Jesse Sharratt grew up in P.E.I., moved to Toronto, built a tech company to seven figures, won a $13.8 million federal contract, then watched it get cancelled after the election. 

He went home to Halifax and started picking up garbage with his toddler.

No, I mean that literally. What started with picking up close to 100 bags of garbage on the walk to daycare became a platform that tags litter by brand and holds corporations accountable for what ends up on the street. My colleague Jennifer Kervin has the full story.

Sharratt's venture studio, Atlas Atlantic, builds on that approach: community-rooted technology, no outside capital by design, more than 20 projects in various stages. His bet is that Atlantic Canada's social fabric (the neighbourliness, the slower pace, the tight community connections) is a competitive advantage for building technology that's supposed to serve people.

Atlantic Canada has been exporting its best people for decades. Sharratt followed that path. Now, he's back and charting a new one.

The same idea shows up from a different angle in an article I wrote this week about AXL, a Toronto venture studio co-founded by Daniel Wigdor, a U of T computer science professor whose early research helped lay the foundation for multi-touch interfaces. In that case, the technology had roots in Toronto. The returns went to Cupertino.

He sees the same thing happening with AI. Canadian students co-founded OpenAI. Canadians lead AI at xAI, Meta, and Google. At the same time, fewer than one in four Canadian firms have fully implemented AI, compared to roughly one in three globally.

"We've done a better job of recognizing the contributions of Canadian scientists," Wigdor says. "But we've done no better in terms of commercializing it here."

Sharratt is keeping value local by staying rooted in the community. Wigdor is doing it by surrounding research with commercial infrastructure before the talent decides to leave.

Some light reading for when you're between chocolate eggs and existential dread about Canadian competitiveness

What agentic AI breaks and why nobody sees it coming Jennifer Friesen spoke with Lenovo's field CTO for enterprise AI ahead of his session at the CIOCAN Peer Forum in Vancouver on April 15-16. His read: only 10% of organizations are ready to deploy agentic AI at scale, and the ones moving fastest are most likely to bleed first.

SR&ED reforms take effect Canada's main R&D tax credit just got its biggest overhaul in more than a decade. Expanded eligibility, higher thresholds, capital expenditures now included, and a new pre-claim approval pathway with an eight-week turnaround. If your organization conducts any R&D work, the eligibility rules have changed enough to warrant a fresh look.

Astronauts called Houston about Outlook this week NASA's Artemis II launched April 1 on the first crewed moon mission in more than 50 years. Within hours, Commander Reid Wiseman radioed Mission Control with an issue: "I have two Microsoft Outlooks and neither one of those is working." Houston was able to help fix it. The toilet had also broken earlier. They fixed that too. No word on whether turning it off and on again was the answer to either.

The technology that helped define the modern smartphone had roots in a Toronto lab. The returns went to Apple shareholders. Many of the researchers who helped build modern AI studied at Canadian universities. The companies they built are headquartered somewhere else.

Canada keeps planting. Someone else keeps showing up at harvest time.

Sharratt's answer is to build something so rooted in its community that it can't easily be uprooted. Wigdor's is to surround the research with the commercial infrastructure before the talent leaves. Both are worth keeping an eye on.

I'm off to hide some eggs. The bunny suit is itchy, but I get to eat leftover chocolate.

— David