An ICTC Overview

Overview | Betting on Red and White

International Investment in Canadian AI

Access the full study, including all sources and methodology, here.

Study Scope

This study examines Canada’s opportunities for leveraging its current strengths in Artificial Intelligence (AI) to attract high-quality foreign direct investment (FDI).

Betting on Red and White: International Investment in Canadian AI also assesses recent domestic and international AI developments, and includes the following:

• AI applications across sectors
• Summary of Canadian AI research and commercialization
• Opportunities and barriers to continued AI expansion

This study extracts insights from industry leaders in over eight countries and seven sector verticals. It builds on ICTC’s 2019 report On the Edge of Tomorrow: Canada’s AI Augmented Workforce.

Terms

Artificial Intelligence (AI): a multi-disciplinary subject, involving methodologies and techniques from mathematics, engineering, natural science, and computer science and linguistics.

Machine Learning (ML): a subset of AI, the application of algorithms capable of automatically learning from past experiences without being explicitly programmed.

Neural Networks: computing systems made of numerous simple, interconnected processing elements that respond to external inputs.

Deep Learning: an AI field closely associated with artificial neural networks. Deep Learning refers to the depth of multiple layers or stacks of neural networks.

(Note: AI and Machine Learning are used interchangeably in this report. A further distinction is between “general” and “narrow” AI. General AI refers to the capacity for human-like cognition. To date, this has not been realized in machines. The current state of AI development is Narrow AI, commonly referred to as Machine Learning.)

Study Context

Canada is increasingly a global hub for AI foreign investment. Regional AI development centres in Canada are Toronto, Montreal, and Edmonton.

Global businesses are turning to AI to generate efficiencies, increase productivity, and solve problems. AI can improve internal processes, enhance customer experience, manage risks, and can even create new products or services.

Current examples of AI include Netflix movie recommendations, Spotify’s “made for you” playlists, Alexa’s vast search capabilities, Snapchat augmented reality filters, etc.

Promising new developments in AI have resulted in significant global growth in research, investment, and AI product development.

• Between 1998 and 2018, peer-reviewed AI research accelerated by 300%
• In 2019, global AI investment totalled $70 billion

Study Findings

AI in Canada

Canada is internationally recognized for academic research into AI. The rapid growth in AI research has recently spurred numerous AI startups across the country.

Currently, Canada has more than 650 AI startups, 40 accelerators and incubators, and over 60 research labs. Nearly 30% of Canadian startups were launched in 2017/18. International investment followed, with global giants such as Uber, Google, Facebook, and Samsung establishing AI research centres and operations in Canadian cities.

• In 2019, Canada ranked among the top five countries for innovative AI-based research
• Canada is becoming an international hub for AI startups, alongside US, Japan, and the UK
• Regionally, Toronto, Montreal, Edmonton dominate AI research and startups, with Vancouver, Waterloo, and Quebec City also attracting attention

Scale AI

Scale AI is an AI innovation ecosystem comprised of industry, researchers, and business for improving productivity across Canada’s economy through the integration of AI. It is part of the Canadian federal government’s supercluster initiative.

Scale AI also aims to advance Canada’s global standing in AI and attract high-quality investment from abroad. It received more than $250 million from the federal and Quebec governments.

To date, Scale AI has supported a total of 14 projects across industries such as retail, natural resources, and digital technology. Combined, these projects received over $32 million in funding to develop and scale their businesses.

Regional Hubs

Edmonton

The University of Alberta is a top-tier institution for computer science and AI. It is the home of Canada’s first Computing Science department (est. 1964), with about 20 faculty members working in AI-related research.

Edmonton-based AI trailblazer, the Alberta Machine Intelligence Institute (Amii) is a key fixture of Alberta’s thriving machine intelligence ecosystem and a key partner in many notable AI achievements over the past 15 years.

• The UofA recently partnered with Amii, with a focus on projects such as a digital chat companion for elderly Albertans, game development, clinical decision making, and financial portfolio balancing
• Google’s DeepMind, the Royal Bank of Canada, Mitsubishi Electric, IBM and Volkswagen have partnered with Amii to conduct research on AI applications and solutions across sectors
• Local startup-support groups also work with Amii to grow talent and accelerate local ecosystem growth
• Edmonton.AI works with Amii. Its mission is to create 100 AI and machine learning companies and projects in the city

Montreal

Montreal has the highest concentration of AI researchers in the world, with world-renowned research talent and more than 9,000 students in AI-related programs.

McGill University was ranked 35th top university in the world by QS World University Rankings in 2019.

The Montreal Institute for Learning Algorithms (Mila) is another key pillar of Montreal’s strength in AI research. Mila was founded by AI expert Yoshua Bengio, one of the world’s most cited computer scientists, famous for his work in neural networks and deep learning.

