Uber has announced its plan to use artificial intelligence to make self-driving cars more efficient on less complex and shorter routes.

According to the company, a laboratory called "Uber AI Labs" will be established in San Francisco for its geometric intelligence works that will help self-driving cars use less data and take shorter and more convenient routes during travels.

Like Us on Facebook

CEO Gary Marcus of Geometric Intelligence said that "deep learning is very data hungry - we're learning it faster." He added that the system could help to cut half of the data needed. For instance, Geometric Intelligence was used in deep generator networks to create new image from a set of images by figuring out what each network neuron was doing. With this knowledge, the system can easily discover driving errors and correct the machine's defective habits. Marcus said that the current era is dominated by big data "but for some problems there's just not enough data."

The collaborative project between Uber and Marcus' team will include expert advise from Zoubin Ghahramani, a machine learning researcher, and Jeff Clune, an expert in evolutionary algorithms. This is the second AI-related purchase by the American online transportation network company. In August, Otto was purchased for $680 million with a goal of improving the truck's self-driving features through Uber's AI driving data.

This partnership between Uber and Geometric Intelligence will help the former to transform itself into a tech giant and pioneer in the AI transportation sector. The company has invested $70 billion on the system, which reveals its aggressive effort to become the biggest player in the ride-sharing platform.

AI expert Oren Etzioni said that the transportation company is reinventing itself to join the Big Four which includes Google, Amazon, Facebook, and Apple.