Authors

Aaron Brindle, Partner, Radical Ventures

Leah Morris, Associate, Radical Ventures

Daniel Mulet, Investor, Radical Ventures

 

Overview: Canada’s AI Flywheel

Predicting what comes next in the rapidly evolving field of artificial intelligence (AI) is simple: follow the talent. From quantum neural networks and AutoML, to AI for drug discovery and natural language processing (NLP), the path to future innovations and commercialization is charted by a small subset of the global technology talent pool, the elite AI researchers who call Canada home.[1]

Canada’s AI pedigree is no accident. Long before their work would define a new era of innovation, the Canadian Institute for Advanced Research (CIFAR) and the Natural Sciences and Engineering Research Council of Canada (NSERC) were funding the work of AI pioneers Geoffrey Hinton, Yoshua Bengio, and Richard Sutton. And, as their breakthroughs made the leap from the lab into real-world applications, vibrant ecosystems of innovation coalesced around Canada’s AI luminaries. Focused government investment, paired with commitments from the private sector, anchored this community in Canada, establishing centres of excellence with the shared mission of attracting world-leading research talent. Today, the Vector Institute for Artificial Intelligence (Vector Institute), the Quebec Artificial Intelligence Institute (Mila) and the Alberta Machine Intelligence Institute (Amii) are major hubs in the hyper-competitive global AI talent landscape.

Like a flywheel, the initial government investment in Canadian AI talent attracted commitments from private industry to build corporate R&D facilities, and spurred direct investments from venture capital to commercialize research in AI-focused startups. The benefits of these investments compound over time, rapidly generating top-tier talent and yielding greater returns for less spend. Once set into motion, the flywheel effect creates sustained growth and momentum in AI-related advancements. Long-term talent retention requires high-paying private sector opportunities and a support network to commercialize research and nurture a new generation of entrepreneurs. Government policies are the lubricant within this flywheel, facilitating the speedy immigration of top international talent, unlocking valuable public data and incentivizing private sector investments in AI technologies and startups.

For investors and enterprises looking to accelerate Canada’s AI flywheel, it is essential to understand the three primary drivers of the ecosystem. This report explores the latest data and developments in talent, industry and policy with exclusive insights from the leadership teams at Canada’s AI research institutes.

 

Canadian AI Institutes Overview

  Alberta Machine Intelligence Institute (Amii) Quebec Artificial Intelligence Institute (Mila) Vector Institute for Artificial Intelligence
(Vector)
Location Edmonton, AB Montreal, QC Toronto, ON
President & CEO Cam Linke Valérie Pisano Garth Gibson
Chief Scientific Advisor Richard S. Sutton Yoshua Bengio Geoffrey Hinton
Core Academic Members[3] 26 27 35
Supporters

Government of Canada

CIFAR

Province of Alberta

Government of Canada

CIFAR

Province of Quebec

Government of Canada

CIFAR

Province of Ontario

Partners[4]

University of Alberta

DeepMind

75+ industry partnerships

McGill University

Université de Montréal

80+ industrial partnerships

University of Toronto

University of Waterloo

28 enterprise & 18 scaleup sponsors

Research Areas

Reinforcement learning

Heuristic search

Precision health

Natural language processing

Games & game theory

Deep learning

Robotics

Privacy, bias & explainability

Social network analysis

Data mining & information extraction

Deep learning

Reinforcement  learning

Computer vision

GANs

Medical imaging

Machine translation

Natural language processing

Object recognition

Societal implications of AI

Machine learning

Deep learning

AI for health

Probabilistic models

Statistical theory

Quantum computing

Sequential decision making

Generative models

Security, privacy & fairness

Industry Relations Contact Christy Holtby, Managing Director, Investment and Partnerships
partnerships@amii.ca
Stéphane Létourneau, Executive VP, Partnership and Corporate Affairs
partenariats@mila.quebec
Cameron Schuler, Chief Commercialization Officer and VP, Industry Innovation
industry@vectorinstitute.ai

 

Talent

Canada excels at producing and attracting machine learning and AI researchers, as well as data engineers and scientists.[5] When talented individuals choose to build their career in Canada, they are choosing to be part of a vibrant ecosystem of startups, enterprises, research institutions and investors.

Attracting Talent
Canada’s top-tier AI talent includes 109 world leading researchers who oversee labs at universities across the country and, in many cases, are Canada CIFAR AI (CCAI) Chairs.[6] They include 2018 ACM A.M. Turing Award winners Yoshua Bengio and Geoffrey Hinton, as well as influential machine learning researchers Richard Sutton, Jimmy Ba, Aaron Courville, and Joelle Pineau. Over the past two years, Canada’s AI research leaders collectively published over 4,000 research papers, and trained over 2,400 master’s and doctoral students.[7]

Amii, Mila, and the Vector Institute benefit from a pipeline of talent emerging from Canadian universities that have machine learning programs ranked among the top  programs globally such as the University of Toronto, the University of Waterloo, Simon Fraser University, the University of British Columbia, the University of Alberta, the Université de Montréal and McGill University.[8] Together, the institutes supervise 1,200 AI trainees. Where like attracts like, attracting the world’s best machine learning researchers is a catalyst to grow a talented and engaged community, which draws private sector investment (through traditional capital or data resources) and increases the likelihood of technological breakthroughs and commercialization opportunities.

Canadian organizations and wealthy individuals have also prioritized talent by creating attractive workspaces across Canada such as the Schwartz Reisman Innovation Centre – a 750,000 square foot complex designed to anchor the University of Toronto’s unique cluster of world-leading AI scientists and biomedical experts. This was made possible through a $100 million gift to the University of Toronto from Gerald Schwartz and Heather Reisman.[9] The Temerty Centre for AI Research and Education in Medicine has similarly been made possible through a $250 million gift from the Temerty family to advance machine learning tools for clinical applications and to further opportunities for entrepreneurship. These centralized locations bring research efficiencies, offering important access to compute power and infrastructure to enable world-class discoveries.

Researchers suggest the key to attracting and retaining top-tier talent is the availability of long-term, reliable funding and opportunities to solve interesting problems. Federal funding groups such as NSERC provided more than $17 million in funding to graduate-level AI projects.[10] Industry partners also continue to invest. These efforts have resulted in 2,054 Canadian-based AI papers being published in 2019.[11] As Geoffrey Hinton, whose research has been funded by government and industry, puts it: “In the long run, curiosity-driven research just works better. Real breakthroughs come from people focusing on what they’re excited about.”[12]

Retaining Talent: Canada as a Talent Platform

The Canadian AI talent retention rate has been lower than other leading nations such as the United States, the United Kingdom, and France.[13] Uneven AI adoption across sectors has been a push factor for talent as jobs are not consistently available in Canada.

