside areas with narrow streets. People looking for shortcuts. The impact was dramatic.” The minutes or seconds you might save by Wazing transformed dozens of peaceful L.A. neighborhoods into loud, exhaust-fumed residential gridlock. Drivers careened down local streets steeper than double-diamond ski slopes. Trucks got pinned on corners. Citizens fought back, reporting phony crashes and traffic jams in a desperate counterattack. Not so fast. A Waze spokesman, oblivious to the irony,
reflexively sneered that “a group of neighbors can’t game the system.”

The headlines from that long-ago December captured the first wave of what Krekorian termed “the ultimate failure of laissez-faire, libertarian thinking,” what happens when corporations wreak havoc on communities for profit: “Locals Upset at Google’s Waze for Sending Traffic to Their Streets,” “L.A. Traffic Is Getting Worse, and People Are Blaming the Shortcut App Waze,” “Angry L.A. Residents Are Trying to Sabotage Waze Data to Stop Side-Street Clogging.” Krekorian’s office reached out to Waze and the operators of similar apps and got “a complete stonewall,” he says. The L.A. Department of Transportation? “They got zero response. Quite an arrogant attitude,” says Krekorian, recalling that Big Tech essentially said, “It’s your problem to fix this.”

Our story begins in Israel, where Ehud Shabtai had a seemingly innocent idea: crowdsource the mapping of the Holy Land in Hebrew and share it for free. It made sense: Israeli mobile GPS systems were lousy in the mid-2000s. Shabtai had studied computer science and philosophy at Tel Aviv University, so as a hobby—simply to get around town—he hacked out FreeMap Israel in 2006. He was, as they say, an accidental entrepreneur. Within two years he teamed with Uri Levine and Amir Shinar to code what would evolve into Waze.

The original Israeli Wazers were playing a game that helped build a product. Drive a Tel Aviv street yet to be mapped, and the app would simultaneously display a trail of pellets, which your digitally represented vehicle would gobble up as you drove along. The Pac-Man-style, Pavlovian design hooked early Wazers and helped make the app “sticky,” fueling a rapid global expansion. Shabtai and company raised $67 million between 2008 and 2011. Borrowing from Tom Sawyer’s playbook, they enticed legions of map dorks to do the essential early mapping for free. Google snapped up Waze for just shy of a billion dollars in 2013, earning Shabtai $78 million and Levine and Shinar $65 million each. A large chunk of that tremendous wealth was built off the keystrokes of dedicated yet unpaid volunteers. That’s hard to comprehend, given the cofounders’ $200 million-plus bonanza. But that’s where we just don’t get it. Tens of thousands of Waze map editors earned another kind of currency: the power to toy with Los Angeles and very nearly every major city on the planet—in effect, to play SimCity with actual cities.

Want to gain a sense of what took the editors down the wormhole? Log in to your Waze account. Fire up the optional Waze Map Editor Academy introductory video and enter a retro-cartoon video-game world fueled by ’80s-era metal rock. A corny, booming voiceover exhorts new editors to “choose your own adventure … and so much more!” There’s a lot of blasting, flashing, and bleeping as cartoon avatars zip around a Pac-Man-like maze of streets, zapping away. Then, the music abruptly fades as the tutorials segue into a perfunctory audio-free guide.

This is where anyone can slip online and get a sense of why so many map geeks get hooked: the omnipotence that comes when you—an average citizen—can change the designation of local roads and suggest edits to “segment attributes” such as traffic direction, unpaved roads, allowable turns. Yes, it sounds crazy, but you’ve now joined the local tech transportation underground, usurping powers once held solely by LADOT. Edit more and up goes your status, like earning Boy Scout merit badges. With no programming skills required and no training in transportation planning, safety or management, suddenly you’re the behind-the-scenes techie whose awesome back-end coding pushes out to drivers’ mobile interfaces within 24 hours.

One need only skim the city of L.A.’s Complete Streets Official Guide to understand how counter to traditional traffic planning it is to empower untrained amateurs—backed by a massive for-profit corporation—to intentionally wreak havoc on the traffic patterns of America’s second-largest metropolis. Weighing in at 244 pages, the guide is meant to “ensure the safety, accessibility, and convenience of all transportation users” and divides L.A.’s streets into two broad categories: Arterial Streets—specifically designed to “carry a large volume of regional through traffic not handled by the freeway system”—and Non-Arterial Streets, which comprise 60 percent of the city’s streets. These Collector Streets, Local Streets, and Hillside Streets have circumscribed purposes “to connect travelers to local residential neighborhoods … and are not intended for cut-through traffic.” Streets are not created equally. This is about flow dynamics and about engineering for disparate volumes and speeds. Arterial Streets are designed to handle far more traffic and speeds ranging from 30 to 35 miles per hour, while Non-Arterial are designed for local traffic and safer, far slower speeds in the 15-to-20-mph range. “Hillside streets are often narrow, winding roadways in residential areas,” states the guide. “They provide access to hillside communities, such as parts of the Silver Lake and Echo Park neighborhoods of Los Angeles.” That is, until. Waze came to town and hacked traffic.

