THE FIRST RESULTS OF THE MONTREAL RESPONSIBLE AI DECLARATION CITIZEN DELIBERATIONS
On June 14, the Université de Montréal team revealed the summary of the citizen deliberations around the Montreal Responsible AI Declaration, at Virage on the MIL campus. Dozens of citizens, experts and stakeholders gathered to find out the first results of the consultations, which were held between February and April 2018, and which allowed over 500 people from all horizons to debate about the responsible development of artificial intelligence (AI).
Marie-Josée Hébert, Vice-Rector of Research at UdeM, Louise Béliveau, Vice-Rector of Student Affairs at UdeM and Yoshua Bengio, artificial intelligence expert, all opened the activity by positioning the work on AI ethics as indissociable from fundamental research. The reflection on the social impacts of this technological revolution and the recommendations that make up the framework for its development must be debated and formulated by academic experts in fields affected by AI, industry and community professionals, and citizens. And this is what the Montreal Responsible AI Declaration has been working on for months.
“Artificial intelligence is purely algorithmic, and thus instrumental,” claimed Marc-Antoine Dilhac, Scientific Co-director of the co-construction and professor of philosophy at UdeM. “Human intelligence must therefore provide a framework for its development, so that this tool can be used for the collective greater good.”
Christophe Abrassart, Scientific Co-director of the co-construction and professor of design at UdeM, then presented the method that enabled the discussions: the use of prospective AI scenarios set in Quebec in 2025. Fictions that relied on 7 values, defined in the preliminary version of the Declaration: privacy, autonomy, knowledge, well-being, justice, democracy and responsibility.
What the citizens said. Generally speaking, the participants recognized that the arrival of AI came with important potential benefits. However, it was also mentioned that AI development had to be done with caution and immediately to prevent abuse, although some consider the possibilities brought on by AI to still be limited. As for potential solutions, all the co-construction tables agreed on 3 potential solutions to guarantee socially responsible AI development, regardless of the field: 1) Legal provisions; 2) Training offered to all and 3) The identification of key independent players for AI management.
We will present public policy recommendations around priority fields of action. To date, we can say that three fields of action have established themselves: digital literacy, diversity and inclusion, and transition and social mutations. The final results will be presented in December 2018.
The Declaration co-construction was launched on November 3, 2017, at the Palais des congrès de Montréal during the Responsible AI Forum. An entire year wouldn’t suffice to cover all the themes of the reflection. The co-construction initiative will therefore continue in September 2018, allowing for discussions about new themes such as the environment, democracy and media propaganda, as well as security and integrity.
Read the entire summary of our citizen deliberations on our website: www.declarationmontreal-iaresponsable.com/bilan-deliberations-juin-2018