Ā鶹AV

News

Experts: Google AI against Go Champion

Published: 9 March 2016

ā€œAfter an extraordinarily close contest, Googleā€™s artificially intelligent Go-playing computer system has beaten Lee Sedol, one of the worldā€™s top players, in the first game of their historic five-game match at Seoulā€™s Four Seasons hotel. Known as AlphaGo, this Google creation not only proved it can compete with the gameā€™s best, but also showed off its remarkable ability to learn the game on its own.ā€ ()

The remaining games can be livestreamed on .

, Professor, School of Computer Science, Ā鶹AV

"It is really a historic day for machine learning, especially reinforcement learning. Older large successes, like DeepBlueĀ  or Jeopardy, involved significantly more engineering; AlphaGo is truly a learning entity."ā€”Doina Precup

Professor Precup is interested in artificial intelligence andĀ machine learning.

She's not available today between 11:00 AM - 12:00 PM and 2:30-3:30 PM. She's not available tomorrow between 10 AM - 11:30 AM.

dprecup [at] cs.mcgill.ca (English, French)

, Professor, School of Computer Science, Ā鶹AV

"The game of Go has been a milestone for AI researchers for several years. Many people thought we would not attain human-level performance for another decade. Ā It's exciting to see the pace of progress!Ā  Of course we'll be watching closely for the results of the other upcoming games."ā€”Joelle Pineau

Two of the main researchers (Arthur Guez, Marc Lanctot) on Google'sĀ DeepMind team are alumni of the School of Computer Science at Ā鶹AV.Ā  Mr. Guez didĀ his MSc under ProfessorĀ Pineau's supervision. She co-directsĀ theĀ .

She's available today between 1:30-2 PM and 3:30-4 PM today. Tomorrow, between 1-2:30 PM. Friday, between 10:30-11 AM.

jpineau [at] cs.mcgill.ca (English, French)

, Director, School of Computer Science, Ā鶹AV

ā€œThe game of ā€˜goā€™ is far more complex and challenging for a computer than chess, yet it was only a matter to time before a computer mastered it.Ā  For decades games were the standard of performance for emerging artificial intelligence. Those challenges are past and we are in an eraĀ  when robotics and intelligent machines will gradually meet and surpass any remaining reasonable intellectual challenge we can propose. This success is especially striking since it exemplified a new class of solution that is much closer to how humans plan this and other games.ā€ ā€”Gregory Dudek

Heā€™s the Director ofĀ . Heā€™s interested in robot navigation, mobile robotics, robot localization, information summarization, human-robot interaction, sensor-based robotics, multi-robot systems, computer vision, vision, recommender system, web services, recognition.

dudek [at] cim.mcgill.ca. 1-514-398-4325. Heā€™s not available Wednesday between 10-11:45 AM and Thursday, 11:00-11:30 AM. (English, French)

Back to top