@EliotPence Each of these will generate decades of operational data — inspections, telemetry, maintenance, outcomes. The training set the next industrial economy runs on. Canada's building the energy; it needs the data layer to match. Otherwise we export electrons and import intelligence
8/ Three pilots, 18 months, published metrics, then a scale/no-scale call in the open. The starting data standard is public — written to be argued with.
Full piece: https://t.co/8Q1TX1JeU3
1/ Canada is about to build at a scale it hasn't attempted in generations. Mines, transmission lines, ports, northern corridors — $200B in, path to $500B.
Every one of those projects generates something nobody's thinking about. 🧵
7/ And honestly: licensing is upside, not the payback. The real return is national capability — the training data Canadian robotics and automation actually needs.
@ItzSuds May have a fit for you. Client iss developing a 380MW+ data centre campus in Colorado. Land secured, fibre on site, grid interconnection underway, and they're looking for anchor tenants. Happy to connect you with the team.
Carney launched a National Council on AI today — 13 sharp people. I read the list twice looking for someone there for the data. There isn't one.
On the missing seat at the table:
https://t.co/OINs0zQUUt
This week a CAD company launched AI that designs for you — and admitted it's trained on historical design archives. Your 30 years of CAD files are the fuel. That's either a payday or an obituary.
Your call.
7/ The US has Mercor, Micro1, Scale — an entire industry doing the unglamorous work of finding corpora, clearing rights, moving them to labs at speed.
Canada has no equivalent. The compute got billions. The data got a conversation.
8/ Three things have to happen:
1. The labs need a Canadian data partner (Cohere’s the obvious anchor)
2. Fund data infrastructure like compute infrastructure
3. Tell corpus owners they’re sitting on strategic assets — most still price them at zero
5/ Canadian health data is fragmented across provinces. Everyone calls this a problem. It’s the opportunity — whoever assembles it with proper consent holds something no lab can scrape.
4/ People sleep on what Canada holds: French/English bilingual corpora (globally scarce), CanLII (best-structured legal corpus in the world), decades of forestry/mining/energy operational records nobody else has.
3/ This isn’t a future problem. The window where Canada’s data is still Canada’s to direct is open right now — and it closes with every deal signed somewhere else.