Skip To Main Content
An abstract image of two fingers about to touch, with high-tech lights shining on them
Data and technology

AI leaders chart course for tech in Montreal

At the ALL IN conference, the top talent in AI discussed key opportunities and challenges within the Canadian landscape

Last week, the leading voices in Canadian Artificial Intelligence gathered at the All-In Conference in Montreal. From responsible design and deployment to impacts on the Canadian economy and support for Canadian AI, the event held many insights into how the accounting profession can support innovation and govern and build trust in technology. Here are some key “Need-to-Know” messages for CPAs across and Canadian businesses coast to coast to coast. \

Read More 

Adopt early, experiment early  

Canadian businesses are lagging behind the U.S. in adopting AI technologies and that’s partially due to a more risk-averse culture. While caution is important, low adoption rates are counterproductive with fast-evolving technology like AI. Experimentation and willingness to learn from failures are critical for finding effective use cases and projects. It's better to address smaller problems that AI can easily solve rather than attempting to apply AI universally at once. 

That’s not to say there aren’t risks that need to be managed, but we need to find productive ways to experiment and support innovation without outright dismissing or avoiding new technology out of fear or lack of understanding. As leaders, CPAs should seek out success stories and share their own. Get involved with working groups and industry think tanks to find opportunities to best deploy AI in responsible ways. We’re all in a time of learning and experimentation.  

Think of experimentation as tuition costs 

One idea that made the rounds on the floor of the conference was the need for a cultural shift to view potential failures as 'tuition costs' for learning what works. The return on investment from failed projects comes from the knowledge gained. However, not every project needs to be greenlit. Organizations should develop a framework to evaluate projects based on valid applications, execution ease, available data, and potential tangible impact. 

Scale responsible AI with proportionality and safety. 

Deploying AI responsibly entails designing it to be safe, transparent, controlled, and aligned with human expectations. It also means using AI proportionally to the task. One presenter used the example of ensuring knife safety by opting for a butter knife when suitable. Similarly, deploying AI tools should match the task's requirements without necessarily using the most powerful or costly models, which might not always be the best solution.  

“There is a lot of ongoing research around responsible and safe deployment of AI, and how governance and safeguards can play an important role in this journey,” says Taryn Abate, CPA Canada’s director of research and thought leadership. “CPA Canada will be zeroing in on this in an upcoming paper, stay up to date on AI here. 

Recognize the environmental cost of AI 

Two narratives exist about AI and sustainability: AI has great potential to address climate issues but is also a significant carbon emitter. Training and powering AI models require immense electricity, alongside the resources needed for the physical infrastructure. Sasha Luccioni from Hugging Face, who spoke at the conference, highlighted the responsibility of understanding AI's environmental impact. A key takeaway is recognizing the importance of considering the environmental cost of running AI models and the exponentially larger impact of training new models versus refining existing ones. 

Call for Canadian investment – drive local demand. 

Discussions underscored the need for substantial investment to build compute infrastructure, fund AI research, and boost Canadian entrepreneurship and investment friendliness. With Canada’s world-class AI talent coming out of top research institutes like Mila in Montreal, the Vector Institute in Toronto and Amii in Edmonton the potential benefits are immense, as are the penalties for not acting quickly.  

One thing was clear at the conference, bridging the gap between research and commercial application is crucial for advancing Canadian AI — and even more crucial to attract necessary investment.  

“CPAs are in position get on the ground-floor of the leading-edge technology of our day by using their expertise and acumen to help assuage some of fears surrounding AI,” says Abate. “This could encourage investment and  help build the Canadian AI landscape. The profession only stands to gain.” 

As part of CPA Canada, professional accountants have full access to countless resources surrounding AI and technology via our digital news section, the award-winning Pivot magazine and our Data and Technology resource page.