• In 2018, Mila’s Bengio (along with French-American Yann LeCun and British-Canadian Geoffrey Hinton) won the Turing Award — the” Nobel Prize” of computing
• Mila attracted substantial investment from government, industry, Google, and Microsoft
• Currently, the Mila community has more than 450 researchers, focussed on deep learning, bioinformatics, computer vision, and neural networks

Montreal has over 120 AI startups, including Canada’s most recognizable AI company, Element AI (founded by Yoshua Bengio).

Photo by Scott Webb on Unsplash

Toronto

Canada’s largest city is an epicenter for AI Startups, currently with about 250 AI startups. Among them:

• ecobee, a developer of residential smart thermostats
• AlayaCare, a cloud-based platform for healthcare practitioners
• Xanadu, a quantum photonic processor and open source software platform

Toronto’s AI presence is backed by the city’s financial services sector strength and world-renowned academics such as Geoffrey Hinton and Richard Zemel.

• Hinton is a University of Toronto professor, often referred to as the “Godfather of Deep Learning.” He was named one of the world’s top 100 influencers in 2016 and leads Brain Team Toronto for Google
• Zemel, also a UofT computer science professor, specialized in machine learning. He was selected as the Google/ NSERC Industrial Research Chair for Machine Learning in 2018. His AI distinctions include the NVIDIA Pioneers of AI Award
• Hinton and Zemel also figure in the Vector Institute, a Toronto AI research and development centre for advancing AI research and applications in deep learning and machine learning
• The Vector Institute secured $135 million in funding in 2017 (over five years)
• Vector Institute partnerships include Scotiabank, Accenture, Shopify, and St. Michael’s Hospital. The hospital partnership led to the creation of an early warning system for patients in need of transfer to intensive care units
• The Vector Institute’s research and commercialization in AI and talent attraction is closely followed by investors and industry partners

Canadian AI Response to Covid-19

COVID-19 derailed anticipated AI growth in 2020. However, the federal government’s spending to combat the pandemic includes 49 COVID research projects for manufacturing, life sciences, and AI (totaling $55 million across 100 projects).

Canadian AI innovators and researchers are pivoting to develop solutions for this global crisis.

• The Vector Institute recently produced a list of tools, including open source research and data sets for anyone interested in contributing to COVID-19 research.
• Toronto-based BlueDot developed AI-based “outbreak risk software” that searches news, reports, and tracks flight paths (and other networks) to help anticipate the spread and impact of the disease.
• A special taskforce was created across Canada’s three main geographic AI hubs to battle the spread of COVID. Mila, the Vector Institute, and Amii joined researchers from the Canadian Institute for Advanced Research (CIFAR) to work on AI projects related to COVID-19.

Image by Scott Webb on Unsplash

Study Interviewee Perspectives on Canadian AI

ICTC completed 20 in-depth interviews with key industry experts (including CEOs, CTOs, Directors, and technical leads) with international companies capable of making investments abroad. These interviews were critical to extracting primary research on AI use cases, investment needs, perceptions of Canada as a destination for investment, and international awareness of Canadian advances in AI.

Why Invest in AI?

The study interviewees predominantly look to AI to solve specific problems in three areas:

• Product enhancement: 50% of interviewees use AI to develop or enhance existing products. Another 25% look to AI to develop new products.
• Generating Efficiency and Improving Internal Processes: About 33% of interviewees are using AI to improve internal processes and generate efficiencies.
• Improving Customer Experience: 20% of interviewees use AI to improve the customer experience, primarily through chatbots, machine learning to analyze customer data, and other tools for building business relationships.

Company Investments into AI

Most interviewees in this study noted that their investments in AI are relatively new.

• Nearly 60% started their investment and use of AI began less than five years ago
• 25% have been using AI for five to nine years
• 15% have been investing in AI for over 10 years

Irrespective of length of experience with the technology, all interviewees have plans to continue growing their capabilities through AI-based investments, (despite that most were unable to quantify the economic value of AI to their organizations).

Barriers to AI Implementation

About 80% of interviewees rated barriers to implementing AI as “insignificant.”

• For those that cited barriers, insufficient talent topped the list
• “Cultural reluctance” to an “emerging or risky technology” was also noted

Awareness of Canadian AI Development

Many interviewees were familiar with the Canadian AI ecosystem.

• 95% of interviewees were aware of developments in Canadian AI
• 75% of interviewees were most familiar with Toronto, Edmonton, and Montreal as Canada’s main AI hubs
• Most interviewees knew about Canada’s Scale AI Supercluster and key educational institutions for AI researchers

Recognition of Canada’s Growing AI Community

Canada is perceived as a top destination for skilled international talent, with an effective immigration system, and a welcoming and friendly culture.

