"The best place in North America to eat, and the best place in North America for AI"

Last Tuesday, McGill University hosted the second annual Montreal AI Symposium, which aims to gather “experts and professionals interested in fundamental advances and applications of artificial intelligence [...]” and strengthen Montreal’s AI ecosystem. The day-long event featured keynote addresses, contributed talks and posters, and ended with a sponsor networking cocktail.

While the presentations themselves were rather technical and specialized -- a clear testament to the calibre of AI talent in Montreal -- perhaps the most remarkable was the organizers’ concerted and comprehensive effort at promoting diversity and inclusion.

This was evident not only in the speakers chosen as keynotes -- both of whom were women at the top of the field -- but so too in the daycare provided, the code of conduct introduced, the demographic statistics they studied via registration as a way of tracking their efforts, and the audience selection process they used to encourage diversity.

I’ve previously written and reflected on other AI conferences where a primary topic of discussion has been the importance of conscious efforts of inclusion and diversity in artificial intelligence, not least as this technology becomes ever present in our day-to-day. But it was heartening to see how blatant these efforts were in all aspects of the symposium, made exponentially more so by the repeated evidence that Montreal, indeed, is the “AI capital of the world” (Mark Maclean, Montreal International, citing his conversations at ICML).

This means that Montreal will have the opportunity to lead efforts at making the AI-abundant world that our future holds a diverse and inclusive one. Still, with great power comes great responsibility; it isn’t enough to be hyper-aware of biases and injustices that exist, and how these are heightened and exacerbated through AI.

Think of the example presented by Professor Margaret Mitchell, in which she cited a well-known study in asking the audience: “a man and his son are in a terrible accident and are rushed to the hospital in critical care. The doctor looks at the boy and exclaims, “I can’t operate on this boy, he’s my son!” How is this possible?”

The answer is simple: the doctor is a woman, and she is the boy’s mother. If you didn’t reach this conclusion, don’t worry: you belong to the majority -- men, women, and self-described feminists -- who automatically associate the word “doctor” with male; for things that are “remarkable,” i.e. female doctors, we qualify them. No one ever says “male doctor.”

Of course, female doctors aren’t remarkable per se. But we come to associate certain images with certain words, using qualifiers to indicate what society has taught us is “remarkable.”

Think of another example, in which word frequency is meant to be indicative of the frequency of instance: the word “murder” is more frequently used than the word “exhale,” and comes up more often in searches, but does this mean that murder happens more frequently than exhaling? Of course not. We just talk more about things which are remarkable.

These examples are meant to highlight dangers and inaccuracies that can come of not making concerted efforts at diversity and inclusion, even when we consider ourselves to be well-intended and socially aware.

Two other critical points that came out of the panel discussion on the ecosystem of AI in Montreal, which featured Yoshua Bengio, Narjès Boufaden, Sylvain Carle, Mark Maclean, Joelle Pineau and chaired by Philippe Beaudoin, included: the importance for businesses to better advertise what exactly they’re looking for from the talent that they seek to hire such that local universities can best prepare their students for the tasks at hand; and, the importance of connecting the Montreal ecosystem to the international one.

Professor Martha Crago, VP of Research and Innovation at McGill, opened the symposium on a light note, saying Montreal is “the best place in North America to eat, and the best place in North America for AI,” and pointed out that as our taxes, in part, have gone to funding research, Montreal has been able to carve out its niche as a city blazing the trail for socially-responsible AI.

The road ahead is a long one, but Montreal has the talent needed to get us there (and the food needed to keep that talent well-fed).