Key takeaways from the All In 2023 conference

October 10, 2023

News Trustworthy AI

By Natalie Richard

Members of the Vector Institute’s senior leadership shared insights with attendees at the All In 2023 conference in Montreal last week. Cameron Schuler, Vector’s Chief Commercialization Officer and VP, Industry Innovation and Deval Pandya, Vector VP of AI Engineering discussed the state of the Canadian AI ecosystem, AI trust and safety, and how AI can be harnessed to address climate change at the inaugural conference. Billed as Canada’s largest gathering dedicated to Canadian AI, the event brought together over 1,4000 participants, 170 conference speakers, and more than 100 AI startups. The first day of the conference coincided with the announcement of a new national AI code of conduct and the release of new data showcasing the robustness of Canada’s AI ecosystem. 

Over the course of two days, All In showcased a series of industry-specific panels. These discussions, led by industry experts and researchers, covered sectors like healthcare and manufacturing, in addition to topics around AI trust and safety, bias and diversity in AI and more. 

Cameron Schuler, Vector VP Industry Innovation, signs a national voluntary AI code of conduct with Minister of Innovation, Science and Industry François-Philippe Champagne.

AI code of conduct

Minister of Innovation, Science and Industry François-Philippe Champagne kicked off the first day by announcing a national voluntary AI code of conduct. The code was later signed by key members of Canada’s AI ecosystem. All of the signees, including CIFAR, Canada’s national AI Institutes — Vector, Amii, and Mila — and Vector Bronze sponsors Cohere and Ada, pledged to assess any potential negative impacts of AI systems.

Safe AI in Canada’s ecosystem

Following the announcement, Champagne joined a panel to discuss Canada’s potential as an international hub for accelerating AI. Yoshua Bengio, Scientific Director, Mila & IVADO, made it clear throughout the conversation that while we should be optimistic about the opportunities of harnessing AI, we cannot forget the risks. 

“We need to act to make sure that we can benefit from everything AI can bring to society,” he said. “That is one thing that Geoffrey Hinton said last spring. For each dollar we invest in making AI more capable, we need to invest a dollar to make sure AI is responsibly applied.” 

The group’s discussion also touched on AI talent, a key point raised throughout the course of All In, especially following the release of Deloitte Canada’s Impact and Opportunities: Canada’s AI Ecosystem – 2023 report that same morning. A key takeaway from the report was that Canada’s AI talent growth had outpaced other G7 countries.

Cameron Schuler, Vector VP of Industry Innovation (fourth from left) takes part in the panel about the progress of AI research.

Major AI institutes and research transfer

Schuler joined a panel on the progress of AI research to offer an industry perspective on advancing Canada’s AI ecosystem, AI research, development and applications, and AI trust and safety. The panel also included Valérie Pisano, President and Chief Executive Officer, Mila; Cam Linke, CEO, Amii; Elissa Strome, Executive Director, Pan-Canadian AI Strategy, CIFAR.

Pisano said that the new code of conduct “echoes what Yoshua Bengio was saying earlier, that a lot of research efforts have been about capabilities in AI but the flipside is that we really need to accelerate our AI safety research and that is where our Canadian ecosystem can play a role.”

“It comes down to practicality,” Schuler added. “We all want the world to be a better place because of AI. How we implement things like the code of conduct or our trust and safety principles is the next challenge.”

The group also dove into the findings of the new national AI ecosystem report conducted in collaboration between Deloitte Canada, CIFAR, Amii, Mila, and Vector.

Blob

Impact and Opportunities: Canada’s AI Ecosystem – 2023 report

  • Canada leads the world in the growth of our AI talent concentration, and is second in the world in the increase in AI patents. 
  • Canadian AI researchers produced more AI publications per capita in 2022 than any other G7 nation.
  • The Pan-Canadian AI Strategy has created significant social and economic benefits for Canada, with our AI sector outperforming that of many of our G7 peers.
  • Employment in Canada’s professional, scientific, and technical-services sector grew by 11.06% from 2021 to 2022—the fourth-largest growth seen in any sector in Canada that year. 
  • In Canada, there were 140,418 actively engaged AI professionals in 2022, an estimated 29% increase over the previous year. 
  • In the past year 20,634 AI jobs were created in Ontario while a further 75,975 AI jobs were retained in the province.

“I don’t think we are hearing companies say ‘no thanks,’” said Schuler when asked why he thought some companies are adopting AI and not others, “what we’re hearing is ‘how do we get started?’ There are challenges there.” 

Players from across industries were present at All In, talking about the transformation that AI is already making possible. Many more were talking about AI adoption for businesses and how to harness AI responsibly to drive their sectors forward.

AI for the environment and how AI is transforming #ClimateTech

There were a handful of panels dealing with climate change and technology at the conference. Highlights included Deval Pandya, Vector’s VP of Engineering participation in How AI is transforming #ClimateTech where panelists talked about AI applications in ocean studies and conservation, coastline management, and agriculture technology. AI’s potential in ocean conservation was emphasized, including its use in detecting marine mammals, tracking marine litter, and improving how coastlines recover from natural disasters like hurricanes and tropical storms. This was punctuated by real-life examples of AI applications like monitoring fish populations and invasive species,, shared by Kendra MacDonald, Chief Executive Officer, Canada’s Ocean Supercluster, and Emily Charry Tissier, CEO, Whale Seeker. 

