In the late 1980s, Geoffrey Hinton, who had been teaching at Carnegie Mellon University in Pittsburgh for a few years, grew increasingly concerned about the political climate in the United States. Disillusioned with Ronald Reagan’s foreign policy and the military funding of his A.I. research, he was eager for a change. When an opportunity arose to move to Canada, he embraced it.
“My wife and I were quite frustrated with the U.S.,” Hinton shared with Observer, “and Canada appeared to be a promising alternative.” Lured by Canada’s robust social system and an appealing fellowship at the Canadian Institute for Advanced Research (CIFAR), he relocated to Toronto in 1987 and has largely remained there ever since, even earning a Nobel Prize for his contributions to the field of A.I.
Hinton’s move marked just the beginning. Generous funding for exploratory research in Canada attracted numerous pioneering A.I. scientists, resulting in significant breakthroughs that laid the groundwork for today’s leading A.I. technologies. In 2017, Canada further solidified its commitment by becoming the first nation to establish a national A.I. strategy, centering much of its innovation in three main hubs located in Toronto, Montreal, and Edmonton.
However, despite its notable contributions to the burgeoning tech landscape, many experts argue that Canada has not fully capitalized on its innovations. The country has seen a significant outflow of talent, with many Canadian researchers relocating to the United States. “Historically, Canada has been home to inventors and pioneers of technology but hasn’t always enjoyed the commercial benefits,” noted Cam Linke, who heads the Alberta Machine Intelligence Institute (Amii), in an interview with Observer.
While efforts to foster competitive A.I. companies in Canada have seen mixed results over the years, recent increases in government funding, strengthened research institutions, and evolving cultural attitudes are beginning to yield positive outcomes. A notable example is the Toronto-based startup Cohere, which raised an impressive $500 million earlier this year—an unprecedented sum for a Canadian generative A.I. startup—thanks to a mix of Canadian, American, and international investors. Although Canada’s “brain drain” remains a concern, Cohere co-founder Nick Frosst expressed optimism, stating, “I believe things are shifting.”
Drawing Top Talent
Long before the advent of companies like OpenAI and Anthropic, Canada stood out as a destination for those passionate about ambitious A.I. research. While the nation may have had less federal funding compared to the U.S., it provided a nurturing environment for those engaged in long-term experimental projects. Hinton observed that, due to its supportive social system and funding for foundational research, “three researchers were quite content to live in Canada”—referring to himself, Rich Sutton, and Yoshua Bengio. The latter two would later earn the title of “Godfathers of A.I.” after jointly receiving the 2018 Turing Prize with Yann LeCun of Meta.
After Hinton established his position at the University of Toronto in the late 1980s, Sutton, an American known for his influential work in reinforcement learning, moved to the University of Alberta due to his disillusionment with U.S. politics. Meanwhile, Bengio returned to his hometown of Montreal to work at the University of Montreal. Their collective presence, combined with Canada’s more welcoming immigration policies, attracted even more A.I. researchers, as noted by Amii’s Linke. “This created a cycle where great talent wanted to collaborate with these pioneers,” he explained.
Although based in different parts of the country, Hinton, Sutton, and Bengio shared a common passion for a specific area of A.I. research that was historically overlooked. Hinton described the division within the field: “There was traditional A.I. and then there were neural networks, which were often seen as oppositional camps.” Traditional A.I. focused on symbolic reasoning, while the neural network approach aimed to emulate the workings of the human brain.
Despite being dismissed as a “crazy theory” at the time, neural networks received backing from CIFAR. For instance, in 2004, CIFAR launched the “Neural Computation and Adaptive Perception” program, which was directed by Hinton and involved contributions from Bengio and LeCun. Hinton noted, “It took time for neural networks to show practical applications, necessitating funding for researchers who could deliver results later on. Securing such funding was far more challenging in the U.S.”
Researchers involved in this program convened annually to exchange ideas, as recounted by Ruslan Salakhutdinov, a professor at Carnegie Mellon University. In 2005, Salakhutdinov had distanced himself from A.I. to pursue a career in banking until a chance encounter with his former mentor Hinton on the street rekindled his academic ambitions. Hinton showcased his latest advancements in deep learning, inspiring Salakhutdinov to return to academia and pursue a Ph.D. under his guidance.
