Skip to main content

Google vs. Uber and the race to self-driving taxis

Google vs. Uber and the race to self-driving taxis


It's an old-fashioned, autonomous face-off between two Silicon Valley giants

Share this story

If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement.

You won't be able to have a fully autonomous car of your own for at least a few more years, but the race to develop the best technology is already tearing up the asphalt. Google is reportedly interested in spinning off its self-driving unit as its own self-driving company under the umbrella of Alphabet. And that means it will need to generate revenue, which could mean ride-hailing. And ride-hailing, of course, means Uber.

So where does that leave us? Well, with a lot of questions. Can Google successfully challenge Uber, or is Travis Kalanick's company simply too large and too entrenched to unseat? Does Google even want to beat Uber? Or is it simply looking for a way to make self-driving cars generate cash? And with neither committing to actually manufacturing their own vehicles, which carmakers will each be partnering with to bring its technology to the road?

Would we Google an Uber? Or Uber a Google?

According to Bloomberg Business, Google plans to start small by deploying its multi-sized autonomous fleet in confined areas like college campuses, military bases, and corporate office parks. This is smart, because it would likely help Google get additional miles under its belt before rolling out its self-driving cars onto city streets.

Those streets fall squarely in Uber's turf — hard-won turf. It took the company five years and countless regulatory battles to expand to the point where it now covers 75 percent of the US population. It has 1.1 million active drivers in 361 cities worldwide, and delivers 3 million rides a day. It has joined the ranks of companies, like Google, that have successfully achieved verb status: you don't call a car, you "uber" a car.

Uber NYC stock 1020 2

But while Uber is dominating the ride-hailing business, Google has the advantage in autonomous car research. It has been developing the technology since 2009 (and employs staff who've been working on it for far longer than that). The company's self-driving cars have logged over a million miles on the streets of Mountain View, California, and Austin, Texas. Uber only got around to opening its own self-driving lab in Pittsburgh earlier this year.

But who isn't developing self-driving technology these days? Tesla's Autopilot system is already on the road in beta form. Countless other automakers have semi-autonomous systems in the works — like GM's Super Cruise, debuting next year — and many are pointing to 2020 as a turning point for fully self-driving rides. Tesla, Nissan, Mercedes Benz, Nissan, BMW, and others are all permitted to test self-driving cars in California, which has become a hotbed for autonomous testing in recent years.

Google has the advantage, but Uber is hot on its heels

Google still has the advantage, and Uber recognizes that, which may explain why the company is so aggressively pursuing Google's engineers and executives. High-definition maps are key to a successful self-driving car project; Uber is building up its mapping division in parallel to its self-driving unit, the former staffed with ex-Google employees and the latter with Carnegie Mellon University engineers. Google Maps product manager Manik Gupta jumped to Uber's mapping division earlier this month. His new boss is Brian McClendon, the former head of Google's entire mapping division.

The staffing decisions seem to be a reflection of CEO Travis Kalanick's fears that self-driving cars could put his company out of business. "That Uber, a five-year-old company has the potential to be disrupted if we don't do this right, is super fascinating," he said at October's WSJDLive Conference. Asked about Google's self-driving success rates, he said, "Getting Google's cars to a 90 percent solution is going to happen soon." But he predicted Google wouldn't get to a "99.99 percent success level" for another 15 or 20 years.

Google, on the other hand, recognizes Kalanick as a serious threat in his own right. "The thing you have to understand about Travis is that he is the definition of a serial entrepreneur in its purest form, with all the strengths and weaknesses that comes with," Eric Schmidt, executive chairman of Alphabet, told Time this month for Kalanick's Person of the Year runner-up profile. "He's a fighter. He is against institutional structures. He has to get up every day and make something. He can be disagreeable in that sense that, well, he disagrees."

Or is it just a media-generated tempest in a glovebox?

Google Ventures invested $258 million in Uber in 2013, an investment so large at the time for both companies that it hilariously led tech reporters to speculate that Uber was unlikely to need to raise any more funds before going public. Two years later, Uber is valued at $62.5 billion, with no signs of an IPO or slowing down.

Experts believe ride-hailing to be the most obvious application of fully autonomous vehicles. And surveys suggest that consumers are excited about the technology, but also concerned about safety, security, and, of course, how much it will cost them. Licensing and regulation are also serious hurdles, and Uber — its wall decorated with the scalps of politicians who dared to oppose it — has an advantage over Google in that regard.

Of course, a lot of this could be a media-generated tempest in a glovebox. Uber's policy is not to comment on other companies, and you could argue that there really isn't anything new here. After all, Google said it was interested in commercializing self-driving technology back in September when it announced John Krafcik, the ex-CEO of Hyundai and former president of TrueCar, would run its self-driving car division.

In fact, we've known for almost two years that Google was interested in using its self-driving technology as a ride-hailing service. In 2014, Claire Hughes Johnson, then an executive at Google's self driving car project, gave a speech that suggested ride-hailing was the most obvious application. "What if you all got here today in a self-driving car that dropped you off and then left?" she asked the audience. "So you may not be able to buy one, but you may be able to drive in one in the next five years."

The question, then, may be whether Alphabet needs to get into the ride-hailing business itself — a daunting task, even without the specter of Uber — or whether it makes money in the long term through partnerships and the licensing of its technology. After all, Kalanick, ever the businessman and opportunist, is said to have mentioned not long ago that he'd buy a half-million self-driving Teslas if he could; why not adorable little (and possibly cheaper) Google cars instead?