As more industries apply AI solutions, however, career opportunities will expand across a variety of new application fields. An estimated 3,683 AI jobs were created and 16,205 AI jobs were retained in Ontario over 2020.[14] In a survey of 151 organizations operating in Ontario, 97% said they had created AI jobs.[15] The Canadian Occupational Projection System (COPS) estimates that demand for technology jobs will grow faster than other occupations with an estimated 100,000 relevant technology jobs being created by 2028.[16] This demand may create a labour shortage in AI-related jobs. To date, AI employment in Canada has been heavily weighted in financial services, healthcare, and general technology (see table below).

Sectors of employment for 2019-20, based on alumni of Vector-recognized AI master’s programs

Source: Vector Institute, 2020.

Canada has positioned itself as a generator of the limited global AI talent, renowned for developing and attracting a higher concentration of AI PhDs than most other nations (see Fig. 1, Appendix).[17] In the United States and Canada, the number of international doctoral students graduating in AI continues to grow, and currently exceeds 60% of the PhDs produced from these programs, up from less than 40% in 2010.[18] Canada continues to lead in the number of AI professionals who conduct research in the private sector.[19] This is, in part, due to a growing demand for talent that spans the need for both fundamental AI research and business applications.[20]

Employee education is an important consideration for industry looking to adopt and develop AI technologies. Canada still lags behind its peer nations when it comes to training industry employees to use AI in their jobs. Canada also has fewer developers and IT staff to create and deploy new AI solutions.[21] Organizations solely focusing on acquiring and educating a new generation of AI experts may overlook the importance of developing and retaining existing workforces.

Creating job opportunities that retain top-tier talent requires a private sector that understands the implicit long-term value of AI and commits resources to building out AI functions within their business. In Ontario, nearly half (44%) of the business executives in a survey of 151 organizations said their organization considers AI to have a strategically important role to play in achieving its business objectives.[22] Comparatively, 23% of companies globally in a Deloitte survey noted that AI was strategically important to their business’ success.[23]

Increasingly, the commercialization of Canadian research is generating demand for technology jobs. Canada’s AI job market has grown over time, now ranking 4th across 50 countries on the LinkedIn AI Hiring Index which measures the increase in AI job advertisements on LinkedIn globally.[24] Between 2015 and 2019, Canada (alongside Singapore, Brazil, Australia, and India) experienced the fastest growth in AI hiring.[25] Canadian AI research institutes, with the support of CIFAR, have developed a community of excellence that retains, attracts, and generates world-class AI researchers in Canada (Case Study 1, Appendix).

The health of Canada’s AI ecosystem can be measured, in part, by the number of well-paying AI jobs for highly qualified AI professionals who have graduated from AI-related programs. In contrast to Canada, talent in the United States remains relatively stable, with less outflow as a proportion of the country’s overall talent pool. This is attributed to many companies being headquartered in the United States where researchers are invited to work.[26] The trending pull for talent to relocate to the United States may be changing as more countries are able to retain their talent. A reduction of inflow for the United States may be accelerated by stay-at-home policies related to the pandemic that allow researchers to work remotely, as well as challenges in reversing a recent history of stricter immigration rules that increased the difficulty for students to stay in the country after graduation.[27]

COVID-19 and Talent

The global pandemic has negatively impacted the Canadian job market overall, with non-STEM (Science, Technology, Engineering and Math) occupations down 3.2% since February 2019.[28] There appears to be some resilience in technology-related jobs with STEM employment rising 1.7% in the same time frame.[29] One caveat is that STEM-related job listings are down nearly 50% from 2019, reflecting a still recovering economy and a general slowing in the pace of hiring in Canada.[30]

While Canada has more outflow of talent on average when compared with other AI-focused nations, Canada also has more inflow than its peer AI nations. Few countries have both higher than average inflow and outflow of talent. Aside from Canada, these countries (sometimes described as “platform countries”) include the Netherlands, Singapore, Switzerland, and the United Kingdom.[31] Canada is a central node in the global AI research network. 79.3% of Canadian AI articles from 2015 to 2019 were the result of an international collaboration (by comparison, over the same period 50.6% of articles from researchers in the United States were international collaborations). Multinational AI research labs based in Canada, play an important role in attracting and retaining a significant percentage of AI talent (see Fig. 2, Appendix). However, Canada’s research institutes remain an essential anchor point for top tier talent.

Canada’s Next Generation of AI Leaders

While the names of Hinton, Bengio and Sutton are synonymous with Canada’s pioneering AI legacy, Canada’s research institutes are actively building the next generation of AI leadership, successfully attracting and retaining some of the world’s brightest AI researchers.

Jimmy Ba
Faculty Member, Vector Institute; Assistant Professor, Department of Computer Science,
University of Toronto; Canada CIFAR AI Chair

Jimmy completed his undergraduate degree, Master’s degree and PhD at the University of Toronto under the supervision of Geoffrey Hinton, Brendan Frey and Ruslan Salakhutdinov. Among his many accomplishments, Jimmy developed the Adam Optimizer, one of the go-to algorithms to train deep learning models. Jimmy’s research focuses on the development of learning algorithms for deep neural networks. In 2015, his team achieved the highest place among academic labs in the image caption generation competition at CVPR. Following his graduation, he remained in Toronto and was nominated as a CIFAR AI Chair and became a Faculty Member at the Vector Institute.

Golnoosh Farnadi
Core Academic Member, Mila; Assistant Professor, Department of Decisions Sciences, HEC Montréal; Canada CIFAR AI Chair

Golnoosh, who was appointed a Canada AI CIFAR chair in January, 2021, has dedicated much of her academic career to developing novel machine learning and AI models to tackle fairness and ethics in AI. Her recent work addresses bias and algorithmic discrimination in decision-making models. Golnoosh completed her PhD in Computer Science at KU Leuven and Ghent University under the supervision of Martine de Cock and Marie-Francine Moens. Following the completion of her dissertation, in 2017, she became a postdoctoral researcher at the Statistical Relational Learning Group (LINQS) at the University of California, Santa Cruz. Prior to joining Mila as faculty, she was a postdoctoral IVADO fellow at Université de Montréal and Mila under the supervision of Simon Lacoste-Julien and Michel Gendreau. Golnoosh has significant collaborative experience, reflected in over 40 publications in international conferences and journals with collaborators such as Microsoft research, UCLA, University of Washington, and Tsinghua University. Among her numerous accomplishments, Golnoosh has received two paper awards for her work on statistical relational learning frameworks.

