As Canada’s tech community gathers for the Elevate Festival starting Oct. 7, conversations around artificial intelligence are taking centre stage.
Occupying an important place on that stage will be Alex Kearney. Kearney is an AI researcher connected to DeepMind, Twitter, and the University of Alberta. She is also developing an AI-powered “behaviour engine” for video games at Artificial Agency, which she co-founded.
Kearney will be at Elevate next week talking about the surprising and vital links between gaming and AI research. Other AI sessions at Elevate involve policy leaders from Cohere and OpenAI, Meta researcher Setor Zilevu, and Spotify director David Nyhan, among other key figures.
BetaKit sat down with Kearney to learn more about her journey from academia to starting her own AI company. She reflected on Alberta’s long history of AI research, how gaming has played a key role in that research, and the benefits and pitfalls of launching an AI startup in Canada.
This interview has been edited for length and clarity.
You were previously a DeepMind scientist. Was it a challenging leap from researcher to helming your own AI company?
It definitely is a transition, because the way you think about problems and what you focus on is different, but I found it personally to be really rewarding.
I spent over 10 years working on AI research, both in industry and academia. But now I’m focused on a very different set of objectives. It’s really just delivering tools so that these designers can create entirely new genres and games.
One of the most invigorating things for me is when we share the tools that we’ve developed with a new game designer or a new company, and just watching folks light up with that inspiration, that moment where they realize, “Oh my gosh, I can do this thing that I hadn’t even contemplated before.”
That really gets me fired up.
What prompted the focus on AI agents for gaming versus other categories?
Games have been a test bed for AI agents for decades. Getting superhuman performance in something like Atari, or a game like StarCraft, those have been high watermarks, from an academic perspective, to show we’re making systems that are competent.
Play is such a fundamental thing, right? It’s a part of how we come to understand the world around us. It’s how we learn, and it’s also how we identify intelligence in many ways.
One of the things that I strongly feel is how chat interfaces let us see the potential of LLMs as intelligent systems. Because you can naturally engage with these systems in a language that’s universal: your own human language. I think through games and play, we’re really going to have that “aha” moment where the public can feel what intelligence is like in a systematic environment.
How do you see the AI industry evolving in Edmonton, and how does that compare against Canada at large?
One of the things that often surprises people is the history of the AI ecosystem in Alberta, and how far back it goes. We’ve been an AI research center for over a quarter of a century.
One of the things that I find heartening is the way that within our Alberta ecosystem, we’re acknowledging that we have this rich tradition in machine learning and reinforcement learning, and we’re looking at ways to broaden it throughout the different academic disciplines.
Over the past few years, there have been a lot of really exciting breakthroughs globally in that area. And that’s a trend that I think is going to continue, and I feel like the university and the province are really investing in that area.
What do you see as the advantages and pitfalls of launching an AI startup in Canada?
I think there’s sort of a can-do attitude. At least within Alberta, there’s a real innovative, take-a-risk attitude that I find, personally, quite inspiring.
But there’s also a humility that counters that. I feel like we are outside of the hype bubble, so to speak, by building in Canada, and specifically by building in Alberta, and that promotes a sense of clarity that I really quite like.
There’s a double-edged sword to that humility, though, which is that we’re often not as loud. We don’t celebrate our successes, even on a federal level, as much as some of our global counterparts.
So I think one thing we can get better at is celebrating our wins, and that’s something that festivals like Elevate really help facilitate.
Do you see events like Elevate as important to fighting “brain drain” to the US? Or even reversing the trend?
I think getting researchers, investors, and innovators in the same room is important. And I think creating the scenario where you get to see what other people are building is important. Success begets success, and seeing positive examples in the ecosystem helps others think to themselves, “I can do that. It’s possible. I could build here. I could raise capital here.”
And so to that extent, having these sorts of celebrations of Canadian innovation and people building locally is important, because those proof points help the next generation see what’s actually possible.
BetaKit is an Elevate media partner.