A new program is reshaping the dialogue regarding the future of artificial intelligence (AI). It raises concerns that the existing path of AI development displays a bias against non-Western concepts of intelligence, particularly those rooted in Indigenous cultures. Abundant Intelligences is a global, multi-faceted initiative that aims to redefine our understanding of AI. Its core idea revolves around integrating Indigenous knowledge systems to craft a comprehensive and inclusive definition of intelligence and intelligent actions, ensuring these concepts are part of both current and upcoming technologies.
A new initiative led by researchers at Concordia is prompting a re-evaluation of the direction of artificial intelligence (AI). It argues that the current trajectory is fundamentally biased against non-Western perspectives on intelligence, particularly those emerging from Indigenous cultures.
In an effort to decolonize the future of AI, they have established the Abundant Intelligences research program, which functions on an international, multi-institutional, and interdisciplinary level to transform how we think about AI. The central aim is to integrate Indigenous knowledge systems to forge an inclusive and robust understanding of intelligence and intelligent actions, and how these ideas can be woven into both present and future technologies.
This complete concept is elaborated on in a recent article published in the journal AI & Society.
“Artificial intelligence has adopted ideas from previous notions of intelligence that followed certain colonial paths, focusing primarily on industrial or production-related concepts,” explains Ceyda Yolgörmez, a postdoctoral fellow involved with Abundant Intelligences and a co-author of the paper.
The authors argue that this scarcity-driven mindset has led to the exploitation of resources and the erasure of Indigenous identities, an impact that affects contemporary discussions surrounding AI, adds lead author Jason Edward Lewis, a professor in the Department of Design and Computation Arts and holder of the University Research Chair in Computational Media and the Indigenous Future Imaginary.
“The Abundant Intelligences research program aims to dismantle the scarcity mindset and create space for various types of intelligence and methods of contemplating intelligence.”
The researchers are convinced that this different approach can lead to AI systems that prioritize human flourishing, support Indigenous languages, tackle urgent environmental and sustainability challenges, and offer innovative public health solutions, among other benefits.
Leveraging local intelligence
The community-driven research program is headquartered at Concordia in Montreal, but much of the localized research will be conducted by specific clusters known as pods across Canada, the United States, and New Zealand.
The pods will be based in Indigenous-focused research and media labs at Western University in Ontario, the University of Lethbridge in Alberta, the University of Hawai’i – West Oahu, Bard College in New York, and Massey University in New Zealand.
These collaborations unite Indigenous knowledge holders, cultural practitioners, language custodians, educational institutions, and community organizations with researchers, engineers, artists, and social scientists to create new computational methodologies that align with an Indigenous-centered viewpoint.
Additionally, the researchers are collaborating with AI experts and industry innovators, believing this program will unlock fresh research opportunities and raise new questions for traditional AI studies.
“For instance, how do you design a robust system using a limited amount of data, like various Indigenous languages?” inquires Yolgörmez. “How can you develop multi-agent systems that effectively integrate and uphold non-human entities within a unified framework?”
Lewis emphasizes that their methodology serves as a complementary and alternative stance to mainstream AI research, especially when considering data sets such as Indigenous languages that are vastly smaller than those currently utilized by industry front-runners.
“Our commitment lies in ethically using data from Indigenous communities rather than just pulling data from the internet,” he states. “While this results in minimal data compared to what larger corporations use, it presents opportunities for innovative strategies when working with smaller languages, which can be advantageous for researchers interested in exploring alternatives to conventional methods.”
“This exemplifies one of the strengths of a decolonial perspective: it allows us to move away from the narrow belief that only one method of action exists.”
