ADVERTISEMENT

Uber Creates AI Lab, Buying Startup Geometric Intelligence

Uber Creates AI Lab, Buying Startup Geometric Intelligence

Uber Creates AI Lab, Buying Startup Geometric Intelligence
Attendees wait for rides in the Uber Technologies Inc. area during the 2016 Consumer Electronics Show (CES) in Las Vegas, Nevada, U.S. (Photographer: Patrick T. Fallon/Bloomberg)  

(Bloomberg) -- Uber Technologies Inc. agreed to buy startup Geometric Intelligence Inc. to help the ride-hailing company create a new artificial-intelligence research lab. 

The startup had 15 employees and was developing new machine-learning techniques inspired by cognitive science that rely on less data. Chief Executive Officer Gary Marcus, an author and professor of psychology and neural science at New York University, will head the new research effort, called Uber AI Labs.

The lab will report to Jeff Holden, Uber's chief product officer. It will try to improve the algorithms Uber uses to match drivers with passengers, and seek new techniques for building self-driving cars. An Uber spokesman declined to say how much the company spent to acquire the startup.

Besides, Marcus, Geometric Intelligence co-founders are Zoubin Ghahramani, a machine learning expert at the University of Cambridge, Kenneth Stanley, a computer science professor at the University of Central Florida, and neurolinguistics specialist Douglas Bemis.

Uber has invested heavily in self-driving cars. The San Francisco-based company has built a massive research operation in Pittsburgh. It hired professors from Carnegie Mellon University, an early hub for autonomous vehicle and robotics research. This year, Uber bought autonomous trucking company Otto in a stock deal worth up to $680 million.

Machine learning is a field of AI that allows computers to analyze and adapt to new data without being explicitly programmed. Uber already relies on these techniques to improve the way prices respond to supply and demand, and to better match riders and drivers.

Calculations for UberPOOL, the company's carpooling product, can be particularly complicated since the company has to match multiple passengers in different locations with the best-positioned drivers, updating calculations as their cars move around.

To contact the author of this story: Eric Newcomer in San Francisco at enewcomer@bloomberg.net.

To contact the editor responsible for this story: Alistair Barr at abarr18@bloomberg.net.