Discussing AI’s transformative potential in agriculture technology, panelists shared AI-powered solutions that could revolutionize farming practices. Use cases highlighted AI’s potential in predicting pest outbreaks, optimizing water and resource usage, and making informed decisions to boost crop yields while promoting sustainability.

Finally, panelists discussed data-sharing challenges in industries like shipping and emphasized the importance of collaboration on climate change. 

“We have to find ways as a community to come together,” said Deval Pandya, Vector’s Vice President of AI Engineering. “Not just research or application, but as an international community of leaders together.”

Vector's VP of AI Engineering, Deval Pandya (seated on the far left), takes part in a panel on how AI is transforming ClimateTech.

During AI for the Environment, panelists talked about accelerating generative AI systems in applications for the environment. They also acknowledged that these AI systems often consume significant resources. There was a general consensus that better, standardized metrics to show environmental impact are needed to level-set AI’s environmental impact and work towards ways to minimize the effects.

David Rolnick, Assistant Professor in the School of Computer Science at McGill University, Canada CIFAR AI Chair, and a Core Academic Member at Mila, emphasized the many ways AI can contribute to climate action including:

  • gathering and analyzing large datasets for informed decision-making, 
  • improving the efficiency of automated systems, 
  • forecasting climate-related events, 
  • speeding up simulations, and 
  • accelerating scientific discovery. 

The panelists stressed that AI alone cannot solve climate change — collaboration across different fields is needed for tangible action. They called for greater transparency and standardized metrics to assess AI’s environmental impact and emphasized the need for responsibly adopting AI.

How AI is transforming the aerospace industry

Leaders in Canada’s aerospace industry gathered to discuss a future where AI doesn’t just assist but fundamentally reshapes aerospace. Facing the immense challenge of managing vast real-time data generated by aircraft, companies like Vector Gold sponsor Air Canada are leveraging AI to enhance on-time performance by analyzing data elements and creating diverse scenarios in real-time. McGill University’s John Gradek envisioned a future of coexistence between humans and AI in aerospace, highlighting its real-world competitive advantages. In sum, for these panelists, the future of this increasingly AI-driven space holds tremendous promise for Canada.

How AI is transforming the manufacturing sector

Diving into AI’s impact in the manufacturing sector, panelists shared case studies demonstrating AI-driven automation, predictive maintenance, and quality control. The focus also extended to smart factories and AI’s role in optimizing resource allocation, reducing downtime, and enhancing overall operational efficiency. Marie-Claude Cote, VP at IVADO Labs Data Science, highlighted Canada’s low AI maturity and slow AI adoption in businesses, particularly SMEs, and stressed the importance of simplifying the narrative around AI to drive adoption. Isabelle Hudon, President and CEO of BDC, echoed this point by raising the need for education to promote AI adoption, emphasizing the role of entrepreneurs and businesses in understanding the benefits of AI.

The panelists discussed the significance of having the right talent internally to handle AI projects and the challenges of scaling AI applications in established companies, touching on AI’s potential to address labour shortages in the manufacturing sector.

Their final comments underscored the critical role of data in AI applications and the need for experimentation and problem-solving, again emphasizing the need for education, talent, and adaptability to move the needle on AI implementation.

How AI is transforming the healthcare sector

The panel discussion on health revolved around how AI may be harnessed to help address the state of the Canadian healthcare system. Panelists highlighted the current inefficiencies within the system as well as the opportunities. They discussed the importance of building platforms that serve both patients and healthcare providers simultaneously, focusing on user experience. The role of AI was underscored in automating mundane tasks, allowing healthcare workers to concentrate on patient care.

Regarding technology adoption, panelists acknowledged that healthcare providers are open to technology as long as it enhances their interaction with patients. However, they recognized that healthcare systems at the national level are slower to adopt new technologies due to systemic factors, and a change in their tolerance for risk is needed for faster AI adoption.

In terms of the future, panelists emphasized the importance of building systems today that account for the automation of tasks in the near future, the need for AI trust and safety among healthcare professionals, and the importance of breaking down silos to efficiently leverage data.

Privacy, fairness and diversity in AI

Panelists looking at privacy, fairness, and diversity in AI highlighted the pressing need to infuse principles of AI trust and safety, privacy, diversity, and gender equality throughout the entire AI development cycle. They stressed the importance of creating regulations and standards to ensure the accountability of AI systems. Education and awareness were deemed essential not only for AI developers but also for the general public to understand the nuances of bias and fairness in AI.

Further, government involvement and international cooperation will be vital in addressing AI’s societal impact. The panelists advocated for a human rights framework to guide AI trust and safety development and deployment. In a world where AI systems wield substantial power, the importance of equitable access was repeatedly underscored.

A call to action from panelists went out to the people utilizing AI, emphasizing the importance of involving end-users, gathering diverse perspectives, and rigorous data collection to prevent AI models from perpetuating biases. 

A common theme across the conference was the need for practicality in implementing AI trust and safety guidelines, as well as the urgency of accelerating AI adoption, AI safety research, and the paramount importance of education, awareness, and collaboration towards responsibly harnessing AI for a promising future.

Related:

Private AI CEO Patricia Thaine presents at Collision 2024
News

ChainML, Private AI, and Geoffrey Hinton underscore the importance of responsible AI development and governance at Collision 2024

Vector Faculty Member Raquel Urtasun speaking on stage at Collision 2024 .
News

Self-driving trucks will be on the road next year says Vector co-founder Raquel Urtasun at Collision 2024

Research
Trustworthy AI

World-leading AI Trust and Safety Experts Publish Major Paper on Managing AI Risks in the journal Science