As the early 2010s rolled around, excitement surged within Canada’s neural network community following significant breakthroughs, like improved speech recognition. Hinton, alongside his students Krizhevsky and Sutskever, gained widespread attention in 2012 after winning an object recognition competition using neural networks. Their success paved the way for the startup DNNresearch, later acquired by Google for $44 million.
With the growing recognition of neural networks, leading researchers such as Hinton, Sutton, Bengio, and LeCun received attractive offers from tech giants like Google, DeepMind, and Meta. Many emerging Canadian researchers also departed for the U.S., drawn by lucrative job opportunities.
In response to the ongoing brain drain in A.I., the Canadian government launched the Pan-Canadian A.I. Strategy in 2017, investing billions into A.I. research initiatives. This effort established three major Canadian A.I. hubs, with prominent researchers like Bengio, Sutton, and Hinton at the forefront.
Even with advancements in A.I. research, Canada’s tech sector has been slow to embrace these innovations. Companies like BlackBerry and Element AI have struggled to leverage neural networks due to conservative business practices and financial hurdles. Moreover, the University of Toronto has faced challenges in promoting entrepreneurial ventures within its academic environment. Students leveraging university resources for startups often had to relinquish a larger share of equity compared to their American counterparts at elite institutions like Stanford and Carnegie Mellon, as noted by Salakhutdinov. The University of Toronto clarified that such equity stakes are negotiated individually, typically ranging from single digits to low double digits.
Additionally, Canada grapples with a notable deficit in computational infrastructure compared to the U.S. Hinton identified this as a significant obstacle for aspiring young researchers. He recounted a case where former student Jimmy Ba encountered difficulties accessing the necessary graphics processing units (GPUs) essential for training large language models while at the Vector Institute, leading Ba to join Elon Musk’s A.I. startup xAI. Hinton expressed concern that Canada may struggle to attain global leadership in A.I. due to these resource limitations, despite its historical success in fundamental research.
While a considerable number of researchers have opted to leave Canada, some have chosen to remain. Additionally, international companies have established research facilities in Canada, creating opportunities for local graduates. Lacoste-Julien, who leads a Samsung lab at Mila, acknowledged the positive influence of these international offices in retaining talent within Canada post-graduation. He noted that although the brain drain issue isn’t fully resolved, progress is being made.
The cultural values prevalent in Canada, which initially attracted luminaries like Hinton and Sutton, might present challenges in competing against foreign adversaries in the A.I. sector. In regions such as Quebec, there’s a strong emphasis on quality of life and social equity over conventional success metrics associated with corporate growth. Nevertheless, some innovative startups are beginning to disrupt the status quo. For example, Artificial Agency, founded by former Google DeepMind researchers, secured $16 million in funding this year for its innovative generative A.I. solutions aimed at enhancing gaming experiences.
Canada’s startup ecosystem is witnessing a resurgence, particularly in cities like Toronto. In 2022, the Canadian A.I. sector attracted a remarkable $8.6 billion in venture capital, establishing the country as a leading destination for A.I. investments. Companies such as Waabi, focused on autonomous vehicle technology, and Cohere, a rising star in the A.I. arena, have garnered significant attention and funding. The growing support from firms like Radical Ventures, a notable Toronto-based venture capital firm, highlights the increasing potential of Canada’s A.I. landscape.
Collaboration between businesses and research institutions in Canada, such as Vector and Amii, has further accelerated the growth of startups in the country. The tendency of companies in regions like Ontario to focus on building and growing locally reflects a noteworthy shift in mindset. The experience of Artificial Agency illustrates the evolving interplay between startups and academia, highlighting a commitment to attracting and retaining top graduate talent.
As Canada’s A.I. ecosystem continues to develop, a new tradition of successful startups is emerging, paving the way for future generations of researchers and entrepreneurs. The shift towards cultivating a robust local A.I. industry brings renewed hope for a sustainable and innovative future within Canada’s technological landscape.