 

 

Industry

Private sector participation in the Canadian AI ecosystem has seen robust growth over the past five years. Corporate research labs and traditional enterprises are increasing their use of – and focus on – machine learning technologies. More than 45 companies have invested in AI research labs in Canada, including the Royal Bank of Canada, TD Bank, Microsoft, Nvidia and Alphabet (see Fig. 2, Appendix). These R&D labs are concentrated in Toronto, Montreal, Edmonton and Vancouver.

Essential to these industry investments is a collaboration with research scientists that brings material value to the private sector (see Case Study 2, Appendix). “The exchange works both ways, with companies learning about the latest AI research developments, and researchers having the opportunity to consider real-world data and applications of AI tools,” states Garth Gibson, Vector’s President and CEO. “We’ve found that close two-way interaction, sharing ideas and experiences, between researchers and industry practitioners has generated some of the best results for our industry sponsors.”

In Alberta, Amii has refined a framework called the AI Adoption Spectrum to guide their consulting projects with industry.

Amii AI Adoption Spectrum

Source: Amii, 2020

Industry Case Study: Amii

OKAKI | Amii

In 2019, OKAKI, which provides the technical support for Alberta’s province-wide prescription monitoring program, engaged the Amii team with their first AI project: to use machine learning (ML) to help prevent opioid overdoses.

OKAKI started out at the Exploring stage of Amii’s AI Adoption Spectrum. But, after two six-month engagements, OKAKI arrived at the Implementing stage. This rapid rate of technology adoption informed the company’s approach to selecting and iterating on new AI and machine learning projects.

OKAKI now has eleven employees in the organization working on ML projects. Additionally, the company has learned how, and where, to access talent for future hires. They have also acquired their first major client for a commercial  project, with four additional ML projects in the works. With Amii’s help, OKAKI has established a solid foundation in AI and ML, with the ability for technical teams and managers to communicate effectively about AI concepts and projects.

“Going through our first ML project together with support and mentorship from Amii gave us the confidence that we can do this, and do it successfully.”

— Dr. Salim Samanani, Founder and Medical Director of OKAKI.

In Ontario, the Vector Institute has over forty sponsors who gain access to their Industry Innovation programs designed to accelerate the application of advanced AI in their organizations. These programs bring together Vector’s world-renowned researchers, advanced compute environment, diverse talent community and AI engineering capabilities to support organizations working to transform AI into business value.

 

Industry Case Studies: Vector Institute

Manulife | Vector Institute

In early 2020, Manulife data scientists from across the globe attended the firm’s first NLP Academy. The NLP Academy is part of Manulife’s Advanced Analytics function. As a Vector sponsor, Manulife scientists engage in Vector-hosted industry AI projects, working alongside top AI researchers on business-relevant experiments including an NLP project. For a major insurance company with large amounts of unstructured text ― from customer call transcripts to benefit submissions and insurance applications ― automating the accurate reading, analysis and sorting of documents offered significant value.

“We aim to become a leading organization in applying advanced analytics in businesses. The NLP Academy is an important step in this journey. Vector’s contribution is very well reflected in the success of our internal academy.”

— Eugene Wen, Vice President of Group Advanced Analytics, Manulife

BMO | Vector Institute

BMO’s sponsorship of the Vector Institute offers opportunities to experiment with leading AI research through projects hosted by Vector’s researchers and Industry Innovation team. Stella Wu, an Applied Machine Learning Researcher at BMO Financial Group, proposed and developed a financial version of BERT – one of the most advanced language representation models available. Wu and Vector researchers used several online financial news sources to add over 182 million finance and market-related terms and contexts to the dataset. They then pre-trained the model with this enriched dataset, allowing BMO to deploy this model to analyze market sentiment.

Supporting this work were the face-to-face conversations, advanced lectures, and weekly feedback sessions that project participants received from Vector researchers and guest speakers. Wu had no previous experience with NLP prior to the project, but was able to access support materials including the latest information from researchers. Aside from producing successful AI projects, the relationship with Vector provides another benefit to BMO: retention. Working with Vector has helped BMO find and retain talent with the research skills and the ability to directly apply a deep understanding of the science to the business.

“Access to Vector researchers significantly accelerates our progress. Having the best researchers in Canada or Ontario available to us increases both the chances of success and the speed in which we can complete this.”

— Yevgeniy Vahlis, Head of Artificial Intelligence Capabilities at BMO Financial Group

Mila created a team of applied research scientists with the aim of helping industry bridge the gap between research advances and applications. Mila has undertaken various collaborations between Mila’s ecosystem of researchers  and industry partners, including the Gates Foundation for AI and drug repurposing, and IBM Canada for the Orion-IBM project to accelerate AI and machine learning adoption using open-source technology.

Industry Case Studies: Mila

Dialogue Technologies | Mila

A virtual health care provider, Dialogue provides a telemedicine platform that aims to facilitate diagnoses through a chatbot that collects relevant information about a patient’s symptoms. Mila is helping the company make the chatbot more effective by using reinforcement learning techniques. The goal is to collect relevant information through dialogue to provide a healthcare professional with the necessary information to identify pathologies.

Institut de recherche d’Hydro-Québec | IREQ | Mila

Mila’s applied research scientists work with researchers at IREQ to predict solar irradiance over a given territory, which is the amount of energy the sun releases when it reaches the ground. The goal is to obtain forecasts that are more accurate than typical weather models in an effort to improve deployment of solar energy and management of the electrical grid.

Natural Resources Canada (NRCan) | Mila

Mila is collaborating with the Geological Survey of Canada, a scientific agency of Natural Resources Canada, to build machine learning models that can predict the composition of minerals and other rock types in the substratum using reflective seismology data. The application of AI aims to address major scientific challenges posed by noise in the data, the limited amount of tagged data and the distribution of data that can vary significantly across geographic regions.

Canadian research institutes also offer opportunities to engage with the ecosystem through consortium projects such as Mila’s involvement in Le Consortium Québécois de Soins-Intelligents (the Quebec Intelligence Care Consortium), a collaboration with institutional partners, SMEs, and large corporations such as Hoffman La-Roche to enable Quebec patients access to their healthcare data, as well as the Consortium Project on Natural Language Processing at the Vector Institute to support the commercialization of NLP projects (see Case Study 3, Appendix).

Case Study: A Partnership between Hoffmann-La Roche Limited (Roche Canada), Amii, Mila, and the Vector Institute

The National Artificial Intelligence Centre of Excellence

The Roche AI Centre of Excellence (CoE), is “an unprecedented collaboration,” according to Valérie Pisano, President and CEO of Mila, combining the expertise of all three national AI institutes under the CIFAR Pan-Canadian AI Strategy. The CoE was announced in November of 2020 and focuses on advancing digital transformation in health. This partnership pairs Roche’s leadership in health and life sciences, and the three institutes’ established track record of excellence in AI research and enabling industry application.

