Shani Gwin's foray into the world of artificial intelligence began with a problem. The Métis entrepreneur from Edmonton previously worked in government communications and found herself one of the only Indigenous people on her team. She was often tasked with reviewing Indigenous content for misinformation.
She noticed that Indigenous people frequently have to battle online hate speech and inform others on First Nation, Inuit and Métis history. It prompted Gwin to create a technology company called wasikan kisewatisiwin, which means “kind electricity” in Cree. It has developed an AI tool that can detect inaccurate representations of Indigenous Peoples in written content.
“A lot of our history online is from a colonized perspective. It’s not validated by us,” Gwin said in a recent interview. “I see this as an opportunity for us to share our truth, our history, that knowledge that we want to put in (the tool).” Gwin is among some Indigenous people who are exploring how AI can help organizations educate the public, preserve culture and language, and address environmental concerns.
With new AI agents popping up at lightning speed, Gwin’s journey took a slower pace. “Sometimes it’s a good thing to be late in the game, because we can actually do this right.” She reached out to the Alberta Machine Intelligence Institute in Edmonton to determine if an AI system could be developed to recognize harmful or misinformed comments against Indigenous Peoples. From there, Gwin worked with Indigenous elders and community members to construct and train an in-house AI model to detect domain-specific content that could be seen as spreading misinformation or racism.
The tool can be used in two ways: with companies’ internal AI systems or as an add-on, similar to writing assistance software such as Grammarly. What it won’t do, said Gwin, is give away traditional teachings. “The goal is never to give away all of our knowledge for free,” she said.
“But it is to make that reconciliation and truth aspect more top of mind for people in their day-to-day and not an afterthought.” Research shows there are limitations on cultural accuracy with generative AI, because of a lack of available information online. Opinions on the role AI can have in preserving language and culture vary from nation to nation. But there are shared concerns over data sovereignty, environmental effects from large dat…
Summary from source