If you don’t work in AI, chances are you’re getting a picture of it built almost entirely from layoffs, lawsuits, and existential risk. Those stories are real, but they offer an incomplete view of what's happening in this country, and just might have you rewatching The Terminator or The Matrix for survival tips. 

Last week, the Alberta Machine Intelligence Institute (Amii) hosted its annual AI conference Upper Bound in Edmonton. Eleven thousand attendees, seven stages running at once, coloured headphones so people could switch between sessions. Digital Journal was there as the official media partner.

Digital Journal’s CEO and executive editor, Chris Hogg, was at the event and felt compelled to publish a dispatch that lays out what's missing from the story your average Canadian is getting.  (You may want a fresh cup of coffee, it's a long one.) 

It's easy, sitting inside the tech bubble, to forget that not everyone has the time to sift through the constant changes in AI, or the hundreds of conference sessions, to get to a more nuanced picture. 

This article is for the average Canadian. It provides context and some of the many things going on in this country and is worth the long read.

Some of that context goes back 40 years. 

The Canadian Institute for Advanced Research started funding AI work in 1983. That's the program that brought Geoffrey Hinton here in 1987 and kept Richard Sutton at the University of Alberta. Both of them still work here, and together they're two of the three living scientists most responsible for modern AI.

Canada now ranks third in the world for AI research impact, behind only Stanford and Google, and is one of four countries with a sovereign foundation model, through Cohere. Edmonton's RL Core has built a reinforcement learning system that could help end boil-water advisories in Indigenous communities. A University of Alberta radiologist has built a pocket ultrasound at a fiftieth the cost of an MRI. The Alberta government built a property management system in-house for $2.6 million instead of the $54 million that the lowest external bid came in at. A medical scribe built at Alberta Health Services is used by nearly 1,000 Alberta doctors, with a version released under an open source licence so any clinic in the world can run it.

One reason many Canadians haven't heard about any of this is the old journalism rule Chris opens his article with: if it bleeds, it leads. The dominant AI story right now is layoffs. Meta, Amazon, Shopify, Telus, Scotiabank, all of them with AI somewhere in the press release.

Federal AI minister Evan Solomon called this "AI smog." He argues that most of the layoffs publicly blamed on AI are American executives covering for past over-hiring by repositioning themselves as forward-thinking innovators, and the smog drifts north every time another U.S. tech giant cuts staff and points at the technology. 

Chris points to Statistics Canada data from 2025 showing 89.4% of Canadian businesses that have adopted AI saw no change in staffing, and to the Bank of Canada's external deputy governor saying earlier this month she hasn't seen widespread evidence of AI replacing Canadian workers.

If you'd like a closer look at what happened each day at Upper Bound, you can read our daily summaries on Digital Journal:

Upper Bound opens with questions on AI ownership, data sovereignty, and institutional knowledge (day one). Mozilla's Mark Surman made the case for open source as a route to ownership. Qohash's Jean Le Bouthillier argued that the standard model of copying data offshore to secure it has run its course. And the University of Alberta's Ross Mitchell walked through the AHS infrastructure already in use across the province.

Day two at Upper Bound pushes AI from theory to practice across industry, government, and talent (day two). Energy economist Peter Tertzakian and former Suncor president Mark Little drew lessons from the energy sector that the AI industry could learn from. Alberta minister Nate Glubish told the room how his team killed a $54 million RFP and built the same system in-house.

Upper Bound speakers urge Canada to hold on to its AI advantage (day three). Solomon announced $24 million for 42 new and renewed Canada CIFAR AI Chairs. Solink CTO Martin Soukup explained why building an AI model that works at 10 customer sites is just the start. From there to 10,000 is mostly systems engineering.

The work being done in AI has the potential to make a significant impact on society, for better or worse. Canada has an opportunity to shape it in a positive direction by continuing to invest in the people, infrastructure, and ideas already in motion, and by telling the story of what's already been built.

"You get what you celebrate," said Amii CEO Cam Linke on stage.

Digital Journal will have deeper coverage of the sessions, speakers, and themes in the weeks ahead. To follow along, visit digitaljournal.com, follow us on LinkedIn, or keep reading this newsletter and share it with friends or colleagues you think should know more.

— David