Microsoft plans to significantly expand its Montreal research lab and has appointed Carnegie Mellon University machine learning professor Geoffrey Gordon as its new research director, the company announced Wednesday.
In a Jan. 23 blog post, the company also said that it plans to double the size of Microsoft Research Montreal within two years, ultimately employing up to 75 technical experts at the facility.
The move illustrates what Microsoft Research New England, New York City, and Montreal managing director Jennifer Chayes called the Canadian city’s status as “one of the most exciting places in [artificial intelligence] right now.”
“We want to be doing the research that will be infusing AI into Microsoft products today and tomorrow, and Geoff’s research really spans that,” Chayes said in the post. “He’ll be able to help us improve our products and he’ll also be laying the foundation for AI to do much more than is possible today.”
Though Gordon is an expert in reinforcement learning, in which systems learn through trial and error, he’s also done groundbreaking work in areas such as robotics and natural language processing, Chayes said, noting that his ability to combine all three areas of expertise will be key to developing sophisticated AI systems in the future.
For his part, Gordon told Microsoft writer Allison Linn that he was drawn to the position because of the Montreal team’s work in AI research and the opportunity to collaborate with the broader Montreal AI community.
“Research has always been about standing on the shoulders of giants, to borrow a phrase from a giant – and it’s even more so in the current age,” he said in the post.
Microsoft began developing its research presence in the city last January, when it acquired deep learning startup Maluuba.
Since then, the Montreal team’s research has contributed to such AI breakthroughs as improved reading comprehension and methods of teaching AI systems complex tasks by dividing large tasks into smaller ones that multiple AI agents can handle (which is how it trained its AI to win the Atari 2600 version of Ms. Pac-Man).
Gordon is especially interested in creating AI systems that possess what users would call long-term thinking: the ability to observe the elements of a problem and solve it by developing a coherent, multi-step plan, a feat that is currently rudimentary in most AI systems, which typically focus on individual tasks such as recognizing images or identifying words in a conversation.
“We have, in some cases, superhuman performance in recognizing patterns, and in very restricted domains we get superhuman performance in planning ahead,” he said in the post. “But it’s surprisingly difficult to put those two things together – to get an AI to learn a concept and then build a chain of reasoning based on that learned concept.”
In addition to Montreal, Microsoft runs international research hubs in Beijing; Bangalore, India; and Cambridge, U.K.