AI Onto-Norms and Gender Vulnerabilities in the African Context

Imagine that AI was food, or a product being sold, or even being offered for free, then we find ourselves in a world that is cheering all of us on to make use of the golden opportunity, and we all start unconsciously getting sucked into the agenda, and every day we try bits and pieces of this product. I relate AI to what Morton (2013) calls a hyperobject, it is complex, viscous, non-local – we interact with it daily, we touch it, we experience it, but we can’t explicitly comprehend what it is. Everyone is trying to explain it, and the more we try, the more we can’t explain it. The more we try to avoid it, the more we find ourselves falling freely into its flashy and beckoning aesthetics embedded in the invisible materialities. Yet as much as we try to make it flashy, even the best experts struggle with uncanny feelings about the anticipated risks associated with AI. I prefer to stay from the dystopic or doomist approach to focusing on the impacts AI is having on people now.

Let’s talk about food, In the African context food holds a very special place. It brings communities together and it can also break them. The way we name food, prepare it, and consume it, differs from one community to the other. The stories about food are exceptionally situated, they are specific for specific communities, and they have practices that are tied to them. ‘Doing’ food, whether it is preparation or eating, is a practice of a specific place, with embedded norms tied to those practices. This is what Annemarie Mol (2013) calls Onto-norms – which are produced and maintained through practices and embedded norms. Onto-norms shape how different people perceive and interact with their world.

Let’s relate AI to the food we prepare and consume. Like food, AI comes in two versions, ready to eat or a version that needs preparation. Whichever way we want to see it, the end practice associated with food is consumption. When food is already prepared, there is a way we will eat it. While some of us cautiously assess the ingredients that prepared it to include the calorific intake, others just consume for the hype of it. When we must prepare it, we regulate the practice of preparation in such a way that the food suits our values and interests, and makes it interesting to eat it. On another note, when food is free and ready it tends to be over-consumed regardless of the value it adds or removes from our bodies.

The way AI is consumed by African women depends on the designer, the product (aesthetics, cost, and hype), and the consumer’s AI onto-norms. While narratives around gender equality in AI narrowly focus on bringing more women into STEM and enriching female data, very little research pays attention to how AI onto-norms are shaping gender norms in the African space, and how tech multinationals are exploiting these onto-norms through algorithmic recommendation and AI-driven content engagement. Our research shows that AI onto-norms are continually being exploited by large tech companies to make money while driving deeply entrenched negative gender norms in those spaces. Are we ready to face these blurry shades of AI exploitation? Or would we rather hold on to the rosy stories of addressing what rationally and contextually makes sense?

If you love this content, hit the like button.

And follow me on Africanaiethicist.com for more or these nuanced snippets.