Last week’s announcement of the sale of Element AI to American firm ServiceNow and the federal government’s decision this summer not to use the COVID-19 tracing application designed by Mila, the Quebec Artificial Intelligence Institute, cast a shadow of the world’s leading expertise. developed by Montreal in the field of artificial intelligence. Valérie Pisano, CEO of Mila, analyzes the situation and defends the advances AI has made in Montreal over the past two years.
Posted on December 9 2020
Element AI may not have been Quebec’s flagship, but it is still a flagship company in Montreal artificial intelligence with its co-founder Yoshua Bengio, who is also Mila’s founder and scientific director. Will the sale of this company damage the credibility of the AI cluster in Montreal?
I do not think so. This transaction with a large operating company like SerivceNow instead proves the relevance of building an artificial intelligence hub in Montreal. This transaction confirms the importance of Montreal’s ecosystem and the talents working there.
But doesn’t the commercial failure of Element AI that led to its sale cast a shadow on our ability to build profitable AI businesses on our own?
Certainly Element AI may have experienced marketing challenges, but the company has also succeeded in developing innovative solutions in the field of AI, and that is why a company has become interested in the sector. . The company will continue its growth in Montreal with new capital and new partners, this is not the first startup where this has happened.
But this transaction should not be forgotten that we have several other AI companies that continue to thrive very well in Montreal and Quebec, such as Coveo, Imagia, StradigiAI, Lightspeed, Hopper …
Mila made an application last spring to combat COVID-19 which is not maintained by the federal government. Did you consider this rejection a failure?
Not allowed. Mila researchers have developed a risk prediction tool whose algorithms have made it possible to prevent exposure to the disease. A unique epidemiological model was developed that resulted in the publication of two scientific studies. We will soon publish the two simulation codes we have developed.
The problem is that we have to send all the data collected to a central server, and some have expressed fears about the protection of users ’personal data even if we don’t collect any data such as name or IP address. In other words, the government prefers the Shopify application, which does not use artificial intelligence.
That said, the Quebec Artificial Intelligence Institute, Mila, was created to attract and retain as many AI skills as possible in Montreal, and you’ve been operating for two years now. What has been your assessment over the past two years?
We started with approximately twenty associate professors at HEC Montreal, University of Montreal, McGill and Polytechnique and about 200 students. There are now more than 50 professor-researchers and more than 600 researcher-students.
Our four universities recruit professor-researchers in AI and Mila supplies research chairs in their respective fields. We conducted our activities with researcher-students-master’s, doctorate and post-doctorate-who mainly came from here, but now half of our students are from all over the world, mainly from India, China, from France and the United States.
Are there sectors of activity where Mila’s expertise is more in demand than others?
Our researchers are active in all sectors of activity. In the field of health, for example, we have teams that develop algorithms to speed up the manufacture of molecules. Many large companies have also partnered with Mila to gain access to our talents, such as Novartis or Roche, which recently opened an AI research laboratory in Montreal and which are working with us. We are a center of scientific research and our work is done in open mode.
Do you believe that Mila made it possible to strengthen Montreal’s position as an important hub for AI in the world?
The federal government commissioned a study from Accenture to measure the impact of creating three Canadian AI labs, and this report told us three weeks ago that the University of Montreal is now leading the way. world for AI research.
Another Montreal International report showed that Montreal was ranked first among the cities in North America for investment in AI and direct investment in technologies increased by 50% between 2017 and 2019.
You also have a mandate to make artificial intelligence more accessible to local businesses. How do you handle this aspect of your mission?
This is an important part of our activities. There’s a lot of appetite in our companies for AI, but there’s also a digital space to fill a lot of society.
We still have more than 50 partner companies such as Hydro-Québec, with whom we have developed algorithms that will make it possible to predict solar radiance to better manage the entire network.
These large companies, such as CN, do business with our team of applied researchers with approximately ten projects underway.
We also have a partnership with Caisse de dépôt, which has access to our technological resources and established on site with approximately twenty start-ups in which it has invested. We also have a team of researchers who oversee our own start-ups, the ones set by our student researchers.
How do you see the future of Mila and AI in Montreal?
We ourselves are a start-up. We quickly settled in, two years ago, we are in the process of operationalization, and there, we want to accelerate our influence. We have a responsibility to participate in public discussion to better understand how artificial intelligence is integrated into life and we must also make progress in AI more seamless for our companies.