Research

Recommended Reading: Computers Already Learn From Us. But Can They Teach Themselves?


The world has achieved significant successes by applying supervised learning algorithms to business problems. With these systems often requiring large investments to incorporate human knowledge, researchers and practitioners are increasingly turning their attention to self-learning algorithms.

Richard S. Sutton, Chief Scientific Advisor at Amii, recently took the opportunity to sit down with Craig S. Smith of the New York Times as part of their Artificial Intelligence special report. In the interview, Sutton highlights the key value of reinforcement learning in enabling the creation of AI systems that learn and act autonomously. From the article:

“Reinforcement learning in computer science, pioneered by Richard Sutton, now at the University of Alberta in Canada, is modeled after reward-driven learning in the brain: Think of a rat learning to push a lever to receive a pellet of food. The strategy has been developed to teach computer systems to take actions.

Set a goal, and a reinforcement learning system will work toward that goal through trial and error until it is consistently receiving a reward. Humans do this all the time. ‘Reinforcement is an obvious idea if you study psychology,’ Dr. Sutton said.

A more inclusive term for the future of A.I., he said, is ‘predictive learning,’ meaning systems that not only recognize patterns but also predict outcomes and choose a course of action. ‘Everybody agrees we need predictive learning, but we disagree about how to get there,’ Dr. Sutton said. ‘Some people think we get there with extensions of supervised learning ideas; others think we get there with extensions of reinforcement learning ideas.’”

Cam Linke, CEO of Amii and an AI researcher in his own right, echoes Sutton’s sentiments on the increasing importance of reinforcement learning.

“Reinforcement learning is the new wave of AI,” says Linke, whose own reinforcement learning research focuses on AI adapting behaviours to improve self-learning. “More and more, we’re seeing companies exploring applications of reinforcement learning for process control and autonomous decision making. We’ve only just begun to scratch the surface of the value that can be created, and business leaders are starting to take notice.”

Learn more about reinforcement learning in the Reinforcement Learning Specialization offered by the University of Alberta and Amii on Coursera, developed and delivered Martha White and Adam White, both former students of Sutton’s as well as Fellows and Canada CIFAR AI Chairs with Amii.