5 min readThanks to a dynamic ecosystem of companies, researchers and entrepreneurs, Montreal has become [...]
Thanks to a dynamic ecosystem of companies, researchers and entrepreneurs, Montreal has become one of the world’s largest artificial intelligence hotbeds. The opening of the new Montreal Institute for Learning Algorithms – or Mila – was a critical step to reinforce the Quebec metropolis’ position as an AI leader on the global stage.
As part of the Espace CPDQ | Axe IA – a program whose goal is the growth of tech companies that specialize in AI – Mnubo is fortunate enough to be able to work at Mila and to benefit directly from this booming ecosystem. In this article, we are taking an in-depth look at Mila and its importance for both Montreal and Mnubo.
Led by Yoshua Bengio, a Montreal-based AI researcher and pioneer of deep learning, the Montreal Institute for Learning Algorithms – commonly known as Mila – inaugurated its 90,000 square foot facility in late January. The goal is to bridge the gap between fundamental research and technology commercialization. This will allow us to compete on the international stage.
The innovation race is taking place before our eyes. Countries like China are investing heavily in artificial intelligence. Indeed, in 2017, Chinese investments accounted for 48% of the world’s total AI startup funding. But startups are only one piece of the puzzle. A study from the Boston Consulting Group revealed that 85% of all Chinese companies were active players in the field of AI. That’s why it is so important to centralize local research to stay at the cutting-edge of technology.
Mila is all about synergy – between the deep learning and machine learning researchers, of course, but also between the growing number of AI tech companies in Montreal. In practice, it means that over 20 professors from McGill University and the University of Montreal – all AI specialists – will work alongside hundreds of researchers and several R&D teams from carefully selected technology companies across Montreal.
Mila will serve as a platform for research in the areas of deep learning and reinforcement learning. It will host corporate labs and startups, but it is excellence in fundamental research that will set Mila apart.
The research on AI that is conducted at Mila is very important. But it is the sustained impact on Montreal’s technology ecosystem that will be under scrutiny. How can the institute benefit as many people as possible?
One of the initiatives taken to help companies take advantage of the “AI wave” is the partnership between the Caisse de Dépôt et Placement du Québec and Mila through its Espace CDPQ | Axe IA. In its initial selection, nine companies were chosen to join the institute. Mnubo was fortunate enough to be part of the chosen few.
The selecting committee used specific criteria, notably:
By backing a number of tech startups in Quebec, Mila helps strengthen the local AI ecosystem. This program will also lead to the creation of numerous jobs in Montreal. In other words, it is a way to monetize cooperation and to add value to it.
Mila, as part of the Espace CDPQ | Axe IA program, selected Mnubo. Our AI and IoT analytics solution will not only benefit from the available resources, but will tap into the collective knowledge and experiences of 350+ AI researchers.
For one of our data scientists, Nathanael Weill, “Mila is a great initiative to train and retain talent by giving students and companies access to one of the best AI centers while also helping with the development of technologies in applied fields.”
According to Nathanael, the Espace CDPQ | Axe IA is “a unique opportunity for Mnubo [and the selected companies], who will have exclusive access to a pool of researchers and students to help develop internal AI solutions.” It is also a great chance to “create a collaborative inter-company workspace where common obstacles can be addressed with support from the field’s best experts.”
From a personal standpoint, Nathanael is taking advantage of this environment to “apply the latest deep learning technologies – from recurrent neural networks to reinforcement learning – to real-life problems we see with our customers.” His next goal is to “make those approaches more systematic in order to improve the performance of our algorithms and lower the development time.”