The New AI Frontier: Tagging Indigenous Values
It’s not ok for AI to tell us how we think — if it doesn’t even know what we think.
We spent last week with Native American students, scientists, educators, and technologists talking, tagging, and telling stories in Phoenix, Arizona. IVOW was invited to the meeting of AISES, the American Indian Science and Engineering Society, to demonstrate how we can teach artificial intelligence about our heritage through storytelling.
I led the workshop together with Tracy Monteith, Senior Software Engineer at Microsoft, a member of the Eastern Band Cherokee and life-long member of AISES, and technologist Chamisa Edmo of the Navajo Nation.
The larger purpose of the workshop was to share stories from American Indian perspectives in a way that can be leveraged to create diverse training data that begins to represent the viewpoints of indigenous peoples. Our message: Let’s not leave indigenous values behind. It’s not enough for AI to tell us how we think if it doesn’t even know what we think.
“Indigenous people are traditionally experiential learners. Generally, we grow up in ways that reinforce observation and exploration first, followed by a thoughtful discussion with someone older. That’s how we approached the AISES workshop,” says Chamisa. “Usually we find ourselves embodying a dominant cultural approach to things and not leveraging our own depth of expertise. This time with Davar and IVOW leading it was different.”
The notion that we can inform machines about indigenous knowledge and help preserve wisdom for future generations was simply mesmerizing for all, even for the young and older engineers in the room!
Cultural Inclusivity in AI
With over 570 tribes in the United States, from every territory in the Americas, there is a vast array of perspectives and opinions. We included a photo-tagging exercise in the workshop that used keywords that are far-ranging, diverse, and unique across groups.
We asked the audience to form small teams and gave them cultural images to review and describe. The task was to write a caption and assign seven tags for each image. The students were engaged and passionate. When it came to the particular image below they captioned it “ceremonial preparation” and chose these tags: plants, sage, cedar and sweetgrass, ceremony, prayers, medicine, sacred, cedar, blessing, protection, spiritual.
If you put that same photo into a popular image recognition app, it will tell you the image is more than 94% a plant, ice cream, and a dessert.
Another image featured pupils at the Carlisle Indian Industrial School in Pennsylvania. Founded in 1879 by the US government, The Carlisle School was part of a larger federally-funded effort to ethnically cleanse and culturally assimilate Native Americans.
Our student participants expressed that no robot or automation tool would be able to sense the sentiment that arises from such an image. The tags the students chose for this image included sadness, tragedy, cultural elimination, whitewash, genocide, Native children, but also resiliency.
The image recognition app tags person, human, audience, and crowd but also, ironically, smile as the top identifiers of this image.
The highlight of the workshop was when Steve Daren, a Diné (Navajo) and one the AISES Council of Elders, joined our session towards the end in a way that was unstructured and unplanned, but organic.
“The energy of our discussion about the children at Carlisle school was still lingering, which is why, I presume, Steve Darden walked up to bring us back into focus,” says Chamisa. “Elders are really good at knowing just the right moment. While the standard room of scientists, mathematicians, engineers, and computer scientists would have probably left after a few minutes, misunderstanding the need and immense knowledge this gentleman carries, this room sat patiently and listened — just as we’ve all been taught our entire lives.”
Diné Elder Steve Darden, who is also on the Navajo Nation Human Rights Commission representing business, spoke about his experience meeting people years ago, the stories they shared with him and the impact those stories had on his family. He talked about the owl that kept watch all day and all night and how many of the people in the room were taught to regard the owl with a sense of cautious reverence. But Steve Darden also said this is a moment to be patient with your Elders to come together to understand the promise of such new technologies.
While the standard STEM professional/student may not have made a direct correlation in this discussion, this room we were in understood exactly what he was saying. That’s the power of storytelling in our world and the complexity of our indigenous intelligence.
In this way, the urgency of western culture seemingly yielded to the ancient ways of the wise. As he spoke, you could feel the reverence and respect in the room. “When our elders speak, we are taught from the earliest days of our life that you are reverent and patient until the elder finishes,” Tracy points out.
I was suddenly weighing elder protocol, guest protocol, temporal protocol, western standards, agenda, personal needs, and teenage attention spans in this seeming “interruption,” but immediately realized the beauty in this one dramatic, suspenseful, elongated moment of our collective lives. Time and tech seemingly tied together like sweetgrass to form a stronger sense of purpose for us all.
