With only 6 months until RE•WORK returns to Canada for the Deep Learning Summit on October 24 - 25 (2019) in the ‘Silicon Valley’ of its namesake, we wanted to take this opportunity to share some exciting developments in the agenda as well as celebrate our first-hand experience of the sheer scope of cutting edge research and technology in Canada from the East to West Coast; a spotlight on Edmonton & Montreal.
It is clear that Canada understands when it comes to AI development and deployment, there is a crucial responsibility to balance innovation with regulation, societal needs with business needs and the importance of encouraging cross-collaboration. Contained within this more general mission, both the development of Deep Learning and the promotion of Responsible AI specifically are championed by the Government of Canada, who invested $125 million in the Pan Canadian AI Strategy (lead by CIFAR) in 2017.
The following year, the Montreal Declaration for Responsible AI was announced, an initiative led by the Université de Montréal designed to guide digital transformation to ensure that everyone can benefit from the technological revolution. In wake of this, it seems only natural that the premiere of our Responsible AI Summit is to be co-located with the Deep Learning Summit in Montreal later this year. In a time where many multinationals & SMEs are being called to review their data handling and privacy processes, it is of paramount importance to review ethical considerations in the development and applications of AI. The Responsible AI Summit will echo the main themes of the declaration - bias, fairness, diversity, accountability - and how enterprises should ensure they are using AI to benefit all, across the business environment and wider society. You can view the schedule for the summit here.
As Montreal becomes the epicenter for Responsible AI - particularly notable as it is the home to the Montreal AI Ethics Institute - we mustn't forget that the Pan Canadian AI Strategy designated three pillars of research; the Vector Institute in Toronto, MILA (Montreal Institute for Learning Algorithms) in Montreal and Amii (Alberta Machine Intelligence Institute) in Edmonton. Recently, there has been increasing focus on Edmonton as a central player in AI, with csrankings.org designating the University of Alberta’s Computer Science Department as the third best in the world. This has been echoed commercially in Edmonton as well, with DeepMind, Huawei, and Borealis.AI opening up offices there. We were lucky enough to travel to Edmonton last week to discover what makes this city’s AI ecosystem so unique and investigate what has driven this rapid expansion.
Edmonton: The Home of Reinforcement Learning
Over the course of our exploration into AI in Edmonton, it became quickly apparent that their ground-up approach to progressing AI within the city has paid off. Reflected in the amount of new AI startups in the area and boom of applications to Richard Sutton’s reinforcement learning lab, Edmonton showcases vast amounts of entrepreneurial spirit and innovation effort. Students once (or still) affiliated with the Computer Science department at the University of Alberta (UoA) are being actively supported by the Canadian government and organizations including Alberta Innovates, Startup Edmonton and of course, Amii.
“The University of Alberta has been a leader in AI research for decades now. As the province looks to diversify its economy and grow industries in addition to, and complementary to, the oil and gas industry, AI has come to the fore as an area of strength we can draw upon. There is also increased interest and support coming from our municipal, provincial, and federal governments. This creates a perfect storm of a) growing demand from industry; b) world-class expertise to draw upon; and c) government support for development and application.” - Katelyn Petersen, Executive Director, German-Canadian Centre for Innovation and Research
As Katelyn suggests, AI growth within Edmonton’s industries is a recent phenomenon. Once upon a time, UoA represented the whole AI ecosystem in Edmonton, thrust into the spotlight by prominent AI researchers, Richard Sutton, Michael Bowling, and Patrick Pilarski. Another remarkable Canadian professor, Jonathan Schaeffer, author of the checkers-playing program Chinook, the first computer program to win a human world championship, shared his thoughts on why their research in reinforcement learning underpins AI in Edmonton;
“Toronto and Montreal are concentrating on Deep Learning. Edmonton has Rich Sutton, so reinforcement learning is our expertise. I believe that reinforcement learning for computers, just like humans, is a critically important component of intelligence. To try something, make a mistake, and then modify your behaviour to reduce/eliminate the chances of the mistake happening again is powerful and general. The goal of AI research is AGI -- Artificial General Intelligence. Surely and AGI computer will have reinforcement learning as part of its core” - Jonathan Schaeffer, Professor, University of Alberta
Also making significant waves in the AI field in Edmonton is Martha White, Assistant Professor of AI at UoA, CIFAR Chair, and Amii Fellow. As part of our Women in AI Podcast, Martha sat down with us to share her motivations and most recent work. Keep an eye out for the full conversation here, but in the meantime here is a sneak preview of what we discussed:
"One of the things I really love about reinforcement learning is that it is a very general formula for AI and gives a lot of room to incorporate many different types of techniques. It allows for a lot of innovation and allows you to think about many different ways that you could actually apply the algorithms that you are developing to a broad set of problems. I also think representation learning is one of the most fundamental questions in AI.” - Martha White, Assistant Professor, University of Alberta
Alongside this academic core of AI experts, it is undeniable the effects that Amii has had in expanding the AI community outwards into industry within Edmonton. During our visit, we sat down with Warren Johnston, Director of Amii Connects and Anna Koop, Director of Applied Machine Learning to discuss their core program areas, including Amii Explores, Connects, Educates & Innovates. As well as nurturing the next generation of AI experts in Edmonton, the team at Amii are leading the way in fostering partnerships between industry and academia; cohesively bringing together industry problems with research solutions in absolute harmony. When asking Warren about Amii's latest contributions, he noted;
“With new programs like Amii Innovates and Amii Educates, we’re helping to translate our scientific and academic leadership into industry, helping businesses build in-house AI capabilities and encouraging AI literacy in Alberta’s workforce. Add to this, Alberta’s young and highly-educated population, and we have the necessary elements for a powerful AI research and economic engine. Edmonton, especially, is the kind of place where risk-taking is encouraged and where the community will rally around great organizations and good ideas.“ - Warren Johnston, Director, Amii Connects
Jonathan Schaeffer also shared his two cents with us on his predictions for specific industries who have a growing interest in developing an AI capability in Edmonton, out with the traditional sectors. He believes that we should keep a close eye on finance and gaming, in particular. For example, through a partnership with UoA, ATB Financial are investing heavily in AI and applying this technology across their business.