• Interviewees praised Canada’s educational institutions for producing high-quality AI talent at the entry level
• Canada’s growing AI capabilities — specifically in research — is internationally recognized
• Top researchers, such as Yoshua Bengio and Geoffrey Hinton, are essential to the recognition of Canada’s AI ecosystem

Barriers for AI Investment

Approximately 30% of interviewees couldn’t identify any barriers for AI investment attraction to Canada; the remainder noted the following investment barriers:

• Nearly 65% of interviewees identified “unclear” regulation as the top barrier
• Over 40% expressed concern about the Canadian talent pipeline (challenges in senior level AI talent recruitment). A greater worry is talent retention (losses to US competitors)
• 25% of interviewees noted concerns over scaling startups (confirmed by a recent report finding by the Impact Centre at the University of Toronto that Canada “dramatically underperforms” the US in scaling private companies)

A clear majority of interviewees felt that Canada was a favourable destination for foreign direct investment in AI and indicated interest in Canada for their own investment strategies.

Despite some hurdles, Canada is a top contender for AI-based FDI and is positioned to make a significant contribution to the global AI ecosystem.

AI Around the World

Globally, many countries are working to use AI to boost economic growth, generate efficiency, and create solutions to important societal challenges, such as the current pandemic. In 2018, there were an estimated 4,500 public AI companies around the world.

US Developments in AI

The US is currently the undisputed leader in AI, with nearly half of the world’s AI companies (over 2000). Many are supported by government agencies like the Department of Justice (DOJ), the Securities Exchange Commission, and NASA.

• The closest AI competitor to the US is China, with over 1,000 companies
• American AI companies typically generate nearly 50% more funding per investment than AI companies located in China

In 2019, Intel made a total of 19 investments in US AI startups, followed by Google at 16, and Microsoft with 11 investments.

The top three AI hubs in the US are:

San Francisco

• San Francisco has some of the biggest startup incubators in the world, with world-renowned institutions, including:
• Stanford University’s AI Lab, SRI International’s Artificial Intelligence Centre (AIC), and Google-NASA’s Quantum AI Lab)

Boston

• Boston is the second largest US AI hub and the premier location for biopharmaceutical developments. It is also home to institutions such as the Massachusetts Institute of Technology (MIT), MIT’s Computer Science and Artificial Intelligence Laboratory, and the Centre for Brains, Minds, and Machines.

New York

• New York City AI development is heavily supported by institutions like New York University’s Courant Institute of Mathematical Sciences, Facebook’s AI Research Group (FAIR), and others

Many US AI-based products have found consumer applications. Assistants like Google Home or Amazon Echo are among top-selling gifts. New US trade restrictions in 2020 (requiring US companies that export AI for geospatial analysis to apply for an export licence), however, could dampen sales of some US AI products worldwide.

US AI Response to COVID-19

The use of AI in the life sciences could play a key role in curbing the spread of COVID-19.

The US National Institutes of Health (NIH) announced funding for researchers and businesses developing solutions to COVID-19. Tech giants Apple and Google are codeveloping technology for contact tracing.

China

In 2018, a quarter of the world’s AI companies were in China.

A significant asset for China in the AI race is the data of its 1.3 billion citizens. Coupled with strong economic growth, China is expected to expand its global leadership in AI.

• China’s top AI companies include SenseTime, currently the world’s highest-valued AI startup. It received investment from Qualcomm, Fidelity International, and Hopu Capital
• Cloudwalk, a facial recognition technology giant in China, makes over 1 billion comparisons of faces against its database each day and has accumulated more than 100 billion data points

China expansion of surveillance capabilities through AI raises questions about data use and AI ethics, and influences China’s ability to source international investment. China currently receives one-third the investment that US AI companies receive ($15 billion in China vs. $45 billion in the US, in 2017).

Emerging Chinese AI Hubs

Of China’s 1,000-plus AI companies, approximately 40% are in Beijing and 15% are in Shenzhen.

Hangzhou City has gained recognition, with Alibaba, one of the world’s largest technology companies. Over 1,000 AI patents have been submitted by Hangzhou-based companies.

Other emerging AI hubs in China include Shanghai and Hefei. Hefei recently established China’s first national library for “brain-like” artificial intelligence technology.

Chinese AI Response to COVID-19

Wuhan was ground zero for the COVID-19 pandemic, and China focussed its AI efforts on combatting the health crisis.

