Many connoisseurs, and not the least, including the big financiers of the development of artificial intelligence (AI) are worried about its rapid advances to the point of asking for a moratorium on research in the field. That’s good, the International Observatory on the Societal Impacts of Artificial Intelligence and Digital Technology (OBVIA) was created in Quebec to analyze and criticize this new world.
“Our mission is geared towards responsible innovation and the common good,” explains Lyse Langlois, CEO of OBVIA, while acknowledging that her organization remains very little known, perhaps because it was developed during the pandemic. . “From the outset, we announced that we were interested in the social issues of technological innovations by proposing innovations in open science and by placing the human being at the center of concerns as well as the critical perspectives in the face of these developments. »
OBVIA was born after a competition launched by the chief scientist of Quebec, Rémi Quirion, in 2018 with an initial budget of $7.5 million. Université Laval won the race with its project linking nine universities and nine CEGEPs from the start, already bringing together 170 researchers. They are now about 240.
The structure links together scholars from the engineering, health and social sciences who share a concern for societal impacts. The acronym OBVIA is reminiscent of the beautiful word “obvier” which means “to take the necessary steps to obstruct something, to ward off possible harm”. Mme Langlois herself comes from the social sciences with a specialization in organizational ethics.
In four years, the Observatory has funded 22 teams to the tune of 5.5 million and 36 scholarship holders for 1.2 million, from middle school to post-doctorate, again following competitions. For example, research has focused on the governance of health data, the use of facial recognition technology by the police or the impact of AI on the professions.
OBVIA also supports an international chair linked to the École Normale Supérieure in Paris on social justice and AI. Funding is for one year at a time. The first incumbent focused on marginalized populations, including Indigenous peoples. The current beneficiary works on the sharing of health data. The Observatory also organizes conferences, of course. The most recent dealt with ethico-legal norms.
And ChatGPT?
And on conversational robots, the scary or captivating subject of the hour? OBVIA preferred to wait before getting involved in the debates. An “education” group is working on a report which will propose guidelines for use in universities this spring. Lawyers, ethicists and specialists in health as well as in culture and media have already identified issues in their various fields. Another group will speak on behalf of CEGEPs at the next ACFAS conference.
“In the United States, there is a big tendency to ban ChatGPT in universities,” says the director. Red lights come on. Innovations are going at a very surprising speed with the possibility of impact in many areas. The media are directly affected by the creation and dissemination of false information. Our democratic mechanisms will also be affected. »
Quebec is home to other centers on the cutting edge of these digital tools that fascinate and worry. The best known, the Montreal institute MILA, brings together hundreds of researchers in the field of deep learning. Its director, Yoshua Bengio, signatory of the recent call for a moratorium, is also a member of OBVIA, whose research themes for the initial phase around responsible innovation are in line with the Montreal Declaration for responsible AI. Everything fits.
Red lights come on. Innovations are going at a very surprising speed with the possibility of impact in many areas.
The chief scientist will receive the results of the first four years of activity of OBVIA this month, in April. The next phase will focus on society, dialogue with citizens, popular education in AI too.
“It’s a shame that we are less well known at home,” said Mme Langlois, interviewed just before the outbreak of the recent professors’ strike at Université Laval. She came from a strategic mission at the University Côte d’Azur, in Nice, where there is one of the three Franco-French centers of research and development around AI with Toulouse and Paris.
“I say this objectively, France lacks an observatory like ours to look at AI work in an interdisciplinary and ethical way. In fact, I’ve been around the world looking for an OBVIA-like structure, and there isn’t one. In a recent German policy, the desire is clearly expressed to create an observatory like in Quebec. »