The Brain Drain of Canadian AI Talent is now Local

As stated in part one and two, fearing that the brain drain of local AI talent by foreign tech giants might dash Canada’s ambition of creating a domestic AI industry, public authorities sprung into action to not only grow but entice the grey matter to stay and build a competitive AI ecosystem. Federal and provincial governments started providing generous public funding to grow the AI talent pool and finance its ability to further Canada’s lead in fundamental AI research undertaken at academic institutes. More importantly, greater investment in this emerging technology would help tech entrepreneurs found startups focused on applied research, spurring commercialized AI solutions.

As for the tech community, aware that applied AI research revolving around the B2C market would be very challenging given Big Tech’s dominance over the internet consumer industry, entrepreneurs would focus on creating enterprise AI solutions for the B2B (and B2G) market where Big Tech isn’t that dominant.

This public/private synergy led to a flurry of private investments in academic institutes and, more importantly, in startups. The subsequent abundance of fundamental and applied research opportunities slowed the brain drain and enabled Canada to achieve the third largest AI talent pool worldwide, capable of building a domestic AI industry able to compete with foreign tech incumbents for market shares.  

Nonetheless, while the brain drain slowed down, foreign tech incumbents still pose a significant threat to Canada’s ambition of creating AI champions. Indeed, having reached a critical mass of AI talents able to conduct fundamental and applied AI research, global tech players are starting to invest in the fledgling Canadian AI ecosystem to capture the latest innovations, hence making the brain drain local.

In fact, the country’s AI talents and their associated applied research in enterprise AI are increasingly exposed to corporate capture by Big Tech. As stated in part two, since the sale of specialized B2B solutions doesn’t yield as much profit as B2C products catering to a larger customer base, Big Tech tends to underserve businesses by providing them with streamlined and general enterprise software. However, the recent AI innovation breakthroughs in Machine Learning (ML) are transforming numerous B2B markets into gold mines. Since AI today has a real shot at revolutionizing organizations by unlocking massive productivity gains, one can predict a strong demand for enterprise AI solutions on the part of businesses. As a result, it would be foolish for Big Tech to let startups capture most of the enterprise AI software market, where some segments have as much potential in terms of revenue as the B2C domain. Consequently, the AI revolution provides the impetus for Big Tech to develop more tailored ML enterprise software, therefore putting them in direct competition with Canadian AI startups.

Tech incumbents are on the lookout for enterprise AI startups that might threaten Big Tech’s bid to conquer B2B market shares. For that matter, Big Tech players have an information advantage when it comes to identifying potentially threatening upstart firms early on. Indeed, they are exploiting AI startups’ dependence on Big Tech’s platforms for distribution, computing, and storage. Having limited financial resources in the early stages, tech founders can ill afford to market their solution to potential customers on their own. Moreover, it would be very costly for AI startups to possess their own computing hardware for processing and data storage. As such, they prefer to develop and host their AI solutions as third-party tools on the cloud services of tech giants such as Amazon Web Services (AWS), Microsoft Azure or Google Cloud. In so doing, tech entrepreneurs get the opportunity to market and sell their product to large numbers of businesses using cloud services. Equally important, they get access to off-premise computing infrastructure to train and operate the enterprise ML software. As a result, Big Tech cloud operators can identify which third-party software is most popular among their cloud users. Once an AI startup gaining market shares is identified, tech incumbents move in for the kill.

One of the most common strategies for Big Tech is to proceed with an M&A of the targeted startup. According to a 2018 report, while there’s been an average increase of 49% in startup deals in Canada between 2013 and 2018, the number of M&As has risen by an average of 50% during the same time period. It turns out that most of those acquisitions have been undertaken by tech incumbents from Silicon Valley.

If a founder is not willing to sell his company, a tech incumbent can copy the startup’s AI solution, therefore cutting the grass under the entrepreneur’s feet. Indeed, with its larger customer base and its greater credibility, a tech giant can rapidly capture large market shares – thus pushing the upstart firm out of business in its early stages.

In addition to M&As and imitations, tech incumbents are increasingly investing in startups in early or late stages through their corporate VC arm. Indeed, this corporate funding constitutes a hedging strategy on the part of Big Tech firms wanting to remain at the forefront of potential incremental AI innovations. If a startup’s solution is very promising, these initial investments pave the way for increased Big Tech participation or outright acquisition. For instance, Google’s VC arm placed a stake in the Canadian startup BenchSci that is developing an AI solution designed to speed up the discovery of new drugs. If the product is successful, a greater participation or an M&A by Google shouldn’t be excluded.

Another approach used by Big Tech to capture this top-notch grey matter – and therefore have a lead in innovation – is to set up their own AI labs in Canada. To make sure those R/D facilities successfully generate incremental innovations, global tech players recruit the best of the best of Canadian AI talents by offering them high salaries that contrast with the smaller and less certain pay in startups that have a big chance of failing. If having the brightest minds isn’t enough to produce cutting-edge applied research, those corporate labs work in close collaboration with local startups to monitor any incremental innovation from the outside that might of interest.  

Furthermore, those AI labs can potentially be a vector to hatch fundamental AI research. Indeed, in addition to keeping an eye on what startups are doing in terms of applied research, corporate R&D facilities pay close attention to fundamental AI research by working closely with academic institutes on the matter. Such are the cases of Facebook’s AI lab and Samsung’s AI centre recently established in Montreal who work closely with the MILA. Since IP protection is lacking in Canadian research institutes, the risk of corporate proprietary hold cannot be excluded. This is compounded by the fact that Big Tech partially funds fundamental AI research undertaken by academic institutes. For instance, in 2017, Google provided Montreal’s MILA and Toronto’s Vector Institute with C$4.5 million and C$5 million respectively.

In sum, as Big Tech firms aim to outsource their R&D efforts to Canada, the country faces the prospect of becoming a digital branch plant economy where AI talents conceive products for foreign tech players instead of building domestic AI champions. This entails no sale revenues for Canada as profits would flock back to the tech incumbents’ countries of origin. That said, one must not forget that Canada’s persistent vulnerability to tech incumbents is compounded by shortcomings specific to the Canadian AI ecosystem. Part four of this series will focus on those internal weaknesses.  

Photo: Artificial Intelligence (2018), by Gerd Altmann via Pixabay. Licensed under Creative Commons CC0

Disclaimer: Any views or opinions expressed in articles are solely those of the authorsand do not necessarily represent the views of the NATO Association of Canada.

About Alexis Amini

Alexis Amini – editor for the Canadian Armed Forces program – is a graduate student in public and international affairs at Université de Montréal (UdeM), Québec. He has a BSc in political science from the same university. Having lived in Djibouti and the United Arab Emirates where he witnessed major geopolitical events, Alexis developed a passion for international security. His research focus revolves around geopolitics, defense policies and political risk analysis. Upon completion of his master’s program, Alexis intends to join the strategic intelligence industry.