When Google's DeepMind wanted to open its first research lab outside the UK, it didn't choose Silicon Valley, Boston, or even Toronto. It chose Edmonton, Alberta—a city of 1.1 million people better known for oil and hockey than artificial intelligence.
The reason? Rich Sutton and the University of Alberta's Reinforcement Learning and Artificial Intelligence Lab. In 2024, Sutton won the Turing Award—the "Nobel Prize of Computing"—for work he largely did in Edmonton.
Why Is Edmonton the Capital of Reinforcement Learning?
Edmonton is home to the world's leading research group in reinforcement learning (RL)—the branch of AI that teaches machines to learn through trial and error, the same way humans and animals learn.
While other AI research centers focused on neural networks or symbolic AI, Edmonton researchers bet on reinforcement learning—a discipline once dismissed as "the field that doesn't work." That bet paid off when RL became central to breakthroughs like AlphaGo and modern robotics.
Who Is Rich Sutton?
Richard S. Sutton is considered one of the founders of modern computational reinforcement learning and the 2024 Turing Award winner.
Key Contributions
Sutton's contributions form the foundation of modern reinforcement learning:
His textbook, co-authored with Andrew Barto, is used by virtually every top computer science department in the world.
"The chances of someone being able to create an AI not unlike human beings by 2030 is about one in four. And for every year past that, the chances go up dramatically." — Rich Sutton, on AGI timelines
What Is the RLAI Lab?
The Reinforcement Learning and Artificial Intelligence (RLAI) Lab at the University of Alberta is the world's leading research group in reinforcement learning.
Research Focus
The RLAI Lab aims to:
- Create new methods for reinforcement learning
- Remove limitations on RL's widespread application
- Develop RL as a model of intelligence approaching human abilities
Research spans mathematics, computational experiments, robotics, game-playing, and computational models of natural learning.
Key Faculty
Recent Publications (2024-2025)
- "Deep Reinforcement Learning with Gradient Eligibility Traces" (2025)
- "AGaLiTe: Approximate Gated Linear Transformers for Online Reinforcement Learning" (2024)
- "Real-Time Recurrent Learning using Trace Units in Reinforcement Learning" (2024)
What Was DeepMind Edmonton?
In July 2017, DeepMind opened its first international research office in Edmonton—its only lab outside the UK at the time.
Why Edmonton?
DeepMind chose Edmonton because of its reinforcement learning expertise:
- Rich Sutton was already a DeepMind advisor since 2010
- Roughly a dozen U of A alumni worked at DeepMind
- U of A alumni contributed to AlphaGo and Atari breakthroughs
- Edmonton was "the best place in the world" for RL research
"A big decision for us to open our first non-UK research lab." — DeepMind, on the Edmonton announcement
The Closure
In January 2023, Alphabet closed the Edmonton office while maintaining its Toronto and Montreal labs. The closure reflected a strategic shift:
- DeepMind Alberta focused on foundational RL research
- Google's AI priorities shifted to large language models (ChatGPT response)
- The Edmonton lab was the only DeepMind office not in a Google-owned building
The Legacy
The closure seeded Edmonton's AI startup ecosystem. Former DeepMind researchers went on to:
- Found Artificial Agency ($17M raised)
- Launch RL Core and other startups
- Continue research at U of A and Amii
- Join companies like Sony
Combined, these spin-off companies now employ more people than DeepMind Alberta ever did.
What Is Amii?
Amii (Alberta Machine Intelligence Institute) is one of Canada's three national AI institutes, alongside Vector (Toronto) and Mila (Montreal).
Key Funding
Canada CIFAR AI Chairs at Amii
Rich Sutton, Patrick Pilarski, and other Amii fellows hold Canada CIFAR AI Chairs, recruiting and retaining world-class researchers in Alberta.
What Is the Alberta Plan?
The Alberta Plan for AI Research is a strategic vision published in 2022 by Rich Sutton, Michael Bowling, and Patrick Pilarski.
Core Thesis
The Alberta Plan argues that understanding intelligence is "the greatest scientific prize in human history" and proposes a focused research agenda to achieve it.
Agent Architecture
The plan describes AI agents with four core components:
Key Principles
- Continual Learning: AI must learn on the fly without pausing
- Meta-Learning: Learning how to learn better
- Temporal Abstraction: Using options and subtasks
- Feature-Based Representations: Finding the right abstractions
The plan deliberately avoids trying to capture domain knowledge, instead focusing on general learning principles.
What AI Startups Has Edmonton Produced?
Edmonton's AI ecosystem has produced several notable startups, many founded by former DeepMind researchers.
