Indigenous Digital Equity in the AI Era: Connectivity and Opportunity
THE ISSUE:
In an era where digital connectivity is crucial to the economy and AI is set to transform how work is done, many Indigenous communities are already at a distinct disadvantage: an estimated 50 percent of indigenous communities in BC still only have access to dial-up internet. In Manitoba, the numbers are even more bleak: only 15 per cent of households on First Nations have access to high-speed internet.
Shawn Gervais, VP Strategic Foresight, Digital Technology Supercluster, Shazia Zeb-Sobani, VP Customer Network Implementation, TELUS, and Kory Wilson, Executive Director of Indigenous Initiatives and Partnerships, British Columbia Institute of Technology discussed some of the implications of adding AI (artificial intelligence) to the mix.
Kory Wilson, Shazia Zeb-Sobani and Shawn Gervais at the 2023 Indigenous Partnerships Success Showcase.
AI holds the promise of increasing productivity, by helping people save time on basic tasks, as well as those more complicated. As Shazia Zeb-Sobani from TELUS notes, “It's actually going to enhance our decision-making capabilities. Just think about the amount of data and the amount of information that we need to pull to make a decision. It can take us hours and days and weeks and sometimes months to get that data and synthesize it and put it in a form of meaningful information. AI can do it very quickly. And, it can really help us open up very different options and evaluate them.”
While many fear that AI will replace humans in their jobs, Shawn Gervais of the Digital Technology Supercluster says it’s more likely that “you're going to lose (your job) to a human using AI.” He adds we should “recognize the reality that every job's an AI job now, and we need to think about what that means from a proficiency.”
If AI is going to be so ubiquitous, then equitable access to it is going to be critical, as those without access risk being left behind.
“Access is the key,” says Kory Wilson of the British Columbia Institute of Technology. “And access can mean a lot of things. It's the actual physical access, being able to get onto the internet, onto the computer, onto the access to the technology, but also the skills and abilities needed to understand and interact with that technology.”
“You (can talk) about AI becoming access, but to access AI you actually need to have access, which is the connectivity,” says Zeb-Sohani. “So, really the digital equity starts by creating a level playing field where everybody living in Canada, especially the Indigenous peoples of Canada, do have access to the robust, reliable connectivity where they can actually get onto the internet, where they can access the global information ocean.”
To tackle the challenge of ensuring access, Gervais says that we should think of AI as infrastructure, rather than simply as a technology. “It's not an optional service. It's not a nice thing to have. It's an essential service … And that means everybody needs to have a pathway to get access so that they can take advantage of the opportunities.
“That idea of thinking of AI as an essential service for Indigenous communities, I think opens up a different perspective. And you can start thinking of ‘Why can't we use infrastructure dollars to invest in AI?’ It doesn't have to be high tech or computer or venture—it's infrastructure dollars. You can embed that in infrastructure, maybe infrastructure co-ops or infrastructure companies where the Indigenous have more than a voice at the table, they actually are active participants and active players.”
When it comes to AI and equitability, there are other issues to consider, including the potential for AI to be biased.
“Going to the core of AI or any kind of computer generated, machine generated (systems) we know there are biases built into those systems,” says Wilson. “We look at any institution in Canada, there's systemic racism and discrimination built into these things. And the machines to a point, they respond to what's put in there. And when you look at knowledge and you look at data, and you look at histories, and you look at storytelling or whatnot, you know, absent often is the Indigenous voice or truth.”
Zeb-Shobani agrees that it’s critical to address such ethical considerations. “How do we make sure that the models, the large language models that have been developed and the algorithms behind it, actually have a respect for different cultures and particularly for Indigenous values and cultures and perspectives?”
Then there are concerns about data ownership.
“I think our challenges around data privacy and the use of data are going to be very different in the age of AI,” says Zeb-Sobani. “We haven't solved data privacy yet in the non generative AI world.”
Gervais adds that it’s not solely the data used in AI, but also the frameworks behind it, that must be weighed. “For example, if we're talking about a person, we'd say, ‘Where did you go to school?’ To kind of get a feel for their background. If I'm talking to an AI, ‘Where did you train? What was the data that you trained on?’ When you talk about working with people, and you're trying to get independent decision making, there's a process we use with people around conflict of interest. So, if I have a conflict of interest, I declare. How does an AI declare a conflict of interest? An AI that has given you recommendations. You know, that's not something that you think about if you're just doing a data model.”
He also notes that “technology tends to homogenize … So even if you have inclusivity, that inclusivity doesn't prevent homogenization. And I think one of the things that, especially when we're thinking about Indigenous reconciliation, we're thinking about honoring different cultural perspectives.
“So, the counterweight is to not have a single model. (It is) to have multiple models with multiple perspectives … If technology creates a convergence and you don't want convergence or you want to honor different perspectives, then you obviously need a different model or a different solution.”
Zeb-Sobani sees inclusive AI development as vital. “This is the opportunity for us collectively across the globe, and especially in Canada, where we can have Indigenous people sitting side by side right from the start and be part of the solution and influence that. We need to embrace AI and we need to be in a capacity where we can influence it.
“The opportunity is to not repeat the mistakes of the past. To invite Indigenous people to sit at the table as we are making big decisions, how to deploy it, use it, govern it, and have them in equal voice in making those decisions. I think it starts with listening … It really starts with creating respect and listening and understanding each other's perspectives, and especially the perspectives of Indigenous led experiences and from Indigenous communities.
“We all have a role to play, and we need to determine that. How can we take accountability directly? We are better together. We are stronger together, and we are more effective together.”