• China accessed citizen data and large-scale acceptance of surveillance technology to develop new interventions for the spread of COVID-19
• AI-assisted temperature testing curbed the spread of the infection on public transit in China (passengers with high temperatures were contacted and advised to self-isolate)
• AI-assisted CT scans were piloted for faster detection of the virus in radiology departments

European Union AI

Europe has a thriving AI industry, with over 3,000 companies (public and private) across sectors, including data analytics, sales, marketing, healthcare, process automation, and image recognition.

EU AI hubs:

• Stockholm (with approximately 1,166 AI jobs per 1 million people)
• Amsterdam (730 AI jobs per 1 million people)
• And Berlin (677 AI jobs per million)

In 2018, Germany launched a digitization initiative aimed at becoming a global leader in AI.

Ethical AI Development: European Commission

The European Commission is seeking to ensuring that AI developments reflect core EU values. The EU is a global leader in the analysis of AI from ethical, legal, and socio-economic perspectives.

Under the EU’s Horizon 2020 framework, over €2.5 billion was allocated to AI-related research and development projects in robotics, big data, health, transportation, and emerging technologies.

This initial investment is expected to be followed by €100 billion of funding under Horizon Europe, the EU’s forthcoming research and innovation framework program set to launch in January 2021.

Key EU Startups

In March 2020, CB Insights ranked the top 100 most promising AI startups in the world. The list included six companies located in four EU countries.

France — Heuritech develops a deep learning powered automatic real-time recognition of objects, including shapes and people

Germany — KONUX smart sensors utilize advanced analytics to enable predictive maintenance for industrial products.

• NavVis develops fully managed digital twins

Spain — Sherpa is a dual platform digital assistant using machine learning

Sweden — Mapillary computer vision createS better maps.

• PerceptiLabs offers a unique way of building and visualizing models used by data scientists, machine learning engineers, and developers

EU AI Response to COVID-19

Many EU organizations focused on battling COVID-19. Hospitals shared data to help train algorithms through “federated learning” (data never left hospitals or touched a private server).

Other efforts included the ImaginingCovid19ai.eu project, which had hospitals transfer data to Quibim, a Spanish-based company using AI to improve the reading of chest CT scans when testing COVID-19 in patients with respiratory disorders.

Impact of AI Across Canadian Economic Sectors

AI has a growing presence across all areas of the economy. The following sectors serve as examples for the application of AI:

Agriculture Ocean Technology: A growing global population will require unique, effective, and climate-neutral approaches to improving agricultural yields. Weather date, soil quality, crop growth, and even animal health can then be improved by machine learning algorithms

• AI is increasingly key in detecting diseases in plants, aiding in pest control, and can identifying farm-wide problems (in combination with drones)
• Machine learning algorithms can draw upon historical data on rainfall, temperatures, and evaporation to predict the likelihood of a droughts

Advanced Manufacturing: AI, robotics, the Internet of Things, and 3D printing are enhancing manufacturing output

• Various human activities may not be suitable for automation but can still be improved through AI-generated insights. (Eg. California-based Drishti uses cameras equipped with deep leaning architecture to generate real-time analytics of human-performed actions analyzed by AI to optimize human performance)
• Business and Finance: Financial advisers, investment bankers, tax advisers, and human resources professionals leverage large amounts of data to make informed decisions. Automated and algorithm-based financial and business products are gaining ground

Digital Technology: AI in the digital technology sector is fast evolving and diverse

• Facial recognition technology automatically “logs in” iPhone owners
• Apple’s Overton AI-based apps answer billions of questions and analyze trillions of data records

Life Sciences: Data and analysis-intensive healthcare applications can benefit from AI for improved diagnosis, patient experience, and accelerate drug development

• Toronto-based ConversationHEALTH offers messaging apps, websites, voice devices, and banner ads, allowing healthcare companies to enter the “conversational age”
• Swift Medical uses AI to identify the severity of wounds and speed healing
• Telecom giant Telus developed an application called Babylon that uses a chat-style symptom checker powered by AI

Natural Resources: Further automation and increasing use of AI is expected in this sector. Digital technologies are improving environmental performance, costs efficiencies, and safety

• In the forestry industry, machine learning is improving the speed and accuracy of tree species analysis, estimating wood volume, and tree dimensions
• Vancouver-based PhotoSat uses machine learning algorithms to transform pictures taken from satellites into 3D models of worksites

Transportation and Logistics: AI is improving transportation, logistics, and supply chains

• Ocado, a supermarket chain in the United Kingdom, uses a robot called the “hive-grid-machine” to execute 65,000 orders per week, dramatically reducing labour costs while optimizing the movement of items to their final destination
• Attabotics, a Calgary-based robotics supply chain company, replaces warehouse shelving with vertical storage structures attended by robotic shuttles
• AI is a significant component of smart mobility options like autonomous vehicles, but many transportation businesses are already using AI to generate efficiencies and manage traffic flow