Should we be happy or worried about the very rapid advances in artificial intelligence (AI) research and its adoption in a multitude of spheres of our lives? We spoke little about it six months ago, even if artificial intelligence was already everywhere without us always realizing it. Since then, everyone and his cousin has been chatting with ChatGPT and generating fake images with DALL-E.
Mixed success: great successes, impressive sham creativity, but also glaring errors produced with as much aplomb as speed. The picture is emerging more clearly: while artificial intelligence can serve the common good and accelerate scientific discovery, it can also feed misinformation, fuel conflict and deepen inequalities.
The greatest experts are now calling for a pause in the development of the most powerful generative artificial intelligence systems, while we find a way to control their territory and power. Whether this proposal is followed or not, the lively debate it has generated allows us to highlight certain important elements of the research that is carried out in universities.
The AI ​​ecosystem is particularly developed in Montreal, where unparalleled expertise in the field and numerous partnerships with the private sector, supported by public funds, are concentrated. Therefore, universities, and mine in the first place, bear a fundamental responsibility to foster, if not to drive, the development of an AI that meets the expectations of humanity, not only in terms of progress and productivity, but also of collective well-being, justice and equity.
I draw three conclusions.
They are direct contributors to the digital transformation of our societies and must invest in the field of AI for digital innovation to truly serve the common good. As we can see, the ambient and legitimate anxiety around the acceleration of AI discoveries illustrates the need to also invest in disciplines that make it possible to understand, contextualize and frame the evolution of knowledge: anthropology, law, ethics, history, literature, philosophy, sociology. I often hear that Quebec needs more engineers and computer scientists. It’s true. But Quebec and the rest of the world also have a crying need for the contribution of the human and social sciences.
Second, universities must live up to the capital of trust they enjoy in order to adequately mark scientific advances in the field of artificial intelligence and promote the adoption of research practices and orientations based on shared values.
The University of Montreal resolutely embarked on this path a few years ago, bringing together the actors and actresses of this sector, with experts in the humanities and social sciences, organizations and members of the public around ethical principles. brought together in the Montreal Declaration on Responsible Artificial Intelligence1. We must now go even further and bring these principles to life.
Finally, universities must review their training programs in order to prepare the next generations for the digital environment that is being built today. The primary and historical mission of universities remains the transmission of knowledge and the full deployment of human potential, but it will quickly have to be adapted to a radically different context. The mastery of digital innovations and AI will not be achieved without the development of critical thinking, autonomy and intellectual agility.
Conversely, programs in the humanities and social sciences will also have to promote the digital and scientific literacy of their students in urban planning, communications or social work, for example.
All this will have to be done in a network, on a planetary scale, because scientific advances in artificial intelligence know no borders. At the University of Montreal, we work with some forty leading universities around the world, to share our expertise on responsible AI, within the U7 group.
Under the leadership of Professor Catherine Régis, from the University of Montreal, the group published a report in 2021 which already anticipated questions that everyone is asking today, and draws up the roadmap for the responsible development of AI. Together, we have the power to create an artificial intelligence that will truly serve the common good and that will advance our collective intelligence.