The Roche AI CoE has its origin in the Roche Data Science Coalition (RDSC), formed in April 2020 as a multi-industry coalition lending expertise and resources to tackle challenges presented by COVID-19. As stated by Yoshua Bengio, Scientific Director at Mila, “recent efforts at Mila, Amii and the Vector Institute, in light of the COVID-19 pandemic, re-emphasize the collective commitment across our scientific community to collaborate and share information in order to drive AI advances forward with tangible societal benefits.”

The CoE will work to deliver quality AI-based digital solutions that optimize and reduce the cost of healthcare delivery, improve health outcomes, and enable Canada to learn and nimbly respond to opportunities and potential challenges in the healthcare system.

“Whether it’s used to design new pharmaceuticals, improve health systems, or personalize medicine, I am particularly passionate about realizing the transformative potential of deep learning in healthcare. It is exciting and heartening to see Roche partner with Amii, Mila and Vector and further affirm Canada as a strategic destination for AI research and adoption. As an industry sponsor of the Vector Institute, Roche is supporting the very programs and talent that will enable it to advance and accelerate AI research and application.”

— Geoffrey Hinton, Chief Scientific Advisor, Vector Institute; VP and Engineering Fellow, Google; Emeritus Professor, University of Toronto.

 

Venture Capital and the AI Startup Ecosystem

A hallmark of a thriving technology ecosystem, such as Silicon Valley, is a critical mass of entrepreneurs who have successfully built businesses and re-invest their experience and capital. As companies grow to become successful, they eventually experience a liquidity event allowing the founders to capture some of the value that they have created. Often, this capital is re-invested in a new generation of companies by way of angel and venture investments. These investments are much sought after sources of smart capital, offering judgment and knowledge alongside capital for building a business. This cycle repeats itself over and over again, with the critical mass of participants growing larger over time.

Toronto, Montreal, and Vancouver have seen robust growth in startup activity over the past 10 years. Overall, private investment in Canada has grown 307% from 2015 to 2019, a faster rate than the United States at 194%.[32] Hubs of entrepreneurial AI activity have emerged around organizations that cultivate entrepreneurship at the earliest stages of venture formation, such as NEXT AI, and the Creative Destruction Lab’s AI streams in Montreal and Toronto. These programs have an outsized impact in seeding and accelerating venture activity in the ecosystem.

Canadian startups, scaleups, and the three national AI institutes primarily collaborate in three ways:

  • Spinouts: The rarest and most valuable are the entrepreneurially minded researchers who spin-out companies based on their academic work. These top-tier scientists often attract global investment to their businesses, which creates the opportunity to grow a business in its local environment. For example, Alán Aspuru-Guzik, an advisor to Radical Ventures, CCAI Chair and Vector Faculty Member, has co-founded two companies that have raised a combined total of US$80 million.
  • Advisory support: Startups often turn to AI researchers for technical advisory roles which help them stay at the frontier of applied AI research. Many faculty at the research institutes are formal advisors to companies. Some scale their advisory role to commercialization programs such as Next AI and the Creative Destruction Lab. The Institutes are supportive of these non-core activities because it helps retain faculty who could earn more by engaging in private sector consulting or leading private research labs.
  • Employment: Local startups play an important role as employers of graduating students, offering career opportunities in Canada. In turn, the startups that engage actively with the AI community benefit from access to talent, top-tier research and global networks of expertise. Anecdotally, all of these play important roles in a startup’s ability to build differentiated technology and attract private sector investment.

Investment Activity

From a private-market investment perspective, there has been a decrease in the volume of investment into pure-play core AI technologies over the past five years. At the same time, the number of AI-powered companies has grown significantly, and the volume of capital raised has increased. This trend may be attributed, in part, to the wider adoption of AI technologies across businesses, which supports our thesis that all businesses are becoming AI businesses.

Fig1. Distribution of disclosed venture capital funding rounds across
industries in Canada, January 2015 to November 2020

 

 

Source:  Radical Ventures analysis of Tracxn data, Jan. 2015 to Nov. 2020 funding activity for AI startups in Canada.

Just ten years ago, there was virtually no venture capital investment in Canada’s AI ecosystem. Today, the tremendous potential of the sector is attracting substantial investments as Canada is home to over 250 active AI startups with publicly disclosed venture financing that raised capital between 2015 and 2020.[33] From 2015 to November 2020, there have been 246 seed rounds, 130 Series A rounds, and 12 rounds at late-stage financings. In 2020, Canadian AI startups raised a total of US$1.16 billion in venture capital, up 50% from 2018.[34] Despite the increase in total funds raised, the number of investment events dropped slightly, signalling a higher amount raised per funding round.[35] Looking at the makeup of investment each year, funding across each stage is up through 2018 with a large increase in late stage capital in 2019 for some mega deals.

Fig 2. Venture Capital Raised by AI startups in Canada, January 2015 to November 2020

 

 

Source:  Radical Ventures analysis of Tracxn data, 2015 to Nov. 2020 funding activity for AI startups in Canada.

Despite a relatively active investment landscape, an increase in venture capital at the earliest stages of company formation will drive entrepreneurial activity tied to Canada’s AI research institutes.

Given its talent bench strength, the Canadian AI ecosystem may be undervalued. Investment in Canadian AI is an order of magnitude smaller than the disclosed investment value in 2019 for the United States (US$25 billion) and China (US$5.4 billion), with total estimated values closer to US$47 billion and US$7 billion respectively.[36] It is worth noting that Chinese investors are estimated to have participated in less than 2% of investments into American AI companies.[37] Globally, private-market AI investment has grown tremendously in the past five years.[38]

 

Policy

Immigration

Ultimately, talented individuals and subsequent investments are more likely to be drawn to countries where governments are actively reducing the barriers to AI adoption. The Canadian policy landscape has enabled businesses to quickly deploy highly skilled AI talent. As part of the Global Skills Strategy (GSS), over 1,100 Canadian employers have used the Global Talent Stream fast-track program since its inception in 2017. This has attracted 24,000 workers to Canada. Policies that have supported Canada as a top-choice for talent include post-study work permits, eligibility for a worker’s family members to find employment, as well as paths to permanent working eligibility and citizenship.

Procurement

The public sector is building AI procurement practices to stimulate a competitive AI marketplace. Canada’s preliminary public procurement guidelines assess the risks for using AI while supporting the potential public benefit.

