Structuring Artificial Intelligence: A Network Analysis of AI Development in Canada and France

“Geoffrey & Yann & Yoshua” is a label that can be seen on a fashionable new T-shirt. For those who get the joke, it refers to the rise of professors Hinton, LeCun and Bengio as iconic figures of the Artificial Intelligence community, a.k.a the “Deep Learning Conspiracy,” a.k.a the “Canadian Mafia.” The question we want to answer here relates to the conditions that have allowed this particular configuration of this particular field, somehow outside of the US. First, we seek to archeologically reconstruct the scientific ties among the actors (for instance, the fact that LeCun was Hinton’s post-doctoral student in Toronto), as well as the network of institutional support provided in the 1990s-2000s, especially the critical contributions from CIFAR (Canadian Institute for Advanced Research). Second, and in line with the “triple helix” of universities intertwined with governments and corporations, we want to focus on the role of the latter by examining, for instance, how instrumental the Google-Hinton relationship has been in the sub-field of image recognition, or how LeCun’s association with Facebook has greatly facilitated the emergence of Paris as a new AI powerhouse (FAIR Centre, Station F, Collège de France, etc.). Third, it is important to note that none of the above developments would be possible without cultural cues in the form of particular beliefs, hopes, and narratives. Specifically, we want to discuss the symbolic handling of the disruptive aspects of AI by major Canadian and French actors (research institutions, business, and government)—how, for example, the “Silicon Valley of the north” narrative is promising to create rather than threaten jobs in the Montreal Region—, and what is at stake in such a framing of the debate with regards to the ongoing overall structuration of AI.