Zoubin Ghahramani is Uber’s Chief Scientist and head of Uber AI Labs, Uber’s research arm dedicated to artificial intelligence (AI) and machine learning. Below, Ghahramani introduces AI Labs’ newest team member, award-winning neuroscientist Peter Dayan.
We are thrilled to officially welcome Peter Dayan to Uber AI Labs! One of the world’s most respected researchers in machine learning and computational neuroscience, Peter joins us from London, where he was director of Gatsby Computational Neuroscience Unit. He is a pioneer in the field of reinforcement learning, where he helped develop the Q-learning algorithm, and made seminal contributions to unsupervised learning, including the wake-sleep algorithm for neural networks and the Helmholtz machine. Peter is also the recipient of the 2012 Rumelhart Prize for his contributions to the theoretical foundations of human cognition, and most recently the 2017 Brain Prize for his work on brain mechanisms that link learning to reward.
Peter’s expertise and interest in behavioral economics, decision making under uncertainty, probabilistic models, and reinforcement learning will help Uber solve the many machine learning and AI challenges we tackle across our business.
“I’ve been a huge fan of Peter and his remarkable work for some time, so I’m truly honored to have the opportunity to work side by side with him on some of the most challenging real-world machine intelligence problems of our day at Uber,” said Jeff Holden, Chief Product Officer and head of Advanced Programs at Uber.
Having worked with Peter for over seven years at the Gatsby Unit, I am a great admirer of his ability to both understand the big picture and dive into the important technical details of almost any problem he is presented with. It goes without saying that Uber is looking forward to the impact Peter’s scientific input will have on helping us create safer and more reliable transportation experiences.
“I’ve spent the last three decades working on understanding decision making—a field that links AI and economics, and control theory with psychology and neuroscience, to study what’s optimal, what’s practical and what actually happens,” said Dayan of his new role. “Uber abounds with fascinating questions along these lines in essentially all its areas of current and future business, and these questions have both research and practical implications that I’m keen to work with AI Labs to explore.”
The most pressing control theory problem Peter has found thus far at Uber? Elevator scheduling in our office. With Peter’s help, perhaps we can leverage our algorithms to find a solution.
On behalf of AI Labs and Uber, we welcome Peter to our team!
Zoubin Ghahramani is Uber’s Chief Scientist, a Professor of Information Engineering at the University of Cambridge, and Deputy Academic Director of the Leverhulme Centre for the Future of Intelligence.
Learn more about Uber’s machine learning and AI innovations:
- Uber AI Labs Open Sources Pyro, a Deep Probabilistic Programming Language
- Meet Michelangelo: Uber’s Machine Learning Platform
- Engineering Extreme Event Forecasting at Uber with Recurrent Neural Networks