Quebec’s national library, Bibliothèque et Archives nationales du Québec (BAnQ), is embarking on an ambitious project to develop a comprehensive database aimed at enriching the cultural and governmental content available for artificial intelligence systems. This initiative follows a recent feasibility study and seeks to bolster AI’s comprehension of Quebec’s diverse society, cultural nuances, and Indigenous languages.
Addressing the AI Knowledge Gap
The increasing reliance on AI tools has highlighted a significant gap in their ability to understand and accurately represent Quebec’s unique identity. Major generative AI systems often struggle to produce reliable outputs regarding local culture, economy, and languages due to a scarcity of region-specific data. Valérie D’Amour, who spearheaded the feasibility study, expressed optimism about the project’s potential, stating, “All scenarios are a little bit on the table right now. We have a lot of ideas and we want to validate the possibilities with cultural stakeholders, as well as with data owners and providers, who will be involved in the discussions.”
The aim of the new platform is not to serve as a distribution channel for creative works but to ensure rigorous control over data access.
A Commitment to Cultural Representation
Marie Grégoire, the president and CEO of BAnQ, emphasized the necessity for AI systems to embody the essence of Quebec’s society and culture. She noted, “That means having Quebec references, whether in small models or large models, whether they come from research or from the business community.” This endeavour mirrors similar efforts globally, such as in Sweden, where vast collections of Nordic texts have been curated to enhance AI understanding of Scandinavian languages.
BAnQ plans to initiate the project using its own collections before potentially integrating data from external sources. This initiative is rooted in a recommendation from Quebec’s innovation council, whose recent report underscored the dearth of Quebec-related data in AI training datasets.
Tackling Cultural Biases
Destiny Tchéhouali, a co-holder of a research chair dedicated to French-language artificial intelligence at Université du Québec à Montréal, highlighted the pressing need for local representation in AI. He warned that the underrepresentation of Quebec culture in existing AI datasets could perpetuate linguistic and cultural biases, particularly concerning Indigenous peoples. Tchéhouali described the proposed database as “strategic infrastructure” that could shape how local content is catalogued and utilised within contemporary AI systems.
Despite the potential benefits, concerns regarding copyright and fair compensation for creators loom large. Grégoire acknowledged these issues, stating that the current landscape resembles “the Wild West,” where data is often harvested without proper remuneration for artists.
A Path Towards Sustainability
The development of this database could transform the relationship between creators and AI. Grégoire argued that by establishing a centralised platform, BAnQ could facilitate better compensation for artists whose works are incorporated into AI training. “By working collectively, cultural organisations would be better positioned to ensure creators are paid and that the sector remains sustainable over the long term,” she explained.
However, some creators express apprehension that their contributions could ultimately undermine their livelihoods. Maxime Harvey, a postdoctoral researcher at the National Institute of Scientific Research, noted that many artists worry about “feeding the beast that will eventually be used to replace contracts they may lose because of AI.”
The feasibility study envisions the platform becoming operational by 2029, although D’Amour indicated that this timeline would be reassessed post-experimental phase. The initiative is expected to require a budget of approximately $10.5 million over the next five years, with the Quebec government contributing $340,000 for the feasibility study and an additional $750,000 for the project’s initial experimentation phase.
Why it Matters
This initiative at BAnQ represents a crucial step towards ensuring that Quebec’s rich cultural tapestry is adequately represented and preserved in the rapidly evolving landscape of artificial intelligence. By addressing the data deficit and the potential biases in AI systems, this project not only aims to enhance the accuracy of AI outputs related to Quebec but also seeks to empower local creators, ensuring that their voices are heard and valued in an increasingly digital world. The implications for cultural sustainability and representation in AI are profound, marking a pivotal moment for the intersection of technology and local heritage.
