The Opportunity for Academia-Industry AI Collaboration

Where do you think Canada currently stands on artificial intelligence research, development and deployment?

A lot of directives coming from the federal level are trickling down to the provincial and municipal levels in terms of investing in AI and developing the right infrastructure for it. These strategies contribute to the development of the computing power and talent that we need, supporting a strong AI ecosystem.

I also see more opportunities in ensuring that we look into the ethical deployments of AI. Given the kind of work that we are seeing across Canada when it comes to machine learning, deep learning, and various facets of AI, there is obviously a significant concern around the ethics and the social consciousness of all of this. Canada can lead in terms of this. Across the country, we are beginning to see a lot research dedicated towards ethical AI and discussions around legal, social, and psychological implications. AI is a multifaceted and interdisciplinary practice, and I think we have a huge opportunity to be leaders in that interdisciplinarity.

With that said, our weaknesses seem to lie in the fact that we work in provincial pockets and I think more collaboration on the provincial level would be worthwhile.

I also see some weakness in our knowledge translation. We do great work, but I do not think the international dissemination of that work is quite at the level that it needs to be. 

“AI is a multifaceted and interdisciplinary practice, and I think we have a huge opportunity to be leaders in that interdisciplinarity.”

In order for Canada to get international recognition for our AI deployments and talents, we need to become more open and creative in how we talk about what we do. Canada needs to present itself as a country where we can take complex, often daunting AI concepts and break them down in such a way that anyone can say they understand it. We have some of the most brilliant minds in AI development but the conversations that we are having around AI need to become more mature in a way that bite-sized information is accessible and applicable to everyone.

Furthermore, lot of our AI developments are confined to urban locales. The vocabulary that we have developed around smart cities is something that I see as problematic because the concept and creation of smart cities essentially excludes rural towns. Instead, we should be looking at AI developments as a holistic smart community. We should be extending what we are doing in data science, machine learning, and neuro-linguistic programming to our rural communities because there will be a lot of impact there. Otherwise, talent will get localized in urban areas, which ends up pushing up housing prices. We have a lot of international and local students studying these disciplines and I think we should start exploring how we can create more opportunities for these students outside urban locales, though initiatives such as the Pilot Immigration Program.

What are the main obstacles SMEs face when adopting and integrating AI technology?

The first opportunity we have in Canada when it comes to SMEs and AI is education: we must help these small to medium businesses understand what AI is, how to produce something with AI, how to invest in AI, and how to diagnose its problems.

From the perspective of academia, we can also assist in the deployment of student and faculty talent, as well as support these businesses in research and development (R&D). The kind of investment that is going into AI R&D trickles on to SMEs from an applied research perspective. This research and investment needs to become more specialized and governed by an educational process and a decision making group that will be able to determine if a business is suitable for AI deployment. We will also need to address from the get-go what impact AI deployment is going to have on the company’s human workforce.

In terms of the actual implementation of AI in SMEs, while it is relatively easy to create AI prototypes and products thanks to applied research, the deployment and the sustaining of these prototypes or products is extremely difficult because you need to have somebody in-house to help, and there is a cost associated with that. 

“The first opportunity we have in Canada when it comes to SMEs and AI is education: we must help these small to medium businesses understand what AI is, how to produce something with AI, how to invest in AI, and how to diagnose its problems.”

Additionally, deploying an AI solution in-house requires mature, localized infrastructure, whether it is from a database or computing perspective, in order to ensure that AI will generate the value you need. Installing and maintaining this new infrastructure will require some handholding and consistent fuelling. This is another cost SMEs can be sensitive to.

Small to medium sized enterprises also tend to have more cohesive, tight, family-like teams. Introducing an artificial intelligence element can have an impact on this, and having that conversation well in advance of any implementation is paramount in ensuring that the community will be ready to embrace this new team member.

Unfortunately, I do believe that there is a sense that AI is an entirely technical subject. It does have a technicality to it, but 90% of the work that we are doing depends on subject matter experts from different sectors who can talk to us about what they want this tool to do and how they can continue to fund our intelligence.

What advice would you give SMEs on how best to implement and convert artificial intelligence developments into productivity and growth? 

I would advise SMEs to look into bodies and investments that already exist, such as MITACS, OCE, The Natural Sciences and Engineering Research Council of Canada (NSERC), The National Research Council-Industrial Research Assistance Program (NRC-IRAP) to name a few. These organizations can help SMEs tap into existing talent and significantly reduce the costs.

When implementing AI technology, I also think it is incredibly important for SMEs to establish from the get-go that this technology does not simply sit in the lap of the IT lead or CTO. This kind of product requires different experts in the organization and all kinds of input in order to be successful and sustainable. If you are going to go down this path of either integrating AI into an existing business process or developing something exclusively based on AI, you will need to remember that AI is like a living, breathing creature that is part of your team and needs to be fed and taken care of.

“One important aspect companies need to prepare for is having every part of their organization, including marketing, business development, engineering, and IT, collaborate on AI deployment.”

One important aspect companies need to prepare for is having every part of their organization, including marketing, business development, engineering, and IT, collaborate on AI deployment. Understanding that AI is a much broader product than what people think of it as is the first step in ensuring success.

What do you think of the collaboration between industry and academia on AI projects right now, and what more do you think could be done?

The structure of an academic institution gives companies access to research and development tools that they typically will not have access to in-house. You have faculty, students, and technical expertise all clubbed into one little team. Furthermore, through partnership with academic institutions, industry can have access to capital investment from our federal, provincial, and regional funding pools, which will reduce the overall costs of AI development. 

“There is a fantastic opportunity for academia and industry to liaise with each other directly and learn together on how to develop, deploy, and share AI prototypes.”

There is a fantastic opportunity for academia and industry to liaise with each other directly and learn together on how to develop, deploy, and share AI prototypes.This partnership only magnifies what both sides are trying to develop. The students who work on these projects are paid and will gain tremendous experience working in industry and with AI. For industry, this is a great opportunity to scout for talent.

What does the Canadian AI sector need to improve on in order to cement our status as world leaders on the subject?

In terms of funding, we need to go back to basics. Before we look at a fully-furnished house, we should look and see if the foundations and infrastructure are stable. Some of Canada’s AI funding needs to be sector-specific and look to solve specific problems, such as developing AI to reduce homelessness or youth crime. The directions on this can be driven by regional, provincial, and federal priorities.

“AI can make access to services easier and showcase how we can provide quality experiences for all Canadians regardless of their status.”

We have a chance to grow and build our Indigenous talent base through the opportunities that AI is giving us. Truth and reconciliation has a long way to go and what we are able to do with AI now can create bridges for us to engage further with talent in these communities. By learning from other worldviews, we will be able to create avenues to develop AI that is as close to being bias-free as possible. 

As Canadians, we also have this amazing opportunity through welcoming new immigrants into our communities. AI can meet the needs that emerge from this. AI can make access to services easier and showcase how we can provide quality experiences for all Canadians regardless of their status.These are areas where if we use AI correctly, we can go a long way.

“When it comes to AI, we do not just need to think outside of the box. We need to burn the entire box.”

We need people who are willing to take a gamble and trust the talent.Often, what I see in innovation projects is that we stay rooted in certain hierarchies. When it comes to AI, we do not just need to think outside of the box. We need to burn the entire box.