Super Pumped : The Battle for Uber (9780393652253)
Page 20
Brazil was in upheaval when Uber arrived in 2015. Unemployment was at an all-time high, and violent crime and murder rates across Brazil were skyrocketing. While the lack of jobs meant many more Brazilians were willing to drive for Uber, the cash bankroll of each day’s earnings made them a tempting target for thieves. At least sixteen drivers were murdered in Brazil before Kalanick’s product team improved identity verification and security in the app.
Kalanick and the other executives at Uber were not indifferent to the danger drivers faced in emerging markets. But they had major blind spots because of their fixation on growth, and their casual application of financial incentives often enflamed existing socio-cultural problems. Kalanick believed that there were things inherent in the Uber software that made it safer than a regular taxi, namely that the rides were recorded and trackable by GPS. He further hoped Uber could fix the problem of driver safety through more tech solutions.
But Sullivan saw all this and knew he needed to act fast. He would build a world-class security organization, divided into branches to handle threats from financial fraud, to digital espionage to physical security. He requested hundreds of staffers—security engineers to handle Uber’s systems, ex-CIA and NSA types on contract to handle on-the-ground operations and field investigations, and many others. Kalanick agreed, and gave Joe Sullivan a blank check.
But Kalanick had one very important requirement: Uber wouldn’t only play defense.
Chapter 18
CLASH OF THE SELF-DRIVING CARS
Travis Kalanick was fuming in the grand ballroom of the Terranea Resort—a seaside haven for the rich off the coast of Rancho Palos Verdes, California. It was the opening night of the 2014 annual Code Conference, a confab for the tech elite. On stage, Sergey Brin was in the middle of a historic speech, but Kalanick was on his iPhone firing off messages to David Drummond. Brin—who was ostensibly Kalanick’s partner and investor—had just unveiled something that could threaten Uber’s existence: a fully autonomous self-driving car.
“The reason I’m excited for this self-driving car project is the ability for it to change the world around you,” Brin told the audience. The technologists, venture capitalists, and journalists were buzzing with excitement. The Google co-founder showed up that night as the keynote speaker in a white T-shirt, black pants, and a worn pair of Crocs. Brin preferred comfort over style.
As the video played, the audience saw an egg-shaped, stark white two-seater vehicle doing laps around a parking lot. It was ugly and small. The front of the vehicle looked like a smiley face, as if Humpty Dumpty had turned into a golf cart. Blade Runner this was not.
None of that mattered. The car didn’t need a steering wheel, so it could be any shape. In the Dumpty-mobile sat two people. Neither did anything to drive the car as it zipped effortlessly around a Mountain View parking lot. As far as Kalanick was concerned, Google’s egg-shaped, self-driving monstrosity was a work of art.
Google, long considered an ally and a partner, seemed to be turning on him. Google’s little car would destroy Uber, and would do it smiling. If Google had a ride-sharing service that didn’t need drivers, they could charge almost nothing, steal all of Uber’s customers, and destroy its business.
Brin was being interviewed on stage by the journalist Kara Swisher, who ran the Code Conference. She asked him point blank if Google had plans to ever create a ride-hailing service, like Uber. Kalanick might have hoped to hear Brin deny it. He didn’t.
“I think some of these kinds of business questions—how will the service be operated, will we operate it ourselves, will we work with partners—are things that we’ll sort out when it’s closer to being widely deployed,” Brin told Swisher, noncommittal. “I think that these initial test vehicles, we’ll probably just operate a service ourselves because it’s going to be a very specialized thing. But longer term, it’s not clear.”
Kalanick was irate. Uber was becoming a major force—both as a tech company and a transportation machine—but it didn’t have a fully autonomous vehicle. It wasn’t even researching one.
Kalanick still believed and acted as if Uber was the underdog wherever it went—a frame of mind that wouldn’t change for his entire tenure as CEO. In the beginning, it was Uber against the greedy, unethical taxi companies who had the sleazy local politicians in their pockets. Later, it was Uber against Lyft, the well-funded startup whose warm and fuzzy branding—those pink mustaches—was just cover for its ruthless executives. And now, it looked like it was going to be Uber against Google, the global corporate technology giant.
