On Monday, Uber sold the Advanced Technology Group—its autonomous vehicle division—after spending six years pitching the public, investors, regulators, and traditional automakers on the supposed inevitability of this technology and its potential to transform Uber.
TechCrunch reported that Aurora, a self-driving start-up, was essentially being paid to take ATG; Uber will be handing over its 86.2 percent ATG stake to Aurora and investing $400 million for a 26 percent stake. The deal, TechCrunch writes, marks one of the last expensive pursuits that [Uber] had yet to either spin or sell off" as it refocuses on ride-hailing and delivery. On top of selling off ATG, Uber also sold its Elevate division to Joby Elevation, essentially paying the startup $75 million to take over Uber’s “aerial mobility” operations that sought to make flying taxis and helicopter rides a reality.
The sales are a huge heel turn from Uber’s own actions and messaging for the past decade, bringing the company back to square one: stuck with unprofitable ride-hailing and delivery operations that require significant subsidies to attract and retain customers.
Under Uber’s previous chief executive, Travis Kalanick, driverless cars were touted as the key to unlocking Uber’s value and eliminating its main costs (driver wages and incentives, for example).
"If we don't get the [autonomous car] software thing nailed, we're not going to be around much longer," Kalanick told USA Today in a 2016 interview. "Will it all take time and storytelling to reassure consumers? Yes. But that's where it ultimately ends up."
At a 2014 RECODE conference, Kalanick said he saw fares as too high because passengers were paying for "the other dude in the car" (the driver). Driverless cars, Kalanick maintained, would not only solve that problem, but end private car ownership—a talking point regularly repeated over the years by Kalanick many times, general managers, Obama’s former campaign manager turned Uber executive David Plouffe, and even Uber’s current CEO Dara Khosrowshahi who has said "one of Uber's goals is to reduce the need for individual car ownership," specifically through self-driving cars and UberPool.
The amount of time and money Uber wasted on its doomed self-driving lark is staggering. Despite acquiring Otto, a self-driving truck start-up, in 2016 and valuing it at $680 million, allegedly stealing trade secrets from Google's self-driving unit, poaching 40 engineers from Carnegie Mellon to found ATG, receiving a $500 million investment from Toyota and another $1 billion from SoftBank's Vision Fund (valuing ATG at $7.25 billion), and despite constant assurances that "it would become profitable by 2021," investors pressured Uber CEO Dara Khosrowshahi to sell.
But while this communicates how much money was wasted on ATG, it obscures how single-minded Uber’s pursuit was and how little evidence there was to back the unsubstantiated claims it made anyone who’d listen.
Both in hindsight and at the time of Uber’s ATG, there was no concrete evidence that driverless cars were on the horizon, let alone made any economic sense. In a conversation with Motherboard in 2017, the new director of the Carnegie Mellon robotics lab that Uber raided to set up ATG in the first place said that driverless car technology was “not even close.” Some advances have been made in the intervening years, but fully autonomous cars that are safe enough to drive around cities at scale are still a ways off.
And as the Financial Times' Alphaville pointed out back in 2015, the economics of self-driving taxis make no sense. Uber has always been, at best, a low-margin business that depends "heavily on being able to overstep regulation, licensing and—most importantly—transfer maintenance, cleaning, insurance and market-risk exposure to drivers." If cars change from personal assets that spend nearly 90 percent of their time idle to investments that yield revenues, Alphaville asked: "Who exactly would fund and manage a driverless fleet on behalf of Uber?"
In 2019, an MIT study reaffirmed FT’s early concerns and found that even if autonomous taxis were occupied by a passenger 100 percent of the time, they would be "fiscally uncompetitive" with vehicles that had a driver. In San Francisco, the MIT study estimated driverless taxis might cost anywhere from $1.58 to $6.01 per mile—the Bureau of Transportation pegs the cost of personal car ownership at $0.59 per mile.
During testimony for its Waymo lawsuit over alleged trade secrets theft, internal documents were revealed that showed Uber was selling its own snake oil internally. One January 2016 presentation projected driverless cars would be profitable for Uber by 2018, while in May a report anticipated "13,000 self-driving taxis by 2019" and by September that number "jumped to 75,000 vehicles." Uber's then-head of self-driving technology, Eirc Meyhofer, testified these were nothing more than "highly speculative" and best understood as "assumptions and estimates."
The same might be said for Elevate, a program that began in 2016 to create flying taxis that would be “more economical than owning a car” and was championed as a cornerstone of Uber’s strategy. It, too, was sold along with ATG.
As Uber’s 2019 IPO loomed closer and closer, Khosrowshahi began to talk about Uber being an "Amazon of transportation" that would serve every transportation need. In two separate interviews with The Verge, Khosrowshahi hammered home his argument that “[Uber’s] future is not just cars.” For the chief executive, the focus is on “Uber moving from being a giant global car-sharing platform moving to a global A-to-B urban mobility platform.”
This gave way to even grander rhetoric as Khosrowshahi began to insist that “cars are to us what books are to Amazon.” At one San Francisco-based conference that year, Khosrowshahi predicted ride-hailing would be "less than 50 percent" of Uber's business in terms of transactions and the core business would be "the Amazon of transportation." The pre-IPO argument went like this: Uber’s lack of profits were, like Amazon’s, temporary and a result of their investment in future growth. If Uber were given the chance to develop enough products and services, it would explosively grow into profitability.
Now that the IPO is done, the rhetoric has changed sharply. By November 2019, six months after its IPO, the tone had changed. In an infamous interview with Axios, he revealed that "Autonomous is a long way off, it's probably 5 to 10 years off." Khosrowshahi instead pivoted to the idea that "very simple routes that are predictable—repetitive" would see driverless cars sometime "in the next 3 to 5 years." While it’s not clear whether autonomous vehicles would be economically viable for car owners, there’s almost no scenario where partially autonomous vehicles make any sense to deploy.
“When Dara Khosrowshahi was hired in 2017 his singular objective was to complete the high-value IPO that Travis Kalanick had failed to produce,” transportation economist Hubert Horan told Motherboard. “The terrible economics of Uber’s peripheral businesses/speculative investments (AVs, urban logistics, intercity trucking, Uberpool, scooters, transit services, flying cars) should have been blisteringly obvious. The obvious problem that spending on these ventures posed for a now-public company (cutting spending would destroy the “long-term growth” narrative while increasing spending would trash quarterly financial results) should have been equally obvious.”
Now that Uber has sold just about every single line of business it used to attract capital and boost its valuation, this era of Uber should go down as being one of the greatest delusions of grandeur in corporate history. Uber is back at square one with a set of core, yet still unprofitable, products. The company has never made a profit and likely never will, no matter how many laws it rewrites because its core operations have abysmal unit economics that cannot be addressed at the national, let alone global, scale Uber desires.
It’s hard to see how any of this was worthwhile for anyone but the investors. Since Prop 22, Uber’s market value has risen over $30 billion and finally yielded a return for every pre-IPO investor. In the meantime, Uber has fought for the right to perpetually cut the wages and livelihoods of misclassified drivers who rely on Uber. Passengers, who stay for the cheap rides, have seen fares increase since the IPO—especially those who with trips leaving or going to non-white neighborhoods—and are already paying a premium as high as 91 percent for Uber Eats.
“After offloading/shutting down AV, trucking, logistics, Uberpool, scooters and flying cars, Uber is back to where it was in 2013—operating taxi services that never had any hope of sustainable profits, plus some delivery services using those drivers and cars that even in a pandemic are less profitable than the original taxi business,” Horan said.