Hey, Canada: time to replace eh with AI
John Stackhouse
Senior Vice-President, Office of the CEO, Royal Bank of Canada. Host of Disruptors, an RBC podcast
It’s been a big week for AI, in Canada and abroad.
Let’s start with the weather, a Canadian obsession. Google DeepMind team published the results of a competition between its generative AI model, GenCast, and Europe’s world-class weather forecasting model. GenCast outperformed the European Centre for Medium-Range Weather Forecasts 97.2% of the time, and was able to produce better forecasts in minutes while Europe’s physics-based model took hours.?
Those machine learning models are getting better and faster. (Check out this article in Nature for more).?
To keep pace — or at least not fall further behind — the federal government made some important announcements this week, which provinces, universities, companies and public sector organizations like hospitals all need to build on. Here’s what stands out for me in the?Canadian Sovereign AI Compute Strategy:
Some this is previously announced money put together in a new package — but old or new, it’s critical now to get the investments flowing … and glowing with success stories that can inspire all Canadians to embrace Gen AI.?
We’re falling behind other advanced economies, especially the U.S. A report published this week from my friends at The Dais at Toronto Metropolitan University , indicates AI adoption is still nascent in corporate Canada, and tends to be found largely in large firms and younger firms.
The new strategy can address that. In fact, a critical part of that strategy is a big bet on Cohere , one of Canada’s AI darlings, to build a data centre that will help it take on the world, and to help other Canadian tech companies take on the world. We haven’t seen Canadian governments bet on companies in this way as much as our U.S. and European competitors do.?
That needs to change, and it would be smart for future governments to think about doubling down on an ambitious AI strategy. Spoiler alert: the Trump administration is going to be more AI-focussed than any government other than China has ever been. And if we don’t keep pace, a lot of our AI talent, data and potential across the economy will move south.
Cohere may also help with that with a new focus on small that it explained this week. With all the excitement about large language learning models, which essentially study as much available information as is out there, Cohere is developing AI agents to study contained data sets, which can be cheaper, easier, and more relevant to a narrowly focused business. A classically pragmatic Canadian approach.
领英推荐
Canada’s new AI strategy can also help the middle tier of our economy speed up its use of these powerful tools, just as we’re seeing so many American firms do. Think how much more competitive our mining, health care, retail and transportation sectors — among others — can be with AI.?
There’s no time to lose. KPMG Canada published a good report this week on the finance sector (a Canadian strength) that shows:
It’s been a big year for Gen AI, and we’re just getting going. Get ready for 2025 to be faster, smarter and more practical than anything we’ve seen.
For more insights, check out:
We’ll have lots more in the months ahead on Canada and Gen AI. Sign up for our newsletters and podcast to keep pace!
Shaping the Future of Ethical AI | AI Governance and Compliance | Digital Transformation Lead | Implementing Scale Agile Framework | Executive Coaching | Enterprise Change Mgmt.
3 个月Well-said
Chief Operating Officer @wearelaivly, Ex EVP IntouchCX - Investor, Futurist, CX, Brand Stewardship, AI, SaaS, Culture & People. Converged intelligences are the future.
3 个月I'm not sure that access to compute is the barrier to adoption that policy is making it out to be. When were talking "AI adoption is still nascent in corporate Canada," compute is already baked into the products that Canadian companies are not using, or available off the shelf from AWS etc. Rather - there's two things at play: (1) Canadian corporate purchasing cycles move glacially. A year to identify the need, a year to go shopping and infight about build vs. buy, a year to budget and do legal and compliance on the solution, a year to implement and realize there's another 24 months of secondary work to do to make anything work, etc etc. (2) Canadian orgs "Classic pragmatism" or plodding conservatism - per examples above, purchasing cycles are achingly slow b/c they reflect a "wait and see" and "our last spend isn't fully depreciated" mindset that makes (in agg) Canadian business slower to move than comps in the US. Government investment in compute is good, but as productivity lever, I think it will be a miss. The value there is for a next wave of innovation, not driving adoption of innovations already out there and readily available.
Owner, Altima Millwork, Altima Homes and Altima Kitchens And Closets | Transforming Spaces with Precision and Style | Elevating experiences with Commercial Millwork
3 个月Insightful article, John! It’s exciting to see how Canada continues to play a pivotal role in the evolving AI landscape. Thanks for keeping us updated on these groundbreaking developments.
Technical Client Advisory Executive @ IBM Quantum | Author of Publications and Patents
3 个月We Canadians founded the AI revolution by funding and incubating talents like Geoffrey Hinton and others at CIFAR. We need to bang our chests a little more about this. At the same time we need to be better at commercialization and innovation. Perhaps we can look to Canadians like Katherine Homuth (though she is not in AI) founder of Sheertex or Tobias Lutke (also not exactly in AI) founder of Shopify as examples. I’m not sure we have any examples at that scale in AI yet? Please let me if I miss someone.
Director of Data Science,RBC
3 个月Very helpful!