His anger turned slowly to fear. Google’s search advertising business practically minted money. That gave Google the freedom to pursue wild projects, even if they lost money or, in some cases, were patently absurd.¶¶¶¶¶ Google’s self-driving car research—years in the making by that point—would be a major cost center for almost any other company in Silicon Valley. For Google, it was a rounding error.
As Kalanick would later tell friends, it was after the 2014 Code Conference that he started sweating. As Brin left the stage, Kalanick kept sending frantic texts and emails.
He needed to talk to David Drummond.
Drummond, as it turned out, was expecting Kalanick’s messages.
Google’s Gulfstream V left the tarmac at San Francisco International Airport the Tuesday after the Memorial Day weekend bound for Los Angeles. Google executives heading to the Code Conference had mulled over how they would tell Kalanick about Brin’s onstage demonstration, which would occur that evening. They decided it made the most sense that Drummond, who sat on Uber’s board, should break the news to Kalanick.
Drummond usually knew how to handle these situations: with empathy. Tall and well-built, he could pass for a fit ex-linebacker. But his hazel eyes and toothy grin made him look harmless, like the smart and affable corporate lawyer he was. Those qualities, in addition to being one of the few African-American executives who had reached the top rungs of the Silicon Valley ladder, made Drummond stand out among his peers. But for all his skill and confidence, Drummond was conflict averse, which kept him from telling Kalanick about Google’s plans until the eleventh hour.
Drummond already knew it was a touchy subject. Kalanick had spies all over the Valley. Uber’s “competitive intelligence” operation—that is, the sprawling, systematic COIN program led by Joe Sullivan and his lieutenant, Mat Henley—grew larger by the day. Kalanick often heard whispers of Google’s self-driving car project, or occasionally, an errant rumor that Google was starting a self-driving taxi service. Every time Kalanick would hear a rumor like this, he’d fire off an email to Drummond.
“We get stuff like this more than I would like,” Kalanick once wrote to Drummond, forwarding intel about a Google self-driving car service. “A meeting with Larry [Page] could calm this down if it’s not true but he has been avoiding any meeting with me since last fall. Without any dialogue we get pushed into the assumption that Google is competing in the short term and has probably been planning to do so for quite a bit longer than has been let on,” Kalanick continued. It went that way for months. Something would pop up, Drummond would smooth things over, and everything would go back to normal until the next rumor came along.
On the day of the Code Conference, Drummond finally called Kalanick and gave him a heads-up about the demo. People familiar with the call would later describe it as tense; Kalanick was understandably upset. He felt betrayed by his own backers.
After Brin’s session came to a close, Drummond asked Kalanick to take a walk with him around the Terranea. Kalanick was melting down, but Drummond laid on the platitudes even thicker than usual, according to someone familiar with the conversation. As a partnerships and biz dev guy, Drummond knew how to do the kind of handling and soothing that his bosses, Page and Brin, never bothered to do.
Kalanick tried to cool off, wanting to believe Drummond. The executive was on Uber’s board of directors. His company had invested hundreds of millions in Uber’s future.
For Kalanick, there was hope, indeed, that Drummond was actually telling the truth.
But later that evening, the idea of Kalanick “remaining calm” flew out the window. Every year on opening night at the Code Conference, the organizers threw a large seaside dinner for the attendees. The most powerful chief executives, however, joined a private dinner elsewhere on the hotel’s campus. That year, Kalanick had scored an invite and brought his girlfriend Gabi Holzwarth, a charming musician and dancer. The two of them were introduced by chance through Shervin Pishevar, an early Uber investor and friend of Kalanick.
Holzwarth, then twenty-four, was a classically trained violinist who grew up studying music. She often played in public as a street performer in San Francisco and Palo Alto, where she was raised. When Pishevar ran into Holzwarth performing music in front of a candy store, he hired her to perform at a fundraiser for Cory Booker, held at Pishevar’s house, where she and Kalanick first met.
Holzwarth possessed a fierce spirit and pursued the arts from a young age. Kalanick loved that fiery spirit, warm personality, resilient attitude, and the way she could speak to pretty much anyone. As Kalanick’s star rose, so did the couple’s profile. They attended dozens of high-level parties together—the Time 100 Gala, the Vanity Fair Oscar Party, the Met Gala—he in his tuxedo, she in designer ball gowns.