Enter the Waze chat room, and you get a sense of how these unruly map editors are messing with L.A.’s streets. The dialogue includes newbie Level 1 editors on up to seasoned Level 6 area manager “champs.” There’s Spankdog, GooberKing, FurryDaddy, Skullzzcode, and Mikey-Pizzaman. … It feels like a den of old-school gamers from the internet’s 1,200-baud modem days trying to recapture their digital youth. And yet there’s something else going on here. Why the pseudonyms? “They all fear legal liability in case of a bigger accident or traffic situation,” suggested one Level 1 editor, who tried to obtain the email and contact information of a higher-level editor to no avail. The subterfuge is a sign that something isn’t right. If the map editors were serving the public good, why wouldn’t Waze and the editors be transparent? But it’s more than that. This is a masquerade party. These aren’t skilled hackers. They know precious little about traffic management. They’re just mapping clerks trying on the personae of techies, with the bizarre reality that their actions often have catastrophic, citywide ramifications.

There’s Spankdog, GooberKing, FurryDaddy, Skullzzcode, and Mikey-Pizzaman. …Why the pseudonyms? “They all fear legal liability in case of a bigger accident or traffic situation.”

But the Dungeons & Dragons-style fantasy frequently clashes with distraught citizens stumbling in from the real world, pleading for help, interactions torn from a Black Mirror episode. “Waze Uses Dangerous or Low-Capacity Road for Alternate Route” is a popular thread on Waze’s California-area community board. The discussion is illustrated with nightmarish photos and videos shared by residents of a would-be sleepy hilltop neighborhood invaded by the algorithm. No street, certainly not your residential 25-mph variety, is safe from being Wazed into a makeshift freeway or thoroughfare. This has created fascinating potential legal issues that may one day be litigated in lawsuits or class actions. These thousands of homeowners and renters have arguably been injured by Waze’s and Google’s successful privatization of formerly public streets. The videos speak volumes: an 18-wheeler rubbing against a telephone pole on a steep hairpin turn. Sunday-evening standstill traffic on an erstwhile side street. An SUV, trapped in gridlock, jamming the accelerator for a death-defying right turn across three lanes of traffic.

A befuddled Level 3 editor from an area deluged with Waze cut-through commuters floats ideas to undo the damage. He suggests setting a segment at each end of a tiny road as “unpaved.” Or marking it as a “gated community.” Another editor proposes reclassifying it as a “Passageway,” a currently unapproved road type. DwarfLord, a Level 5 Los Angeles-area manager, lays down the law. To resist is futile: “Despite the name, Waze thinks Passageways have the same capacity and safety as freeways,” he opines, inadvertently underlining the madness of humans granting machines the last word, even if the city street guide caps maximum speeds at 10 to 15 mph for these local roads. DwarfLord’s deeper message: The Waze algorithms don’t care about the societal cost they inflict and neither does Waze if the algorithm calculates that a cut-through may save seconds. “The instant the time penalties work out,” DwarfLord claims, “Waze will just as happily send a thousand Wazers down a Passageway as it would [a single] one.” Not only that, but unlike Apple and Google maps, which calculate driving times based on legal speed limits, “Waze takes the speed of what people are actually driving,” a Level 1 editor confides. “If it’s a single-lane residential street and the speed limit is 15 mph, but people are speeding through at 35, they will model the time off that data.” Waze gives the subjective feeling that the Waze route is the shortest. “A Waze editor showed me how the Waze route down my insane, super-curvy street would save you 17 seconds,” the Level 1 adds. “The app tells you to take these crazy roads—it’s the culture of shortcutting, a kind of Hunger Games.”

The Waze algorithms don’t care about the societal cost they inflict and neither does Waze if the algorithm calculates that a cut-through may save seconds.

