Late last year, DeepMind began more closely collaborating with fellow Alphabet-owned company Google by founding a Mountain View-based team. The AI research company is now opening its first international AI division in Canada.
In the process, it is hiring (via Bloomberg) three esteemed artificial intelligence researchers from the University of Alberta in Edmonton. Professors Rich Sutton and Michael Bowling both have previous connections with DeepMind.
The former is an expert on reinforcement learning — which has computers learn by trial and error — and previously served as DeepMind’s first scientific advisor. Leading the new lab, Sutton helped supervise another researcher who later led the AlphaGo team.
Meanwhile, Bowling also uses reinforcement leaning to train systems that can play poker better than professional players. He previously created a platform that allowed others to build systems that could play old Atari console games. In 2015, DeepMind used this work to create a rapidly learning AI that could play the games without having prior knowledge and still be better than humans.
The last hire is Patrick Pilarski, who has a particular focus on building artificial limbs that use AI. This particular line of research lines up with DeepMind’s increasing interest in applying AI to the field of health, though this has encountered some setbacks.
These three researchers will remain as part-time professors at the university to teach and supervise graduate students.
Canada has become a hotbed of artificial intelligence research in recent years. Just last year, Google Brain — an internal AI division — founded a Toronto office, while Google made an investment to fund more research in the country.