At the Code Conference’s private dinner, Kalanick and Holzwarth were seated with the heavy hitters; powerful CEOs of the Valley’s biggest companies. He should have enjoyed his meal—this was the ultimate validation of Kalanick’s success and influence. Instead, he spent most of the evening watching Sergey Brin chat up his girlfriend.
Holzwarth was polite, good at conversing with even the most awkward engineer. But Brin, who was going through a messy, public divorce after having an affair with one of his employees, ignored Kalanick, oblivious to the optics. Before the meal ended, Kalanick snapped an iPhone photo of Brin’s cozy-looking chat with Holzwarth and texted it to Drummond. He later told Drummond he saw Brin place his hand on Holzwarth’s leg, and that he believed that Brin’s behavior was a liability for Google.
Despite Drummond’s talents, there was no smoothing over what happened that night. After the dinner, Brin asked Holzwarth to hang out with him and talk by the pool later that evening. Kalanick stewed. Google was going out of its way to screw him. And now he watched as the man who was going to kill his company tried to steal his girlfriend.
When it came to self-driving cars, Kalanick was further behind than he could truly appreciate. Larry Page—a transportation obsessive—had invested more than a billion dollars and tens of thousands of employee hours into the problem by the time he and Sergey felt comfortable enough to show their egg-shaped car to the world. No one was more determined to bring robot cars to life than Larry Page.
No one except, perhaps, for Anthony Levandowski. The lanky, cantankerous engineer was still working on “Project Chauffeur,” the company’s pet name for autonomous vehicle research. But his position was growing tenuous.
For one, Levandowski was a poor leader. He would get in constant fights with colleagues over the speed—or lack thereof—at which Google was willing to work on the self-driving project. But Page liked the fact that Levandowski didn’t always play by the rules. Men like Levandowski, Page believed, would bring Google to the next phase of autonomous vehicle research.
Levandowski had a certain effect on Page, too. While often divisive, Levandowski could be charming. The two men would dine together occasionally—that in itself a peculiar activity for Page—imagining a future driven by robotic cars. Whatever flaws Levandowski had, Page needed him in the Googleplex.
But by 2015, Page’s personal attention and millions of dollars in bonuses wasn’t enough to keep the golden boy of autonomy happy. Levandowski was tired of his risk-averse colleagues. He was tired of being told “No.” Google never seemed comfortable with the project, and dragged its feet, he thought. Levandowski believed they could do better. That he could do better.
And so he started pitching a few of his trusted Google colleagues on a new idea: long-haul trucking, a space whose last great innovation was No-Doz. He practiced his pitch on co-workers at dinners he hosted off-campus: imagine a world where self-driven trucks moved goods constantly from city to city. A world in which the sleep-deprived truck driver was no longer a threat. Trucking was an enormous industry, employing 7.4 million Americans and creating $738.9 billion in revenue each year. Trucks drive 5.6 percent of all vehicle miles in the United States, according to Department of Transportation data, and are responsible for some 10 percent of highway fatalities. Automating it would be worth billions. It helped that self-driving trucks didn’t directly challenge Google. At least, that was what Levandowski told his colleagues. They would call it Ottomotto—Otto, for short.
By 2016, Levandowski had left Google, taking with him a small cadre of colleagues, including a close partner named Lior Ron, who worked for years on Google’s popular Maps software. Levandowski made his feelings clear in a final email he sent to Page: “I want to be in the driver seat, not the passenger seat, and right now [it] feels like I’m in the trunk.”
Less than six months later, in the summer of the 2016 presidential election, Otto was up and running. Levandowski and Ron, who became Otto’s co-founder, had already expanded the startup to forty-one employees. They had logged more than 10,000 miles on the road with their experimental equipment in three Volvo trucks. With them came fifteen former Googlers, more than half of them autonomous vehicle engineering specialists, a rare and valuable breed in Silicon Valley.
In a rare move, Otto took no venture capital at all. The entire group of Xooglers—the preferred noun for “ex-Googlers”—were rich, and could afford to fund the project themselves. Levandowski was richest of all, having made millions off the sale of his companies to Google years earlier.