Artificial Agency
Artificial Agency develops technology allowing game characters to interact with players more naturally and dynamically—using the same RL techniques developed at U of A.
Keen Technologies
Rich Sutton partnered with legendary game developer John Carmack to pursue AGI research at Keen Technologies, bringing Alberta Plan principles to a well-funded startup.
AltaML
AltaML is Alberta's largest AI employer, applying machine learning to health care, supply chains, financial services, energy, agriculture, and the public sector.
"I don't know if we would've started this business if the talent wasn't here. It's a huge advantage." — Nicole Janssen, AltaML Co-Founder
AltaML also runs talent accelerators, bringing in ~50 interns every four months and partnering with the Alberta government on GovLab.ai.
Other Notable Companies
How Big Is Edmonton's AI Ecosystem?
Edmonton's tech sector has grown significantly, driven by AI:
The University of Alberta has Canada's oldest computer engineering program and ranks in the global top 5 for AI research.
Who Are the Key Researchers?
Michael Bowling
Bowling leads the Computer Poker Research Group, which has built some of the world's best poker-playing programs.
Patrick Pilarski
Pilarski focuses on AI for medical devices and prosthetics, exploring how humans and machines can learn together.
Frequently Asked Questions
Why is Edmonton important for AI?
Edmonton is the world capital of reinforcement learning—the AI technique that teaches machines to learn through trial and error. The University of Alberta's RLAI Lab, founded by Rich Sutton, has over 100 researchers and produced the 2024 Turing Award winner. DeepMind chose Edmonton for its first international lab specifically because of this expertise.
Who is Rich Sutton?
Rich Sutton is a professor at the University of Alberta and the 2024 Turing Award winner (with Andrew Barto). He's considered one of the founders of modern reinforcement learning and has contributed fundamental algorithms like temporal difference learning and policy gradients. His textbook is the standard worldwide.
What is Amii?
Amii (Alberta Machine Intelligence Institute) is one of Canada's three national AI institutes, part of the Pan-Canadian AI Strategy alongside Vector (Toronto) and Mila (Montreal). Based in Edmonton, Amii has supported 300+ companies and facilitated $600M+ in venture financing.
What happened to DeepMind Edmonton?
DeepMind opened its first international research lab in Edmonton in 2017, led by Rich Sutton, Michael Bowling, and Patrick Pilarski. The lab closed in January 2023 when Google shifted priorities toward large language models. However, the closure spawned several AI startups, including Artificial Agency ($17M raised).
What is the Alberta Plan?
The Alberta Plan is a strategic vision for AI research published in 2022 by Sutton, Bowling, and Pilarski. It outlines a research agenda focused on creating agents that can learn continuously, adapt to complex environments, and approach human-level intelligence through reinforcement learning principles.
What AI startups are in Edmonton?
Notable Edmonton AI startups include Artificial Agency (gaming AI, $17M raised by former DeepMind researchers), AltaML (applied AI, 120+ employees), and Keen Technologies (AGI research with Rich Sutton and John Carmack). Former DeepMind staff have also joined companies like Sony.
How does Edmonton compare to Toronto and Montreal?
Toronto (Vector) leads in deep learning and enterprise AI, Montreal (Mila) leads in deep learning fundamentals and AI safety, and Edmonton (Amii) leads in reinforcement learning. Together, they form the "AI Triangle" of Canadian AI research. Edmonton is smaller but highly specialized.
Related Reading
The Alberta AI Ecosystem
- Rich Sutton: Father of Reinforcement Learning — The 2024 Turing Award winner who built Edmonton's AI reputation
- Amii Complete Guide — Alberta's national AI institute
Canada's AI Triangle
- The AI Triangle: Canada's Three Superpowers — How Toronto, Montreal, and Edmonton collaborate
- Vector Institute Complete Guide — Toronto's AI research hub
- Mila Complete Guide — Montreal's deep learning powerhouse
The Research Pipeline
- Geoffrey Hinton: Godfather of AI — The Nobel laureate who shaped Canadian AI
- Yoshua Bengio: Deep Learning Pioneer — Turing Award winner and Mila founder
- CIFAR Complete Guide — The institute that funded deep learning
The Broader Ecosystem
- Toronto-Waterloo Corridor Guide — North America's second-largest tech hub
- Canadian AI Unicorns — Billion-dollar Canadian AI companies
Join AGI House Canada to connect with Canada's AI research community. Scan the QR code to join our WhatsApp group or subscribe to our newsletter.
Data current as of January 2026. Statistics from Amii, University of Alberta, Edmonton Global, and industry reports. DeepMind closure details from The Logic and BetaKit.
