BMW and Uber ratcheted up their war over the weekend as BMW announced its own self-driving car testing in Germany and Uber announced the acquisition of AI startup Geometric Intelligence. The move by BMW to keep up with the ride-sharing titan indicates not merely that Uber’s progress has rattled traditional car companies, but also that company’s pivot toward being an AI services company is moving rapidly.
News that the German manufacturer will start testing their own autonomous vehicles on the streets of Munich marks the second announcement within a week regarding regional testing of self-driving cars following the publication of a test run in Waterloo, Ontario last week that was led by the University of Waterloo’s Centre for Automotive Research, with participation from BlackBerry.
Forty cars will turn the center of the Bavarian capital into a testbed, according to Reuters, while mirroring the techniques being used by Uber in Pittsburgh. A specially-trained driver will be ready to take over any wheel, according to Klaus Buettner, BMW’s VP of autonomous driving operations.
BMW had announced a tripartite deal with Mobileye and Intel to develop a fully autonomous car by 2021, but the executive team seems to be going full-speed ahead.
BMW has made moves to enter the vehicle services market with a number of their own ventures, including car-sharing DriveNow, space-finding app ParkNow, electric charging network ChargeNow, and their ridesharing service ReachNow (operating in Seattle). But none of that will much matter unless they can keep up with Uber in Pittsburgh or NuTonomy in Singapore and Boston. There is also American rideshare giant Lyft who raised $500 million from General Motors (GM) in January 2016 with an aim toward developing its own autonomous autos.
“Uber and Lyft do not operate their own fleets of cars. Owning the fleet means you can make offers that Lyft and others are unable to provide,” BMW’s Chief Executive Harald Krueger told Reuters.
But that point might be moot as Uber seemingly makes very rapid strides toward becoming first an artificial intelligence and machine services company and only a ride-hailing service second. Based out of New York, Geometric Intelligence’s entire 15-person team will now transition and constitute the core of the new Uber AI Labs in San Francisco. The startup’s named investors before the acquisition were Adam D’Angelo, Drew Houston, Fontinalis Partners, Jaan Tallinn, Jeremy Stoppelman, Liquid 2 Ventures, Steve Blank, Mark Achbar, and Esther Dyson.
“In spite of notable wins with machine learning in recent years, we are still very much in the early innings of machine intelligence,” Uber Chief Product Officer Jeff Holden blogged. AI Labs will be led by Geometric Intelligence CEO Gary Marcus and include the company’s CTO Doug Bemis, CSO Ken Stanley, and CSO Zoubin Gharamani.
This will be Uber’s second major AI acquisition in the second half of 2016 after it dropped $680 million on Israeli self-driving truck company Otto.
While rumors that tech giants have shown weariness about advancing their autonomous vehicle projects, traditional manufacturers are picking up the slack. The sheer magnitude of building a vehicle manufacturing and testing infrastructure at scale that could catch up to established American, European, and Asian car giants might intimidate some of Silicon Valley’s technology giants to the backburner on autonomous autos. Apple in October reassigned staff working on the project but seemed to confirm this week they were still pursuing their own vehicle, making good on a $1 billion investment in Chinese rideshare company Didi.
Lyft and Uber have avoided that by strengthening their own alliances with corporations that know the vehicle market and have the manufacturing infrastructure to produce self-driving vehicles. Ford announced in August it would develop an autonomous car for the private market within five years. GM’s deal with Lyft also makes that clear.
“Uber is in the business of using technology to move people and things in the real world,” Holden continued. “With all of its complexity and uncertainty, negotiating the real world is a high-order intelligence problem. It manifests in myriad ways, from determining an optimal route to computing when your car or UberEATS order will arrive to matching riders for uberPOOL.”
“It extends to teaching a self-driven machine to safely and autonomously navigate the world, whether a car on the roads or an aircraft through busy airspace or new types of robotic devices.”
While much is being made about Uber’s moves with autonomous cars for the purpose of TaaS — Transportation as a Service — it has been lost on people that a lot the work in maintaining success solely as a ferrying company for individuals is definitely an unproven business strategy. Traditional taxi companies and startup darling ridesharers have always struggled to make a profit.
Uber and some of the world’s most influential technology insiders have been unrelenting in prophesying a new transportation economy that does not have mass personal car ownership at its core, but they have yet to prove that mobility service companies will be able to make a profit to operate that sort of infrastructure.
But Uber’s business is not going to solely depend on ride-sharing or vehicle manufacturing and repair. It will be AI, and a complex kind at that. Just as Amazon expanded from selling books to selling everything and now sponsoring some of the best innovators in the tech world, so too will companies like Uber expand their own offerings in order to remain relevant and (eventually) gain a profit without burning through literally billions of dollars worth of cash.
If there was any debate about Uber’s role in Silicon Valley as a technology company, this acquisition and BMW’s coincidental announcement to keep up with the startup juggernaut should put those doubts to bed. Uber is making a push for survival beyond just low-priced 3:00 AM cab rides for drunk bar patrons and extensive, expensive legal battles with city governments around the world. They have pivoted, and their AI solutions will likely find their way into industries well beyond grounded and wheeled vehicles.