Scale AI supercluster initiative: Another model for government support

Based in Montreal, Scale AI invests in companies across a range of industries that are deploying AI to enhance their supply chain. The initiative seeks innovative, collaborative industry-led projects that are leveraging AI to help make supply chains more intelligent and more efficient, such as:

●       Supply chain operators

●       Supply chain solutions providers

●       AI and digital technology providers

Scale AI has supported Amii’s first accelerator stream, Supply Chain AI West. The stream is focused on helping founders to establish a strategic AI direction while also bringing together a community of experts to provide advice and access to their extended networks. These startups will be seeking strategic opportunities to partner with industry with the aim of driving rapid product development and deployment.

Learn more at Scale AI Projects

A macro-level commitment to develop responsible AI has spurred continued investment. The Canadian government has invested $750 million in AI since 2016.[39] According to CIFAR, that number expands to one billion dollars when including initiatives where AI is not the focus but is embedded – such as the Digital Supercluster in British Columbia and the Next Generation Manufacturing cluster in Ontario.[40]

Data

AI models learn by viewing millions of real life examples to make predictions based on variations in information. Generally, the quality and quantity of the data will impact the accuracy of the model. Once the model is trained, more data is needed to test the model’s accuracy when making predictions based on information it has never viewed before. In some cases, such as medical data, there are strict laws in place to ensure privacy is maintained such as the GDPR, HIPAA, and PHIPA. Companies with significant resources are able to form partnerships, generate their own data, and use strategies that are often not available to startups.

Even when large data quantities are available, data quality still remains an issue if algorithms are to function as intended. Data that is noisy, dirty, or incomplete will slow down the efforts of the most competent AI developers. Quality data requires meticulous data governance to ensure  developers have adequate access to data for statistically accurate analysis while protecting the privacy of the individual. Across Canada, efforts have been made to tackle this issue through dataset creation efforts.

Montreal Video Annotation Dataset (M-VAD)

Mila has supplied an annotated video dataset available through their website. This has been combined with the Max-Planck-Institute for Informatics Movie Description (MPII-MD) dataset in the Large Scale Movie Description Challenge. The challenge is sponsored by Nvidia and focuses on generating a more realistic and practical setting of multi-sentence movie description generation.

 

For more information see the Large Scale Movie Description Challenge

These efforts have been particularly salient in the healthcare space. Amii notes that Canada’s single payer health system may offer a data governance opportunity to leverage health data in a manner that respects the privacy of patients, preserves the security of the data and algorithms, while also advancing our ability to detect, trace, treat and cure. According to Cam Linke, CEO at Amii, “this is a key conversation that would help unlock significant potential for AI”. Amii is actively collaborating with organizations to collect, curate, and make research datasets available.

Ontario’s Ontario Health Data Platform (OHDP)

The Ontario Health Data Platform (OHDP) was created by the Ontario government as part of an effort to better detect, plan, and respond to the COVID-19 pandemic. The provincial government and the Vector Institute for Artificial Intelligence repurposed AI computing infrastructure, to support data-intensive analysis and decision-making. The result was a federated high-performance computing environment for secure and accurate linkages of large health data sets. These data sets ensure privacy while allowing for ML-powered analytics in an effort to support Ontario’s ongoing response to COVID-19 and its related impacts.

Responsible AI

Canada has taken a global leadership role in the development of AI and technology regulations. Since 2019, Canada has reviewed existing regulatory requirements and is currently conducting a reform of the Personal Information Protection and Electronic Documents Act (PIPEDA). The implications of this reform have yet to be fully determined. The Government published Canada’s Digital Charter in 2019, which outlined 10 principles to build a foundation of trust for Canadians in the digital sphere.

Canada is currently involved in leading a global online consultation on AI ethics spearheaded by Mila and Algora Lab (UdeM) in partnership with UNESCO. This builds on the 2017 Montreal Declaration for a Responsible Development of Artificial Intelligence which articulated ten fundamental values to underpin ethics principles for AI.[41] Canada is also co-leading the Global Partnership on AI (GPAI) with France. This is the first global initiative bringing together 14 countries and the European Union to guide the responsible use and development of AI.

Simultaneously, organizations such as the Schwartz Reisman Institute for Technology and Society, facilitate collaborative projects with a cross-section of researchers, industry stakeholders, government, and civil society to build AI and other advanced technology solutions that are high value and globally actionable, as well as safe, responsible, and inclusive. Another example is the International Observatory on the Social Impacts of AI and Digital Technologies (OBVIA) led by Lyse Langlois, Professor of Industrial Relations and Director of the Institut d’éthique appliquée de l’Université Laval. The Observatory will bring together nearly 20 universities and colleges, as well as nearly 90 research centres, including Mila to maximize the positive outcomes of AI and technology for communities, organizations, and individuals.

AI for Humanity at Mila

Socially responsible and beneficial development of AI is a fundamental component of Mila’s mission. Learn more at Mila.

Diversity and Inclusion with BiaslyAI

With the goal of eliminating gender and racial bias in written text, researchers Yasmeen Hitti, Carolyne Pelletier, Andrea Eunbee Jang and Inés Moreno are designing a tool to track gender and racial bias in AI data and “clean” it to be as neutral as possible. The idea is to train an algorithm with different data by asking different questions to avoid gender and racial discrimination in written text.

Detecting Trolls on Social Media

Canada CIFAR AI Chair, Reihaneh Rabbany, and her team at the McGill University School of Computer Science are using data mining and graph integration techniques to identify the misuse of social media to influence political activity. With a focus on analyzing real-world interconnected data, the tools in development aim to detect anomalies and suspicious behaviour on social networks during political events.

 

 

Pan-Canadian AI Strategy

Established in 2017, the Pan-Canadian AI Strategy is a $125 million national AI strategy led by the Canadian Institute for Advanced Research (CIFAR). The strategy aims to further establish Canada’s international leadership in AI. CIFAR plays a significant role in AI research in Canada and around the world, supporting the fundamental breakthroughs in the field through the Learning in Machines & Brains program. The Pan-Canadian AI Strategy provides funding to professors and research across institutes through CIFAR. For example, Irina Rish, a CCAI Chair and Mila professor was awarded a Canada Excellence Research Chair in the summer of 2020, along with a budget of $34 million over seven years.

The strategy’s objectives are as follows:

  • To increase the number of outstanding AI researchers and skilled graduates in Canada.
  • To establish interconnected nodes of scientific excellence in Canada’s three major centres for AI in Edmonton, Montréal and Toronto.
  • To develop global thought leadership on the economic, ethical, policy and legal implications of advances in AI, and;
  • To support a national research community on AI.

CIFAR’s leadership of the Pan-Canadian AI Strategy is funded by the Government of Canada, with support from Facebook and the RBC Foundation.

 

Headwinds

Despite Canada’s pioneering AI pedigree and growing talent base, strong competition is mounting globally. The 2020 updates to the Tortoise AI Index, ranking state capacity for AI, reveal worrying trends for Canada.[42] This year, Canada held fourth place on the global stage, but shifted down ranks significantly in multiple categories across implementation, innovation and investment.