In the concluding exercise, we held up a large Fijian tapestry and asked students to imagine their tribe’s cultural engine within the tapestry. What data would they need to gather and what ontology? “It was an excellent thought experiment,” says Tracy. “And also a reminder that despite the diversity, there is a shared lived-life experience that binds us toward a common understanding of the human experience.”
Despite the fact that Artificial Intelligence is informing so many aspects of our lives, it is clear that AI tools are missing cultural data and context, and by extension the traditions of millions of underrepresented communities are not being recognized.
As part of the closing ceremony of AISES, I attended a remarkable social powwow. The ceremony began with a processional led by American Indian veterans carrying the AISES Feather Flag.
Earlier that day, I fed an image of the Native American feather flag into a major image recognition app. This deep learning-based visual analysis service couldn’t even begin to identify it properly — because it doesn’t have enough training data. This sacred staff with eagle feathers was instead identified by the AI tool as 99.9% a handrail or a banister.
Reflections from Tracy Monteith, Senior Software Engineer at Microsoft, a member of the Eastern Band Cherokee
While all tribal students and professionals have learned to walk in two worlds both professional and traditional, elders have not had the time to adjust to the rapidly changing pace on which technology thrives. Humans often fear that which they do not understand so there is a natural tension between the ways of the past and the ways of the future.
As tribal members we all need the wisdom of our ancestors to help guide us toward a better future by weaving together our need to build nurturing families and communities, but in addition we must carve out a path of the inclusion of technology so we can ensure a better future with an eye for the seven generations that will follow. Our youth can shepherd a technology-enabled future as our elders keep us grounded in the past.
Increasing the diversity of AI data is an imperative to a more inclusive society and opportunities. As modern societies wrestle with global challenges of climate change, food and water scarcity, displacement, and increasingly hostile living conditions, we can do well to listen to the stories of our elders.
They have lived through the most uncertain of circumstances and furthered stories where we can learn how our ancestors have persevered through millennia of change. These stories from our ancestors teach us views and methods on food, shelter, and stewardship of our natural resources.
When the United Nations speaks of sustainability, the ways of indigenous people are the blueprint of sustainability, conservation, and stewardship. The story of the Three Sisters soup is a microcosm of behavior since it conserves water, enriches soil, provides nutrition, increases sustainable planting methods, and creates micro-climates. It also produces a future of food security since the seeds of corn, beens, and squash can be dried and eaten in the harsh winters and saved for the next year’s planting season.
Reflections from Technologist Chamisa Edmo of the Navajo Nation
So often we are told and shown that we need to perform and exist in specific ways to be taken seriously in highly technical fields. In our efforts to put our ‘best foot forward,’ we oftentimes find ourselves favoring one foot over another in our pursuits.
This sense of hiding and not valuing our perspectives negates our identities as the original scientists, engineers, mathematicians, and technologists. In our fast-paced, hyper-productive world, there is immense power in slowing down to assess what is best and how we can create conscious reverence when it comes to technology and our communities. Utilizing our experience and knowledge base to build culturally relevant AI tools is one way we can ensure we are not erased in the future.
Contact Davar Ardalan if you’d like to join us and support expanding this project: Davar@ivow.ai
Davar is the founder and chief storytelling officer at IVOW — intelligent voices of wisdom. Realizing that there is a gaping hole in AI algorithms that will define our future stories, Davar created IVOW to champion culturally conscious data strategies across multiple industries from academia to development and enterprise.
Ardalan, who is also an Executive Producer at National Geographic, served as co-chair of the Cultural Heritage and AI track at ITU’s AI For Good in 2020. Prior to this, she was Deputy Director of the White House Presidential Innovation Fellowship Program in Washington DC and before that a long-time journalist at NPR News, where she was Senior Producer of the Identity and Culture Unit.
INDIGENOUS KNOWLEDGE GRAPH team at IVOW / Davar Ardalan, Founder and Storyteller in Chief; Kee Malesky, Director of Editorial Content; Karim Ardalan, CTO; Nikki McLay Creative Director; Nisa McCoy, Head of Product, Miroslav Milovanovic, AI Scientist. In collaboration with Wolfgang Victor Yarlott, AI Researcher Florida International University of the Crow Tribe of Montana; Tracy Monteith, Senior Software Engineer at Microsoft of the Eastern Band Cherokee; Technologist Chamisa Edmo of the Navajo Nation; Alva Lim, Co-Founder and Director of Agora Food Studio by the Timor-Leste Food Lab.
Related stories:
