As a software developer, I sometimes use AI to generate complex code, which I find saves considerable time and error. I can see exciting possibilities in a tool that can help deal with complex issues such as the climate emergency, healthcare research, and improving regenerative agriculture.
On the other hand, as a political geek, I am very concerned about the growing risk to democracy and freedom that comes with AI. The generation and dissemination of misinformation and the ability to track people wherever they go is distressing. The potential for propaganda to make people unable to determine what is real and what is not is a huge risk.
Thus, I feel rather conflicted when there is talk of building an AI data centre in Île des Chênes. It would be hypocritical to just say “no” without some reasoned discussion. So here it is.
First and foremost, we are at the edge of a precipice in our destruction of our climate. We absolutely must stop burning stuff to generate energy. AI data centres are notoriously power hungry, and this one is no exception with the proposal of six gas-powered turbines. Any idea that creates more fossil fuel energy sources and emissions is a betrayal of future generations.
Another grave concern is the amount of water AI data centres use. The processors generate a lot of heat (a result of all that energy they use) and they need to be cooled to be efficient. That means water. Lots of water. In some places, AI centres can use millions of litres per year, causing droughts for residents. Locally in Southeast Manitoba, we are still facing a threat to our aquifer from Sio Silica, and an AI data centre represents another potentially dire cost to residents.
It won’t provide a load of jobs beyond some maintenance tasks to keep the machines running. It doesn’t really solve the problem of extracting Manitoba’s government from dependence on American data handling technology – this is not a classic web server / data storage site. It is a collaboration between an Alberta tech company and an American chip maker funded by Blue Owl, an American equity firm. Using our resources to make their shareholders wealthier.
So, without a hard “no,” I am feeling extremely cautious. We must require this project to align with climate and conservation goals and be accountable for the externalities (the cost the community and environment will pay). 350 acres is a lot of land – roofs and parking lots and land covered in solar panels could generate a lot of electricity and maybe have some spinoff for the community (whose residents complain about power problems already). Maybe explore the opportunity to put some agrivoltaics in there, too (growing crops or raising animals around solar installations has very positive outcomes). Water use could be reduced using a closed-loop system with chillers, but this is more expensive to build and more energy intensive and thus is not compatible with the profit motive.
We need to make sure the private investors are held to a higher standard than just burning stuff and draining aquifers. Otherwise, the residents of Southeast Manitoba may well end up subsidizing their profit with our air, water, and future.