Most notably, Canada’s operating environment fell from fifth place to thirty-second. This may be based on low rankings for privacy legislation implementation, gender diversity, and a lack of public trust. Possible indications of public misgivings were demonstrated through the loss of the Sidewalk Labs waterfront development project in Toronto last year.[43] Additionally, Canada slipped nine places in terms of development, a category requiring industry collaboration to support contributions to new models, techniques and products. The report also cites declining incentives to seek patents as a contributing factor to a slip in the rankings.

These issues will, in part, be addressed by industry accelerating AI adoption through commercial partnerships that provide access to data and the development of practical AI applications. AI adoption will be further aided through up-skilling the existing labour force to improve comfort levels with technology and to ensure workers are AI-ready. Such steps have the ancillary benefit of shoring up public awareness and understanding around the value of AI-based technologies.

 

How to Engage Canada’s AI Ecosystem

Endorse the Pan-Canadian AI Strategy

The United States has a significantly greater capacity to fund R&D efforts – spending almost US$1 billion in nondefense AI R&D in 2020. Similarly, China had an estimated nondefense AI R&D budget between US$1.7 and $5.7 billion for 2018. While the United States and China surpass Canadian funding on a per capita basis, Canada continues to compete on the basis of top quality talent producing impressive research results. In 2020, Canada moved from 7th to 6th place globally when ranked by total article share in the AI field from 2015 to 2019. Although funding is important, recent policy developments have made Canada a more attractive labour market for AI talent.  Endorsing the renewal of the federal Pan-Canadian AI Strategy is the quickest route to prioritizing investments in talent and building on Canada’s AI research legacy.

Advocate for Data Access

Access to data remains an issue for AI communities globally. While Canada is the top country in Open Data Barometer’s assessment of 30 countries on access to data, collaboration with industry will cement this lead through the provision of high quality data opportunities for AI research.[44] For instance, while Canada has a large opportunity based on its single-payer healthcare system, competing priorities and unforeseen complexities in executing a centralized database has, to date, alluded attempts to fully actualize this resource for research.

In building out data strategies, consideration must be given to ensuring secure and private AI mechanisms that protect both algorithms and datasets. Further consideration should also be given to safe and effective methods for human-AI collaboration, as well as improving the overall safety, privacy and security of AI systems. Organizations can set up advisory panels drawn from research institutes including academia, government and industry to help determine a set of relevant principles for the company’s AI work.

Invest in Canadian AI

Investing in research institutes offers a unique opportunity to contribute to Canadian AI research while benefiting from AI advancements tailored to business outcomes. Each research institute offers various types of involvement dependent on the needs of the partner. Engagements range from industry knowledge exchange to collaborative projects. These projects contribute to commercializing AI in sectors through quality data access and engaging job opportunities. The opportunity to engage with top-tier talent has attracted some of the largest Fortune 500 companies to Canada to expand their existing operations or to establish new R&D hubs (see Fig. 2, Appendix).

Collaboration and partnership opportunities also occur through engaging with startups and scaleups directly. Investing in venture capital funds offers a frontrow seat to industry trends, an increased capacity to source partnership opportunities for the application of AI technologies to products and services, and most importantly, fuels the growth of local technology companies poised to deliver outsized returns and global impact.

 

2021 Ecosystem Milestones

Amii

Amii has identified three specific goals for 2021:

1.     Foster AI and machine learning academic research by allocating funding to 26 primary researchers and by supporting academic talent and retention efforts of its postsecondary partners.

2.     Support 45 companies to build their AI and machine learning capabilities and capacity.

3.     Train and upskill 1,500 individuals including graduate students, technical team members, managers and executives and technicians through Amii education sessions in-person and online.

Mila

Mila will continue to focus on its four strategic pillars:

1.     Strive to attract, train and retain a growing and diverse pool of talent recognized for their extensive expertise in machine learning.

2.     Achieve the highest levels of scientific leadership in the development of innovative approaches to machine learning for AI.

3.     Collaborate with a range of organizations to accelerate economic and social innovation.

4.     Seek to stimulate a democratic dialogue on the potential of AI and the importance of its ethical and responsible development.

Vector

In 2021, the Vector Institute will execute year two of its Three-Year Strategy, which focuses on four pillars:

1.     Become a top 10 global centre for machine learning and deep learning research.

2.     Expand partnerships with Canadian industry through programs for talent, training, and applied AI projects.

3.     Towards better whole-life health, enable effective and appropriate research access to health data.

4.     Contribute thought leadership about Ontario and Canada’s role in AI, including economic and societal impacts.

 

 

Appendix

Collaboration Case Studies

Case Study 1:  Mila and Vector

Ecosystem Research Collaboration
Stony Brook Medicine | Mila, Vector InstituteTeams at Mila and the Vector Institute, led by Joseph Paul Cohen and Marzyeh Ghassemi respectively, collaborated to build a public dataset and AI models that predict the severity of COVID-19 pneumonia. The teams partnered with Drs. Tim Duong and Haifang Li at Stony Brook Medicine in New York who offered a test platform to trial algorithms and a large team to label and identify how algorithms in development can fit into clinical practice. The  project aims to build an effective predictive model for clinical settings.

 

Case Study 2:  Research-Private Sector Collaboration

Novartis Canada Biome
Novartis Pharmaceuticals | MilaAnnounced in October 2020, the Novartis Canada Biome – a collaboration between Novartis Pharmaceuticals and Mila – will bring together emerging tech companies and AI researchers to act as an on-ramp for the discovery and development of scalable digital solutions for patients and healthcare providers. Based in Montreal, the Biome is an extension of Novartis activities dedicated to AI and digital innovation for healthcare solutions. Startups that have joined the Biome network include ConversationHEALTH (the developer of a conversational AI for healthcare professionals, patients and consumers) and Amblyotech (a digital therapeutics company that was subsequently acquired by Novartis).

 

Case Study 3:  Collaboration with Research Institutes through Consortium Projects

Research-to-Application
Consortium Project on Natural Language Processing | Vector InstituteTraining NLP models is resource intensive, requiring days of processing on hardware that may be prohibitively expensive for many organizations. The Vector consortium project on NLP involved 37 participants from 16 companies that worked with Vector researchers in workstreams focused on various NLP-related experiments. Knowledge transfer between Vector researchers and industry participants occurred as industry sponsors recreated and trained an advanced NLP model on applications in health, law and finance for tasks such as machine translation, sentiment analysis, and question answering and subsequently applied these insights within their companies.