The upshot? If the algorithm decrees it may save time, however minuscule, no street is safe. And then comes the kicker, where it becomes clear that L.A.’s citizens, even the editors, are powerless, and the game is no longer fun. It’s just impotent, unpaid techie map slaves watching as the algorithm overwhelms its masters like the bucket-wielding brooms in The Sorcerer’s Apprentice. Moans DwarfLord: “Every technique we could try will either have little effect or will have serious drawbacks.” Adds another editor, “The benefit to using the app is gone.”

In April 2015, Mayor Eric Garcetti trumpeted a data-sharing agreement with Waze that would “help reduce congestion, increase safety, and improve driving experience around L.A.” The mayor spoke proudly of this shining example of the city’s “data initiative,” and how the Waze partnership would “get people where they want to go, faster.” The city would share information with the company on traffic accidents, street repairs, and the like; Waze would share its rich. real-time data mined from users about traffic conditions. That was the plan. But traffic worsened. There was a problem with the mayor’s public product endorsement: This wasn’t a partnership. The proof could be seen just two years later, when the city, dissatisfied with the results, chose not to renew the contract. Explained a spokesperson for LADOT: “When the agreement expired, the City Council directed the L.A. Department of Transportation to establish and lead a new pilot program with mapping apps that would allow a more active partnership.” The spokesperson added that the plan was to ensure drivers would be “directed toward roadways … adequate to handle the volume of diverted traffic.” But there was more at play here than the absence of a truly collaborative relationship. The devil was in the details: advanced math, economic theory, the nightmarish complexity of modern traffic in the car capital of the world, and something called a Nash equilibrium.

Alex Bayen, a French-born mathematician who earned his Ph.D at Stanford with a thesis on automating air traffic, is the director of the Institute of Transportation Studies at UC Berkeley. His team of doctorates helped develop the first traffic apps for Nokia more than a decade ago. Bayen understood Waze’s early appeal. Sales of iPhones and Androids had yet to fully take off in 2011. “The knowledge of what Waze was offering was known by only a small percentage of people,” he says, “and that gave the opportunity to come up with an interesting shortcut.”

Getty Images

Then the math kicked in, and the shortcuts quickly became everyone cuts. By 2014 there were 171 million smartphone users in the U.S. The number of Uber and Lyft drivers in Los Angeles—virtually all reliant on traffic apps—exploded. Angelenos started crying foul. “It’s like a cancer,” says Bayen. “It takes some time for the doctor to see it.” The professor “knew it was going to happen from theory because we were studying the problem in computer simulations.” Many dismissed the predictions as “the problem of academics. ‘It’s not real.’” But the minutes once gained by using Waze dwindled to seconds or disappeared. Traffic throughout L.A. worsened, and Waze, in particular, with its rebellious brand built on diverting drivers onto side streets, created what transportation wonks call “negative externalities”—and not just in the traffic-congested neighborhoods. Ironically, when too many vehicles ditch the freeway at once, the exit lanes back up, clogging the freeway itself.

The Waze problem, Bayen argues, today applies equally to Google, Apple, and any other routing agents, in part because they’re all competing to get you to a final destination while selling the often illusory promise of speed. “None of them have any incentive to give you anything else than the fastest path. If they did not give you the fastest path, you would switch apps,” says Bayen. “There’s no incentive to do something societally efficient. The only way is to give you selfish routing.”

Then the math kicked in, and the shortcuts quickly became everyone cuts.

But John Nash, the brilliant economist and hero of the 2001 film A Beautiful Mind, demonstrated that when you give people “selfish solutions” in competitive, “noncooperative games,” everyone loses. Counter to the hyped PR tech narrative, Waze, Google, and Apple were all “steering toward a Nash equilibrium, and that is not a good thing,” argues Bayen. By promoting the myth that their apps would get you there faster, the tech firms were effectively and collectively creating a perfect traffic storm for everyone. “The overall benefit is less than equitable [in a Nash equilibrium],” says Bayen. “Factor in that there is no incentive for the companies to help society with better routing guidance, and any claim by the corporate PR teams that they are providing better routing guidance is immediately disproved by John Nash.”

When Waze wrecked Tim Grant’s Hollywood Hills neighborhood a few years ago, he and his husband filmed videos shared on YouTube and Twitter accounts they named Wazed and Confused. Tim became a Level 1 editor, removed his residential route from the map and felt immediate gratification. The traffic instantly returned to normal. “It was like turning off the faucet,” he says. But a higher-level editor promptly turned the faucet back on and marked the road off-limits to Level 1s, citing a “malicious editor.” Grant says Waze then got his Twitter account shut down and had a lawyer threaten him.

