8 months ago, I had the idea of creating a research paper study club at Western University. My direct motivations seemed obvious, but less obvious was the intended effect.
Was I trying to help people become bone fide AI researchers? Not really. Was I trying to help people find software jobs in AI companies? Not really. Was I trying to start a structured AI interest club? No. Was I trying to create a new friend group? Only a little.
7 months ago, an incredibly smart and kind group of folks gathered together each weekend to look at ML papers. I joined them a few times, and I was amazed by the depth we were able to achieve. Seperately, we built a good foundation for modern AI education at Western AI, the leading club for AI education at Western.
The lesson I learned was that when the environment is right, catalysts and consistency go hand-in-hand to create greatness.
Now, a school year later, here are the environment predictions I want to make for 2026, and what risks Western can take to meet the future.
You can build effective technical systems without the whole computer science / engineering environment.
AI Alignment will be a top 3 priority, especially in Canada. So will nationalized AI security, but this is less applicable to a university environment (especially one so far from Ottawa)
Western CS will continue to have faculty shortages and no dedicated building, let alone AI research faculty/facilities.
Motivations for students will lean even more towards baser instincts of power and money.
The key turning point for the remainder of this decade is to build better. We have the ability to, for the first time in living memory, build our way out of sticky situations instead of just redefining and politicizing (though those things will be heavily present as well). It’s not a remedy for the core societal and financial issues of our time, but at least we can make this layer a lot better.