 

Figure 1:  Student Programming and Opportunities

Nurturing Workforce-ready Graduates
Vector Institute AI Master’s ProgramFrom eight founding faculty Members in 2017, Vector has expanded to a diverse group of over 500 researchers representing 15 universities across Canada, including 381 students and researchers at Vector. With support from the provincial government, Vector continues to expand Ontario’s workforce-ready AI talent pool by working with universities across the province to develop master’s programs that build core skills and complementary areas such as business and health. These programs respond directly to employers’ needs while attracting and retaining top students with competitive scholarships.Ontario universities have responded quickly to employers’ increased demands for AI talent. As of March 31, 2020, Vector recognized 22 AI master’s programs across Ontario for training graduates with the skills and competencies sought by industry. Of these 22 programs, four are new degree programs and 12 are programs which have updated curricula offering AI-specific minors, concentrations, and courses.These programs are contributing to an increase in AI candidates ready to enter the Ontario economy. As of March 31, 2020, there were over 1,000 AI master’s students enrolled in a combination of Vector-recognized programs and individual AI study paths at universities across Ontario.Mila’s academic network There are more than 500 students collaborating with Mila including master’s, doctorate, post doctorate, interns, visitors, professional master’s, and DESS (Diploma of Specialized Higher Education). In the fall of 2020 there were approximately 1,200 applications for all positions, with a 40% increase since fall 2018 in MSc and PhD positions including a 60% increase from applicants who identify as female in the same category. Mila has increased the number of offers made by 180% and awarded a total of 349 scholarships to students between April 2019 and March 2020. Mila now has a community of approximately 400 alumni across faculty members, students, and interns.Student funding at AmiiAmii funds student research and positions at institutions such as the University of Alberta through their fellowship program and the Canada CIFAR AI Chairs program. With a significant talent pipeline in Alberta, Amii supports the work of early-career AI and machine learning researchers. Students graduating from Amii have gone on to lead industry research labs and have taken top positions in academia. As of September 2020, Amii funded a total of 153 students including 80 master’s of science students, 56 doctoral students, 7 undergraduate students, 6 postdoctoral researchers, and 4 research associates.The CIFAR AI4Good National Training Program
CIFAR and its partners engage hundreds of Canadian and international students, from high school students to postdoctoral fellows, providing them the opportunity to develop the technical skills, expertise and networks that they need to be successful in their future careers. The program has a dedicated focus on engaging underrepresented groups in AI, and supporting programs that advance AI approaches that deliver positive societal benefits. A wide range of both Canadian and global organizations are involved in the program such as Amii, Mila, the Vector Institute, the University of British Columbia, IVADO, Global Goals AI, Simon Fraser University, and AI4ALL.

 

Figure 2: Corporate Research Labs (Exemplary Selection)

Lab Location Core Research Faculty / Key Researchers Partnerships
DeepMind  Alberta Edmonton, AB Information asymmetry (DeepStack) / Reinforcement learning Richard Sutton
Michael Bowling
Patrick Pilarski
Adam White
Amii
University of Alberta
DeepMind Montreal Montreal, QC Natural language processing Doina Precup
Shibl Mourad
Mila
McGill University
Ericsson
Montreal Global AI Accelerator (GAIA)
Montreal, QC Telecom & 5G applications Nimish Radia Mitacs
Facebook AI Research (FAIR) Montreal, QC Computer vision / Conversational AI / Integrity / Natural language processing /  Ranking & recommendations / Systems search / Speech & audio/ Theory / Human & machine intelligence Joelle Pineau
Mido AssranKoustuv Sinha
Michael Rabbat
Shagun Sodhani
Pascal Vincent
Amy Zhang
Mila
McGill University
Google Brain Toronto, ON Montreal, QC
Edmonton, AB
Multidisciplinary

David Fleet

Geoffrey Hinton

Hugo Larochelle

Dale Schuurmans

 

Mila
Vector Institute
University of Toronto
Johnson & Johnson – Janssen Research and Development, LLC Toronto, ON Speech & audio in disease detection – Alzheimer’s disease focus Frank Rudzicz
Katie Fraser
Liam Kaufman
Maria Yancheva
WinterLight Labs
LG Electronics AI Research Lab Toronto, ON Multidisciplinary  Dr. I.P. Park (President, CTO) University of Toronto
Nvidia Toronto AI Lab Toronto, ON Computer vision / Machine learning / Computer graphics Sanja Fidler (Director) University of Toronto
Vector Institute
Microsoft Research Lab Montreal, QC Deep learning Fernando Diaz Mila
Roche Canada – National Artificial Intelligence Centre of Excellence (Roche AI CoE) Toronto, ON AI and deep learning in digital health and life sciences   Amii
Mila
Vector Institute
Royal Bank of Canada
Borealis AI
Toronto, ON Montreal, QC
Vancouver, BC
Reinforcement learning / Machine learning / Security / Model validation / Fairness
– Financial services & banking

Foteini Agrafioti, Chief Science Officer

 

Amii
Mila
Vector Institute
University of Waterloo
Simon Fraser University
Samsung Advanced Institute of Technology (SAIT) Montreal, QC AI Processors Simon Lacoste-Julien Seoul National University
Mila
University of Montreal
TD Bank / Layer6 Toronto, ON Machine learning – Financial services & banking Tomi Poutanen Vector Institute
Thales Group
Centre of Research and Technology in Artificial Intelligence Expertise (cortAIx)
Montreal, QC

Multidisciplinary

 

Mila
IVADOVector Institute
Thomson Reuters – Toronto Technology Centre Toronto, ON Multidisciplinary Vector Institute

Uber

*closed

Toronto, ON Probabilistic programming /  Bayesian inference / Core deep learning / Reinforcement learning / Neuroevolution / Safety / Autonomous vehicles Raquel Urtasun University of Toronto
Vector Institute
 
Adobe Systems Inc. Toronto, ON Multidisciplinary
Etsy – Machine Learning Center of Excellence Toronto, ON Multidisciplinary
Huawei  –
The Noah’s Ark Lab
Montreal, QC
Toronto, ON
Computer vision / Natural language processing / Search & recommendation / Decision & reasoning / AI theory

 

End Notes

 

[1] One in ten of the papers selected for Oral Presentations at NeurIPS 2019 were authored by researchers based in Canada. Oral Presentations represent the most prestigious class of papers at NeurIPS, with an acceptance rate of 0.5% in 2019.

[2] While this primer focuses on the three established Canadian AI research institutes to date, Radical Ventures also recognizes British Columbia (BC) as a strong node in the Canadian AI Ecosystem. This includes CAIDA as the main AI research organization at the University of British Columbia (UBC) with over 100 professors and research associates. UBC contributed 17 percent of Canada’s AI publications in 2019. BC is the forefront region for computer vision innovation in Canada with Simon Fraser University ranking as the top university in computer vision in Canada on CSRankings (2010 to 2020) and 11th in US and Canada.

