Ghahramani is a titan who might have to draw on his unparalleled expertise to rework Uber’s autonomous technology
Uber is appointing Zoubin Ghahramani as its new Chief Scientist, company CPO Jeff Holden announced on their official blog late Tuesday. Ghahramani came to Uber via the recent acquisition of Pittsburgh-based Geometric Intelligence, of which he was a co-founder, but this role ostensibly gives him even more power in the Silicon Valley titan.
“I have realized what a fantastic place Uber is for machine learning and AI researchers,” Ghahramani wrote in the post. “There are a huge number of opportunities for both near-term high-impact research and longer term challenges to work on; the resources both in terms of data and computation are plentiful; and there are many talented and brilliant colleagues to work with.”
No one can doubt Ghahramani’s influence on the world of machine learning and artificial intelligence. He is a founding member of the Gatsby Computational Neuroscience Unit, founding director of the Alan Turing Institute, and a critical member of Carnegie Mellon’s headline-grabbing Machine Learning Department.
Wired reported in early December he was planning to split his time between the University of Cambridge and work hours for the startup colossus. The blog post here does not indicate a change in that plan. The Uber acquisition promised to keep all 15 staff members of Geometric Intelligence as the core of the company’s new AI Labs division. In short, Ghahramani was already extremely influential in the core department of Uber’s technology development.
His appointment is the first real good piece of public relations Uber has gotten since a deluge of scandals hit the company beginning in late February. In quick succession, Uber’s HR department was accused of covering up sexual harassment, the company was sued by Google subsidiary Waymo for allegedly stealing LiDAR designs, and Senior VP of Engineering Amit Singhal was forced to resign over undisclosed sexual harassment allegations back when he was with Google.
Ghahramani steps into an executive role as a definitely dignified replacement to Singhal, but Ghahramani was going to have influence on the company anyway. It also reassures investors on Uber’s own AI credentials following the risk of losing valuable IP in the lawsuit with Waymo.
Then again, Ghahramani faces a serious situation right now that perhaps only his own expertise can answer. Should Waymo’s injunction against Uber’s deployment of self-driving systems be honored by the court (and furthermore should Waymo ultimately defeat Uber and its subsidiary Otto), Uber will face an existential crisis. Uber will either have to pay out the nose both to license Waymo’s tech and pay settlement fees, or they will have to start virtually from scratch on an original autonomous concept (and pay settlement fees).
“We have to navigate around the real world, develop perception and action systems for our self-driving cars, and understand, predict, and make more efficient the experience for our riders and drivers, Ghahramani said in the announcement. “At a larger scale, we are trying to model and optimize entire cities, and reimagine the future of transportation through, for example, urban VTOL aviation.”
“The probabilistic ML approaches I work on are clearly useful for this, but we have assembled research talent across a much wider range of ML and AI approaches including deep learning, reinforcement learning, and optimization, as well as problem domains such as language and robotics. We are continuing to recruit across all these areas and more, for both talented researchers and engineers.”