By: Michael Feldman

Uber has taken another definitive step toward eliminating human drivers from its fleet of vehicles.  This week, the ride service pioneer created a research division, known as AI Labs, while simultaneously acquiring machine learning startup Geometric Intelligence for an undisclosed sum.

The new lab will become the incubator for products and technology that Uber can use to further automate the operation of its vehicles and service. The announcement summed it up thusly:

“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…”

The new lab will be initially populated by the 15 employees that made up Geometric Intelligence, a New York-based startup that assembled a team of young AI university researchers and enthusiasts. Heading the new Uber lab will be Geometric Intelligence founding CEO Gary Markus, a professor at NYU. Although the startup says it has patents pending, it’s noteworthy that the company has no product in the offing and has not published any research, at least not in the public domain.

One area the Geometric Intelligence team appears to be developing is a Bayesian approach to machine learning, which can be used to detect patterns using relatively few examples (sometimes as few as one).  In a more typical deep learning approach, the model requires reams of data consisting of thousands or even millions of examples of what the application is trying to identify. Being able to train a model with a small number of examples has a real utility in navigating real-world driving environments, and could provide Uber with a competitive advantage against its rivals in the autonomous vehicle space.

But the addition of the AI Labs division could have even more wide-ranging goals.  According to a write-up in Wired, the move by Uber is part of a more ambitious strategy to establish itself as an AI company, which would allow it to branch out into other transportation markets – and even into other industries. The article quotes Oren Etzioni, the CEO of the Allen Institute for AI, who says “They want to join the Big Four,” referring to Google, Amazon, Facebook, and Apple.

Uber, a company that is estimated to be worth between $60 to $70 billion, certainly has the resources to swim in the deep end of the AI pool, should it so desire. Of course, catching up to the likes of Google or Facebook will take some doing. The good news for Uber and like-minded companies is that there are hundreds of AI startups available for the taking – 1,685 currently on AngelList -- with more created every day. For businesses looking to add AI technology or talent to their enterprise, it’s a buyer’s market.