Ever since the launch of the Pan-Canadian artificial intelligence (AI) Strategy in 2017 – the first ever national AI strategy – Canada has been the focus of sustained attention from the academic, IT and business communities. So, what are the key objectives of the Pan-Canadian AI Strategy? What progress has been made to date? What is the long-term potential for Canada to develop as a globally significant centre of expertise in AI? And how best can IT professionals contribute to this ongoing development?
National AI strategy
Founded in 2017, the pioneering CIFAR Pan-Canadian Artificial Intelligence Strategy (CIFAR) works in partnership with the country's three national AI Institutes - Amii in Alberta, Mila in Montréal, and the Vector Institute in Toronto - to achieve four key objectives. Firstly, to increase the number of ‘outstanding’ artificial intelligence researchers and skilled graduates in Canada. Secondly, to establish interconnected ‘nodes of scientific excellence’ in the country’s three major centres for artificial intelligence in Edmonton, Montréal and Toronto. Thirdly, to develop ‘global thought leadership’ on the economic, ethical, policy and legal implications of advances in AI; and, fourthly, to support a national research community on artificial intelligence. As Elissa Strome, AVP Research and Executive Director - Pan-Canadian AI Strategy, explains, since its establishment three years ago, CIFAR has ‘already made great progress towards its objectives.’
“The AI Institutes are all thriving hubs of their respective AI ecosystems, and there is a great deal of collaboration, exchange and training happening across all three Institutes,” she says.
Strome also points to a number of key highlights over the past three years, including the fact that, in March 2018, Canadian researchers Yoshua Bengio (Scientific Director and Canada CIFAR AI Chair, Mila; Université de Montréal) and Geoff Hinton (Scientific Director, Vector Institute; Google and University of Toronto), along with their colleague, Yann LeCun, won the ACM A.M. Turing Award ‘for conceptual and engineering breakthroughs that have made deep neural networks a critical component of computing.’ In December 2019, CIFAR and the AI Institutes also announced another round of Canada CIFAR AI Chairs, bringing the total to 80.
Another ongoing strategic priority for CIFAR is its AI4Good National Training Program, which is focused on advancing AI training and skills for women and other underrepresented groups. The annual CIFAR Deep Learning and Reinforcement Learning Summer School (which ran virtually this year) also attracts who Strome describes as ‘the top grad students from AI from around the world, with a focus on diversity and inclusion, to learn the state-of-the art in research advances from top scientists.’
“In summer 2020, we also launched a report from the AI for Health Task Force, which outlines specific actions that Canadian governments must take to build the foundations for a national strategy on AI for health to capitalize on Canada's advantages in this space. AI for Health is a tremendous opportunity for Canada,” she says.
Cutting-edge research
In addition to the CIFAR-wide work in pursuit of core objectives, the three key member organisations of CIFAR are also engaged in a wide-ranging variety of AI projects in their own right. For example, the Alberta Machine Intelligence Institute (Amii), an Edmonton, Alberta-based non-profit institute - originally founded in 2002 as a research centre at the University of Alberta – continues to support world-leading research in AI and machine learning and translate scientific advancement into industry adoption. As Spencer Murray, Director of Communications & Public Relations at Amii, explains, a key objective of the Institute is to grow AI capacity through advancing leading-edge research, delivering what he describes as ‘exceptional educational offerings’ and providing business advice – all with the goal of building in-house AI capabilities.
“We advance leading-edge research in AI by funding academic research and empowering industry leaders to invest in Alberta’s world-leading talent and expertise,” he says.
Other key objectives are to grow business capabilities and capacity in AI for start-ups, SMEs and enterprise clients to support growth, improve operations, and solve complex problems – and to build Alberta’s AI workforce by funding research positions to teach the next generation of professionals, and by developing and delivering educational training for technical teams, managers, and executives.
Key ongoing projects at Amii include an initiative to develop an AI companion to counteract loneliness in the elderly and research efforts aimed at addressing a range of impacts of the COVID-19 pandemic. The first cohort of Amii-funded Machine Learning Technicians have also recently presented their ‘capstone’ projects – with recruitment for the next cohort currently underway – and the Seventh Competition on Legal Information Extraction and Entailment, organised by Amii researchers, is scheduled to launch in October 2020.
“With diverse expertise in AI and machine learning, our researchers lead the world across many subfields and application spaces, including reinforcement learning, precision health, natural language processing, heuristic search and game theory,” says Murray.
“We strive to be the leading AI research institute in the world by 2025,” he adds.
Future prospects
Commenting on the broader prospects for the continued growth of the AI sector across the country, Murray observes that Canada already has ‘significant research expertise’ in AI and, as organisations like Amii continue to work to ‘connect industry into the leading-edge research and talent pools available in the country,’ he predicts ‘greater levels of industry adoption of AI.’
“Already Canada is a leader in the science of AI – and we’re continuing to develop our capabilities in the business of AI. One of the things we stress at Amii for industry adoption of AI is the importance of marrying domain knowledge with technical expertise. That’s why we offer courses such as our Machine Learning Technician Certification to individuals and as part of a greater corporate engagement,” he says.
“We’ve seen that the biggest challenges industry is currently facing in AI adoption around the world is access to talent and expertise – meaning the individuals needed to do the work of building useful models but also the know-how to apply AI within a business context. The ability to translate a business problem into an actionable plan for AI – and then translate back the work of AI technicians to leadership will be crucial going forward,” he adds.
Meanwhile Strome points out that Canada is already ‘doing really well’ at attracting and training technical talent in AI and, moving forward, believes the country has a ‘great opportunity’ to continue to grow its expertise in the societal implications of AI.
“We need more social scientists and humanists on the social, ethical, economic and legal aspects of AI and more collaborations between scientists, engineers and social scientists. Biologists and physicians who understand and can apply machine learning will also be critical to our growth,” she adds.