AI data centres are coming to Alberta. Let’s get the numbers right.
I’ve run data centres since 1994. I’m not opposed to building more of them. I am opposed to approving enormous AI campuses on vague promises about jobs, water, power or public benefit.
There is plenty of nonsense on both sides of this debate. A bad claim does not become true because it supports the side you already agree with. This site is my attempt to sort out what is real, what is exaggerated and what Alberta should require before saying yes.
What I’m looking at
Power. How much does the project need, where will it come from and who pays for the infrastructure?
Water. How much water is actually consumed? A closed cooling loop, an evaporative tower and a gas power plant are not the same thing.
Noise. What reaches the neighbouring property at night, during generator testing and during normal operation?
Land. An industrial site beside existing heavy industry is a very different argument from productive farmland beside homes.
Ownership and benefit. If Canada supplies the land, gas, electricity and approvals, what stays here?
Cleanup. Get the obligations and financial security settled while the developer still wants permission to build.
Start with the projects
The project map tracks publicly identified Canadian AI data-centre proposals. The project pages pull together what has actually been filed, what has been promised and what is still missing.
If you hear about a project I have missed, the useful clues are usually the company name, legal land description, power application, municipal file or even a proposed fibre requirement. Early projects do not always arrive with “AI data centre” written on the front page.
Where do you draw the line?
I also want to know what people actually object to. Industrial land? Farmland? Potable cooling water? Natural-gas generation? Foreign ownership? Or the idea of a data centre at all?
The public vote asks those questions separately. It is not a scientific poll and there is no score at the end.
The AI part matters too
These projects are being built because AI companies expect the compute to be worth a great deal of money. That raises another set of questions about Canadian access, ownership, labour displacement and who benefits if AI becomes as productive as its builders expect.
AI futures looks at those scenarios. Canadian Compute looks at what Canada could ask for while the infrastructure is still being built.