[3] “Core Academic Members,” Mila, Accessed December 18, 2020, mila.quebec/en/mila/team/; “Faculty Members,” Vector Institute, vectorinstitute.ai/#people; “Researchers,” Amii, 2020, amii.ca/about/our-people/

[4] “Industry Solutions,” Amii, Accessed December 18, 2020, amii.ca/industry-solutions/; “Industrial partnerships,” Mila, mila.quebec/en/mila/industrial-partnerships/; “‘Partners,” Vector Institute, vectorinstitute.ai/#partners.

[5] “Global AI Index Methodology Report,” Tortoise Media, 2019; “Global AI Index Methodology Report,”Tortoise Media, 2020.

[6] “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020, p. 4; “Canada CIFAR AI Chairs Program Surges Past 100,” January 19, 2021, “https://cifar.ca/cifarnews.

[7] “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020; “Canada’s AI Ecosystem: Government Investment Propels Private Sector Growth,” University of Toronto, 2020.

[8]  As per CSRankings, 2020; “Pan-Canadian AI Strategy Impact Assessment Report,” 2020, p. 5.

[9] All values in Canadian Dollars unless otherwise stated.

[10] “Artificial intelligence: 2019-2020”, accessed November 14, 2020, NSERC Award Database, nserc-crsng.gc.ca/. Note: this includes “expert systems”, “knowledge representation”, “learning and inference theories”, “logic programming”, “natural language and speech understanding”. This number increases to $22,277,813 when including Robotics (“computer vision”, “pattern analysis and machine intelligence”, and “intelligent machine systems”.

[11] “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020, p. 4.

[12] Kate Allen, “How a Toronto Professor’s Research Revolutionized Artificial Intelligence,” Toronto Star, April 17, 2015, thestar.com/news/world/.

[13] “2019 Canadian AI Ecosystem,” JF Gagne, 2019, jfgagne.ai/2019-canadian-ai-ecosystem/; “2020 Canadian AI Ecosystem,” JF Gagne, 2020, jfgagne.ai/2020-canadian-ai-ecosystem/.
Note: Measured by location of arXiv author affiliations.

[14] “Ontario AI Snapshot: The state of the province’s AI ecosystem in 2019/20,” Produced by Deloitte on behalf of the Vector Institute, 2020.

[15] “Ontario AI Snapshot: The state of the province’s AI ecosystem in 2019/20,” Produced by Deloitte on behalf of the Vector Institute, 2020.

[16] Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020; Ministry of Employment and Social Development of Canada. (2019) Canadian Occupational Projection System (COPS). Note: Technology occupations include NOC 0213 Computer and information systems managers, NOC 2161 Mathematicians, statisticians and actuaries, NOC 2147 Computer engineers (except software engineers and designers), NOC 2171 Information systems analysts and consultants, NOC 2172 Database analysts and data administrators, NOC 2173 Software engineers and designers, NOC 2174 Computer programmers and interactive media developers.

[17] “Global AI Talent Report,” JF Gagne, 2019, jfgagne.ai/talent-2019/.

[18] “The AI Index 2019 Annual Report, ” AI Index Steering Committee, Human-Centered AI Institute, Stanford University, Stanford, CA, December, 2019, hai.stanford.edu/.

[19] Global AI Talent Report 2020,” JF Gagne, 2020, fgagne.ai/talent-2020/.

[20] Ibid.

[21] “Future in the balance? How countries are pursuing an AI advantage,” Deloitte Center for Technology, Media & Communications, 2019, deloitte.com/content/. Note: countries include Canada, Australia, China, France, Germany, United Kingdom, United States

[22] “Ontario AI Snapshot: The state of the province’s AI ecosystem in 2019/20,” Produced by Deloitte on behalf of the Vector Institute, 2020.

[23]  Ibid.

[24] “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020.; Data Analyzed from LinkedIn, 2020.

[25] “Global AI Index Report,” Tortoise Media, 2019, tortoisemedia.com/intelligence/global-ai/.

[26] “Global AI Talent Report 2019,” JF Gagne, 2019, jfgagne.ai/talent-2019/; “Global AI Talent Report 2020,” JF Gagne, 2020, fgagne.ai/talent-2020/.

[27] “Global AI Talent Report 2020,” JF Gagne, 2020, fgagne.ai/talent-2020/. Note: Canada increased its score on invites to talent by 2.285 points from 2017 to 2019, demonstrating an ability to win back talent from the United States.

[28] “Magnetic North: How Canada Holds Its Own in the Global Race for Innovation Talent,”  Innovation Economy  Council, December,

2020, innovationeconomycouncil.com.

[29] Ibid.

[30] Ibid.

[31]  “Global AI Talent Report,” JF Gagne, 2019, jfgagne.ai/talent-2019/.

[32] “What investment trends reveal about the global AI landscape,” The Brookings Institution, September 29, 2020, brookings.edu/techstream/.

[33] Tracxn data as of 11/1/2020 of publicly disclosed funding rounds. Excludes public companies, those backed by private equity funds, and companies with undisclosed funding rounds.

[34] When comparing our analysis to data from the Canadian Venture Capital Association (CVCA) we note that AI investments in the first three quarters of 2020 (~US$743 million) represented ~23% of the total VC investment in the country over the same period (~US$3.5 billion), and was up slightly from 19% in 2019.

[35] There were 87 funding events in 2019, a drop from 98 in 2018.

[36] “What investment trends reveal about the global AI landscape,” The Brookings Institution, September 29, 2020, brookings.edu/techstream/.

[37] Ibid.

[38] “The AI Index 2019 Annual Report, ” AI Index Steering Committee, Human-Centered AI Institute, Stanford University, Stanford, CA, December, 2019, hai.stanford.edu/; “What investment trends reveal about the global AI landscape,” The Brookings Institution, September 29, 2020, brookings.edu/techstream/.

[39] “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020.

[40] Ibid.

[41] “Montreal Declaration for a Responsible Development of Artificial Intelligence,” Research Impact Canada & Université de Montréal, 2017, recherche.umontreal.ca/english.

[42] “Global AI Index Methodology Report,” Tortoise Media, 2019; “Global AI Index Methodology Report,” Tortoise Media, 2020, tortoisemedia.com/intelligence/global-ai/.

[43] “Sidewalk Labs announces it’s no longer pursuing Toronto waterfront development,” Global News, May 8, 2020, globalnews.ca.

[44]  “Pan-Canadian AI Strategy Impact Assessment Report,” Accenture & CIFAR, October, 2020; Open Data Barometer, 2020, opendatabarometer.org.