DeepMind’s first international AI research office lands in Alberta

Alphabet’s DeepMind is expanding beyond its U.K. home, with a new international office based in Edmonton, Alberta, Canada. The research-focused office is being set up in close partnership with the University of Alberta, and will be led by UAlberta academics Rich Sutton, Michael Bowling and Patrick Pilarski, in the latest example of Canadian AI-focused scholars getting the chance to contribute to private-sector endeavors while continuing to work on their academic research and teaching, too.

Alongside Sutton, Bowling and Pilarski will be Adam White, returning to Alberta as an adjunct professor, as well as six other researchers who co-authored a recent scientific paper about “DeepStack,”¬†an AI system that made headlines by beating human pro poker players at no-limit Texas hold’em.

The arrangement with the University of Alberta includes continued funding for research from DeepMind, and Communications Director Aaron Brindle from Google Canada says the “aim is to attract even more world-class talent to the region,” with the intent of contributing “to the growth of Edmonton as a technology and research hub.”

All three of DeepMind Alberta’s lead researchers share a focus on reinforcement learning, which attempts to mimic the way humans learn to improve and repeat good outcomes and train avoidance of bad ones. Sutton also acted as DeepMind’s first-ever advisor prior to its acquisition by Google in 2014, and Alberta alumni employed at the company have helped develop its existing AlphaGo and Atari-playing AI systems.