But Otto’s secret weapon wasn’t the well-lined pockets of its founding team, nor the foresight to charge into an open field. It was that Levandowski was finally free of Google’s corporate and legal bureaucracy. Now he could do things his way. Back at Google, he was scolded for breaking rules and bending regulations. At Otto, Levandowski had no such restrictions.
When the startup was ready to film a demo of its self-driving hardware kit—which could be fitted, off-the-shelf, to existing big-rig trucks—Levandowski called the lobbyist who convinced Nevada regulators to write a new law for Google self-driving cars, and had him request a permit for Otto to film on a stretch of highway in the state. After the Nevada Department of Motor Vehicles declined his request, Levandowski ignored them and filmed it anyway. The sweeping, aerial views of a stark-white eighteen-wheeler dotted with Otto’s black signage stood out marvelously against the warm tones of the Mojave Desert. A regulator grumbled that Levandowski’s move was illegal. He never faced any actual consequences.
For Levandowski, it was worth it: everyone who saw the launch video loved it. If he had played by the rules, as he had mostly done at Google, he’d still be waiting for approval. Inside Otto, engineers printed out orange-colored stickers and pasted them around the San Francisco headquarters with a message they knew Levandowski would love: “Safety Third.”
Their meeting felt almost predestined.
Travis Kalanick and Anthony Levandowski were first introduced to one another in 2015 at the suggestion of Sebastian Thrun, a former Google executive and bigwig in the world of self-driving cars. Soon after, as Levandowski was preparing to leave Google and start anew, he began meeting with the Uber CEO in secret.
The two men clicked immediately. Levandowski, a born futurist and affable, six-foot-seven showman, roused something inside of Kalanick. The men, both in their forties, imagined a future filled with self-driving vehicles, with Levandowski’s engineering talent fueled by Kalanick’s enormous ride-hailing network. In Levandowski, Kalanick felt he had found a “brother from another mother,” he’d later say.
That first meeting developed into a series of clandestin
e conversations. Levandowski went to work for Google during the day in Mountain View, and would then return to San Francisco in the evening to meet Kalanick and talk about their future partnership. To keep from attracting attention, they would arrive separately at San Francisco’s Ferry Building, a beloved city landmark. After each picked up a bag of takeout, they’d walk north and westward, up the pier and toward the Golden Gate Bridge, where they’d begin discussing their self-driving dreams.
Kalanick knew almost nothing about autonomous tech, but Levandowski filled him in on technical details. A self-driving car needed an enormous amount of equipment just to understand the terrain it traveled, much less navigate it safely. The chassis was fitted with lasers, 360-degree cameras, an array of sensors, radar beacons. Lidar, short for “light detection and ranging,” helped the car’s software absorb terabytes of data about the landscape.
Anthony’s laser “is the sauce,” Kalanick once wrote on a whiteboard in a meeting. They would spend hours making up code names and communicating in secret slang. If the autonomous vehicles could drive themselves, Kalanick mused, they could create a “Super Duper” version of Uber, or “Uber Super Duper.” Instead of taking 30 percent of driver earnings—the company’s current business model—Uber would instead take the entire fare. That meant billions upon billions more in revenue. The abbreviation, USD, lent itself to a cool internal codename: “$,” a simple dollar sign.
They talked like teenagers obsessed with a science project. Kalanick would arrive home buzzing with excitement after a meeting with Levandowski, waving his phone in his girlfriend’s face. “Look how far we went this time!” Kalanick said, noting the steps his iPhone pedometer had tracked across the city.
Levandowski would eventually create a separate company, Otto, right after leaving Google, as if he were interested in pursuing his own trucking startup. Then he would take venture investment meetings up and down Sand Hill Road—the famed home of top-tier Silicon Valley venture capitalists, including Andreessen Horowitz and Kleiner Perkins—to drum up money for the endeavor. But he would shrug those meetings off, opting instead to raise no outside capital. (Since they were operating independently mostly for appearance’s sake, it made no sense to fork over equity to outside investors.) Then came the coup: Uber would acquire Otto for millions, a grand proclamation of Uber’s intentions to pursue self-driving technology.