Alongside larger corporations such as ATB, there is a burgeoning AI startup scene in Edmonton and we were keen to investigate why. We spoke to Dornoosh of Medo.ai, an AI-Augmented 3D Ultrasound platform, who emigrated from Singapore to Edmonton in 2015.
“A startup company does not mean anything without its people. Finding well-educated talents in Edmonton is relatively easier than anywhere else. This pleasant opportunity is further enriched by the active collaboration of UofA and the industry. At the UofA, students are encouraged to apply for internship positions at startup to apply their knowledge into solving real-world problems. The community has a very important role in the culture here in Edmonton and as an entrepreneur one can heavily count on the support he/she can receive from the community.” Dornoosh Zonoobi, Co-Founder, Medo.ai
Our time in Edmonton only further confirmed to us that the city has cemented itself as a center for AI research, and we are excited to see how corporations in Edmonton adapt to utilizing machine learning in the future. However, it was time for us to check in on the latest news in Montreal, also a popular destination with startups and tech giants - Facebook, Samsung, Denso, IBM, Ubisoft, Google, and Thales to name a few - alike.
Montreal: The Silicon Valley of Deep Learning
MILA is arguably the center of AI in Montreal and it is their mission to ‘create a unique space for innovation in artificial intelligence and technology transfer that will make use of interactions with industry and spark the emergence of start-ups while integrating the social impacts of technology in its projects’. Founded by recent joint Turing Award winner, Yoshua Bengio, MILA has furthered the ground-breaking contributions of Deep Learning and we are excited that Yoshua will be sharing some of his latest developments in this space with us in October at the Deep Learning Summit. Other recently confirmed speakers who will be joining Yoshua at the summit include Hugo Larochelle, Director of Google Brain Montreal, Doina Precup, Research Team Lead at DeepMind, Pierre-Marc Jodoin, Full Professor at Université de Sherbrooke and Ari Morcos, Research Scientist at FAIR, as well as many more leaders in the field still to be announced.
During our visit to Montreal we were very privileged to visit MILA and honored to meet Myriam Côté, Director AI for Humanity, who shared an insight into how she is spearheading a strategic plan for the fourth pillar of MILA; the development of socially responsible and AI (the others being Training, Academic Research & Technology Transfer).
Myriam will be joining us at the Responsible AI Summit to share her latest inroads into driving AI for Humanity at MILA and some exciting partnerships and projects they are undertaking to achieve this. We will also be joined by Ashley Casovan, Director for Data Architecture and Innovation within the Government of Canada, Richard Mallah, Director of AI Projects at the Future of Life Institute, Hilary Nicole Ervin, Senior Strategist, Ethical AI Trust & Safety at Google and both Abhishek Gupta and Mirka Snyder Caron from the Montreal AI Ethics Insitute at the summit.
Lastly, our final stop before leaving Montreal was the Institute for Data Valorization (IVADO), a multidisciplinary center of advanced expertise in data science, artificial intelligence, and operations research, to catch up with them and hear about their developments since we saw them last at the Deep Learning Summit in 2017. Commenting on their most exciting recent developments, Nancy Laramée, Partnership Director at IVADO mentioned;
“Currently we are focused on growth management, as we work on scaling academic research to meet the increasing needs of the industry and deliver collaborative research projects that transform scientific discoveries into concrete innovative applications, economic opportunities, and benefits to society. IVADO also greatly invests in the next generation of data scientists and has awarded close to $8M in research scholarships so far." - Nancy Laramée, Director of Partnerships, IVADO
Although we only scratched the surface of AI in Edmonton & Montreal on our trip, we are looking forward to returning to Montreal in October to further these conversations and celebrate the research and work of leaders in Deep Learning and Responsible AI. You can also keep an eye out for more in-depth interviews on our blog with the key players in the AI space in Edmonton!
Registering Your Pass for Montreal?
Super Early Bird Tickets to the Deep Learning Summit and the Responsible AI Summit are available until June 7th. Don’t forget, one ticket gives you access to BOTH summits and a brand new third deep-dive track, which will explore topics including ‘Choosing Which Deep Learning Method is the Right Tool to Use’, ‘Questions Your Startup Should Have Answers For’, and ‘Strategies for Scaling AI Responsibly in Enterprise’. Find out more information about how to register here.