In August 2018 Tim’s city councilman, David Ryu, sent a letter to the city attorney suggesting legal action against Waze and Google. “Waze has upended our City’s traffic plans, residential neighborhoods, and public safety for far too long,” Ryu wrote. “Their responses have been inadequate and their solutions nonexistent…. If we do nothing, Waze will lead us on a race to the bottom—where traffic plans are ignored and every street is gridlocked.”

The letter got headlines, but Ryu, who declined several attempts to be interviewed for this story, has said little on the subject since. He told constituents that Waze would designate a liaison to work with the city, and then the company stopped returning calls. So complete was the stonewalling that Ryu beseeched constituents who had contacts—any contacts—within the Waze-Googleplex to prevail upon them to get his office’s calls returned. Waze declined to be interviewed for this article or to answer questions. Instead, a Waze spokesperson issued a statement through a San Francisco branding firm explaining that Waze is “committed to partnering with cities and transit authorities,” adding that “we’ve met with officials many times in an effort to help address the city’s traffic issues, and have never declined a request for a meeting.”

Councilman Krekorian has had enough of Big Tech PR spin. He knows who is causing the traffic. He wants it to stop. “Throughout L.A. there are small residential streets that were not designed to accommodate the dramatically increased flow of traffic that these apps are causing,” Krekorian said. “In many cases such streets have become completely gridlocked, defeating the supposed advantages of the apps while ruining the peace and quality of life in entire neighborhoods. Even worse, these apps often route a grossly excessive number of cars onto narrow streets with blind curves and unregulated intersections, and often without sidewalks for pedestrians or room for bicyclists.”

Today, with cut-through hillside routes known to hundreds of thousands of L.A. commuters, it’s no longer clear the Waze problem can be fixed by shutting off the tech faucets. “The genie is out of the bottle—it’s a classic issue of tech getting out in front of public policy,” says Dr. Susan Shaheen, the codirector of UC Berkeley’s Transportation Sustainability Center. “How do you regulate these services for the public good, particularly when access to the data is not good?” One possible solution: taxing driving. “To get the genie back into the bottle may require a mechanism as strong as a road charge,” says Shaheen. “People are going to pay more [to] access the road. Local service streets may not be accessible to the public.”

“The genie is out of the bottle—it’s a classic issue of tech getting out in front of public policy.” —Dr. Susan Shaheen

Krekorian, however, is holding out hope. This spring he led a pilot program to curb Waze traffic on L.A. side streets. LADOT has submitted an initial report with recommendations to the council and is taking the lead on reaching out to the app companies to strike a more evenhanded data-sharing agreement. “We can seize this opportunity to create a new partnership with one or more of these companies,” Krekorian says. “The time is right. We have the ability to provide data to the companies.” The competitor who does reach an agreement, says the councilman, will have “a distinct advantage in the car capital of the world.”

Adds Marcel Porras, the chief sustainability officer for the city’s Department of Transportation: “Mapping app algorithms have had a negative impact with cut-through traffic. The LADOT has put together a matrix of some of the sensitive areas, and the city is directing us to negotiate a pilot program to address these issues. That’s fantastic. We have clear direction. We’re hoping we can find agreement with Waze, Google, Apple, and other mapping firms. If not, [the issues] will come back into the hands of the City Council.”

But it remains far from clear that the tech companies want to share data and alter their algorithms for the public good.

That’s too bad. Because transparency and improved collaboration between tech firms and L.A. might save lives. Only the city has in-the-moment data on planned closures and emergencies. During the Malibu wildfires Waze showed an algorithmic disregard for human life, reminiscent of Hal in 2001: A Space Odyssey, mindlessly sending drivers toward fireballs. Tweeted Andy Ruther at the time: “I don’t trust Waze. It just told me I could save 8 minutes by driving right thru a wildfire off the 405 Freeway.”

Tim Grant, for his part, says bring on the tech but stop the corporate PR slogans. Do it for real, for the community. Not just to sell ads or promote a false narrative of speed. “You know, if they’re going to use big data, use all the data,” says Grant. “Not just, ‘Hey there’s a street here’ but ‘Hey, there’s a street here with blind turns, no sidewalks, people pushing strollers, people walking their dogs. It’s narrow, and it’s just not appropriate.’ Take it off, take it off, take it off!”

RELATED: The Surprising Wisdom of Waze

Stay on top of the latest in L.A. food and culture. Sign up for our newsletters today.