The Power of Mila#

At the heart of Montreal’s AI ecosystem stands the Montreal Institute for Learning Algorithms (Mila), a research institute that embodies everything unique about the Canadian approach to AI. Founded in 1993 by Yoshua Bengio, Mila traces its foundations to the Laboratoire d’informatique des systèmes adaptatifs (LISA) at the Université de Montréal and the Reasoning and Learning Lab (RL-Lab) at McGill University.

Unlike the often-siloed labs of the US, Mila is a unique consortium uniting researchers from Université de Montréal, McGill University, École Polytechnique de Montréal, and HEC Montréal under one roof. This physical proximity fosters unparalleled collaboration, with approximately 1,000 students and researchers and 100 faculty members working together as of 2022.

Government as a Catalyst#

The pivotal role of the Canadian Institute for Advanced Research (CIFAR) cannot be overstated. In the early 2000s, when neural networks were considered a dead end by most of the AI community, Geoffrey Hinton approached CIFAR with an idea. He had become convinced of the power of neural networks and their potential for deep learning in machines.

By early 2004, Hinton was leading CIFAR’s Neural Computation & Adaptive Perception program (NCAP), which included Yoshua Bengio and Yann LeCun among other neuroscientists, computer scientists, biologists, electrical engineers, physicists, and psychologists. This was a top-down strategic decision to own a future field, providing the crucial, patient funding that kept neural network research alive during the AI winter when US funding dried up.

The “Canadian Mafia” Identity#

The bond between Yoshua Bengio, Geoffrey Hinton, and Yann LeCun represents more than just professional collaboration—it’s a brotherhood forged in the crucible of academic adversity. These three researchers, sometimes referred to as the “Godfathers of Deep Learning,” received the 2018 ACM A.M. Turing Award together for their foundational work on deep learning.

This created a powerful, supportive, and distinctly Canadian network that operates with a long-term view, prioritizing fundamental research over quick commercial gains.

The Godfathers#

Yoshua Bengio stands as the heart of the Montreal ecosystem. Born in France to a Jewish family who had emigrated from Morocco, Bengio received his education at McGill University before becoming a faculty member at the Université de Montréal in 1993. His theoretical work on deep learning architectures, including word embeddings and attention mechanisms, was foundational to modern AI.

As of August 2024, Bengio has the highest Discipline H-index (D-index) of any computer scientist and is the most-cited living scientist across all fields by total citations. His decision to stay in Montreal and build Mila was the single most important factor in the city’s rise as an AI hub.

Geoffrey Hinton, while based in Toronto, represents the unwavering champion of backpropagation and neural networks.

The CIFAR Lifeline#

The CIFAR program that funded Hinton, Bengio, and LeCun for decades is the unsung hero of the AI revolution. Founded in 1981, CIFAR provided the patient, fundamental research funding that allowed these researchers to support students, run workshops, and keep the flame alive during the AI winter.

This represents the ultimate proof of the power of patient, fundamental research funding—a lesson that many other nations are now trying to replicate.

The Open Science Ethos#

The Canadian AI community has historically been deeply committed to publishing and open sourcing their work, accelerating global progress and attracting talent to their doors. Mila members have contributed significantly to open-source software, with Theano being one of the early programming frameworks for deep learning that originated at MILA.

The Bengio Flywheel#

Bengio’s reputation has created a powerful flywheel effect: his standing attracts top global talent to Mila, which produces groundbreaking research, which attracts massive investment from tech giants (Google, Microsoft, Samsung, IBM all have partnerships or labs in Montreal), which feeds back into Mila and spins up startups, which reinforces the reputation.

In December 2020, Mila teamed up with IBM to accelerate AI and machine learning research using open-source technology, integrating Quebec institute’s open-source software, Oríon, with IBM’s Watson Machine Learning Accelerator.

The Pan-Canadian AI Strategy#

In 2017, Canada launched the world’s first national AI strategy, the Pan-Canadian Artificial Intelligence Strategy. This strategy strengthens Canada’s leadership by fostering world-class research and top-tier talent in AI, with three National AI Institutes: Amii in Edmonton, Mila in Montreal, and the Vector Institute in Toronto.

The government’s commitment is substantial: Budget 2021 provided more than $443 million in federal support over ten years, including up to$208 million to CIFAR. This investment has helped Canada rank 4th in the world on the Global AI Index and third among G20 countries in net talent migration for people with AI skills.

The Startup Scene#

Montreal has given rise to a vibrant AI startup ecosystem. Bengio co-founded Element AI in October 2016, a Montreal-based artificial intelligence incubator that turned AI research into real-world business applications before being acquired by ServiceNow in November 2020.

The Policy Voice#

Yoshua Bengio has become a leading global voice on the ethical and societal implications of AI, advocating for its responsible development. He helped draft the Montréal Declaration for the Responsible Development of Artificial Intelligence and continues to raise awareness about the importance of mitigating potentially catastrophic risks associated with future AI systems.

In 2023, he was appointed to the UN’s Scientific Advisory Board for Independent Advice on Breakthroughs in Science and Technology, positioning Montreal not just as a tech hub, but as a thoughtful leader in the AI conversation.

Contrast with Predecessors#

The Verdict#

Montreal’s legacy is a powerful testament to the idea that sustained investment in basic science and a collaborative culture can, over time, yield world-changing returns. They didn’t just participate in the AI revolution; they persevered through its darkest days to help create it.

The Montreal model demonstrates that with patient capital, strategic government support, and a commitment to open science, even a mid-sized city in a cold climate can become a global powerhouse. They are the ultimate “patient gardeners” of AI—nurturing ideas through long winters until they bloom into world-changing technologies.

The Canadian approach offers a compelling alternative to the Silicon Valley model: instead of winner-take-all competition, they chose collaboration; instead of quick exits, they chose long-term cultivation; instead of proprietary research, they chose open science. The results speak for themselves.