Posted by NewAdmin on 2025-01-16 12:23:19 |
Share: Facebook | Twitter | Whatsapp | Linkedin Visits: 82
How AI is Helping Indigenous Communities Preserve Culture and Language
Indigenous languages are vanishing at an alarming rate, but AI offers hope for their preservation, say Indigenous technologists.
Indigenous researchers face a race against time: Of the 4,000 Indigenous languages worldwide, one goes extinct every two weeks as its last speaker passes away. “Within five to ten years, most Native American languages in the U.S. could disappear,” warns Michael Running Wolf, founder of Indigenous in AI, a global network of Native, Aboriginal, and First Nations engineers.
Running Wolf has dedicated his career to combating this crisis. He leads First Languages AI Reality, a project by the Mila-Quebec Artificial Intelligence Institute, which is developing speech recognition models for over 200 endangered Indigenous languages in North America.
However, a critical challenge persists: there are too few Indigenous computer science graduates who possess both the technical expertise and cultural knowledge needed for these projects. Running Wolf stresses the importance of respecting the data, which often includes deeply personal and culturally significant information from speakers who have passed away. “The data isn’t just social media posts—it’s the essence of communities’ identities,” he explains. “We must ensure communities maintain control over their data.”
Despite years of research, Running Wolf estimates there are only about a dozen Indigenous North American AI scientists. “Only one or two Indigenous Ph. D.s in AI and computer science graduate each year,” he notes. Indigenous representation in tech remains alarmingly low: they make up less than 0.005% of the U.S. tech workforce, hold just 0.4% of computer science bachelor’s degrees annually, and have only one board member among the top 200 tech companies. In 2022, Native-founded companies received a mere 0.02% of venture capital funding.
Still, the few Indigenous engineers in the field are making significant strides. Organisations like First Languages AI Reality, IndigiGenius, Tech Natives, and the Wihanble S’a Center for Indigenous AI are training Native American, Alaska Native, and Native Hawaiian students to use technology to preserve their languages and cultures. “Traditional AI frameworks often treat data as proprietary, which can harm Indigenous communities,” says Running Wolf. “Our goal is to demonstrate that language preservation can be achieved ethically using culturally appropriate AI protocols.”
One success story is Kyra Kaya, a computer science student at Yale University and a participant in the Tech Natives program. Inspired by her Native Hawaiian grandmother, Kaya developed an AI tool to recognise spoken phrases in Hawaiian Pidgin English, a stigmatised creole language spoken in Hawaii. “I wanted to challenge the idea that Pidgin is a ‘lesser language,’” Kaya says. By training the tool on phrases her grandmother and relatives used, she hopes to create an app that will make the language accessible to Hawaii residents. “AI has the potential to either amplify or silence marginalised voices like ours,” she adds. “That’s why Indigenous representation in tech is so critical.”
Encouraging interest in tech at an early age is another key strategy. IndigiGenius’ Lakota AI Code Camp, held annually in South Dakota, immerses Native teens in a three-week program to develop apps that document Lakota culture, including sacred plants and common words. Since its launch in 2022, the camp has trained 33 students, many of whom have returned as mentors or pursued further computer science studies. To expand year-round access, IndigiGenius also launched T3PD, a program training high school teachers to integrate culturally relevant computer science lessons into their curricula, bringing technology to classrooms that previously lacked such resources.
“Making AI education culturally relevant is essential,” says Andrea Delgado-Olson, Executive Director of IndigiGenius. “We’re using Indigenous knowledge to connect technology with traditions.”
AI is also bridging gaps in other areas of Indigenous heritage. Madeline Gupta, a Yale student with Chippewa ancestry, is developing a virtual reality experience for Native youth to explore their ancestral lands. After the U.S. government displaced tribal members from their lands between 1819 and 1969, many lost their connection to their heritage. “Of the 50,000 people in my tribe, only 2,000 live on the reservation,” Gupta says. She hopes her VR project, which incorporates stories and 3D spatial videos of Mackinac Island, will help youth feel a deeper connection to their culture.
AI is also being used creatively by Indigenous artists. Suzanne Kite, an artist at Bard College’s Wihanble S’a Center for Indigenous AI, integrates AI into her work to explore Lakota dream language. Kite used machine learning to translate dreams into Lakota women’s geometric patterns used in bead work and quilts, as well as a graphic score for an orchestra. “Rather than adopting Western views of AI, I draw from Indigenous knowledge systems to create ethical art,” she explains.
While these projects are in their infancy, Running Wolf envisions a future where they’re no longer needed. “My dream is to revive endangered languages and empower new generations of Native speakers to create ethical technology,” he says. “One day, this technology will serve as a testament to a difficult era, reminding us of what was saved.”