TED Talks Daily - How a dung beetle inspired a better AI model | Nanjira Sambuli

Episode Date: February 4, 2026

There's a common African proverb: "When elephants fight, it's the grass that suffers." Policy researcher Nanjira Sambuli says we must apply this thinking to today's AI evolution, asking: When tech gia...nts battle for dominance, who gets trampled in the process? She introduces a new ethical compass for AI, showing how people across the continent are charting a different path for the future of tech.Learn more about our flagship conference happening this April at attend.ted.com/podcast Hosted on Acast. See acast.com/privacy for more information.

Transcript
Discussion (0)
Starting point is 00:00:06 You're listening to TED Talks Daily, where we bring you new ideas and conversations to spark your curiosity every day. I'm your host, Elise Hu. Ubuntu is a foundational philosophy rooted in South Africa, but practiced widely across the African continent that emphasizes interconnectedness, community, and the collective well-being of society as the basis for our moral compass. But what does this philosophical practice have to do with artificial intelligence? In this talk, tech policy, visionary Nangira Sambuli uses the principles of Ubuntu to imagine a world where AI can be both ethical and inclusive. What can the African savannah teach us about AI? Take this journey with me. Across Africa, Proverbs are a cornerstone of the oral tradition through which indigenous knowledge and wisdom has been passed down from generation to generation. The Yoruba people of Nigeria, say that a proverb is the horse that can carry one swiftly
Starting point is 00:01:18 to the discovery of ideas. One of my favorite proverbs says, when elephants fight, it's the grass that suffers. And it has guided me in making sense of merch in the world today, especially with all the developments in AI. The elephants can symbolize great powers, be they nation-states, corporations, broligarchies, while the grass comprises people, geographies,
Starting point is 00:01:44 ecologies considered resources to exploit, wastelands, or charity cases. Great power competition, as we're living through today, emphasizes the power of the metaphorical elephants in their quest for dominance over resources, ideas, and innovations. In the case of Africa, not only are we caught in the middle, but we comprise a key battleground over which elephants are fighting over natural and human resources to power their intelligent age.
Starting point is 00:02:13 Meanwhile, our perspectives and ambitions tend to be drowned out by the din of the elephants. But what can we discover when we look past the elephants fighting and zoom in on the grass? Well, for one, we can learn a simple, yet profound concept. I am because you are. This sums up a set of value systems that emerged among the Bantu people of Africa
Starting point is 00:02:39 and is known as Utu in Eastern Africa and Ubuntu in South and Africa. Utu and Ubuntu instilling us a profound appreciation of humanity as a quality we owe one another. And it's not just about the relationship between humans, Ubuntu is also about our relationship with nature and the spiritual or cosmic. Ubuntu reminds us we ought to be developing technologies like AI for the benefit of all of humanity and our ecology.
Starting point is 00:03:10 across Africa, we have exciting examples of embracing the wisdom of Ubuntu to inform data governance, AI product design, and community building. I call it Ubuntu. AI development today treats data as if it's an abundance, natural resource, and we hear this in sayings like, data is the new oil. But already the limits of this paradigm are being realized, as researchers have been sounding the alarm, that high-quality data to train AI models
Starting point is 00:03:42 is drying up. Through Ubuntu, we conceptualize data differently, and we appreciate that it represents lives, cultures, and communities, so that data governance for us is about the meaningful participation, informed consent, self-determination,
Starting point is 00:04:02 and community ownership of data sets from which language, nature-based knowledge and indigenous wisdom are derived. Thank you. And this has inspired a concept like data justice in our African policy frameworks. Data justice matters because it means that rural women in Africa, for example, who possess unique knowledge about agriculture, food production, medicine, and environmental protection are represented and visible in data systems and agri-tech solutions.
Starting point is 00:04:38 Then, when we've been told ours are low-resourced languages, we're resourcing our languages. Conventional AI wisdom demands large language models, but African practitioners are making do with little language models. Driven by efficiency as a core value and inspired by our relation to nature, initiatives like Lelapa AI have developed lightweight African language models that are serving our communities,
Starting point is 00:05:08 without requiring extensive resources. Their Incuba small-language model has been inspired by the Deng Beetle, which can roll up to 250 times its body weight. It's small, but mighty. Incuba is trained on 0.4 billion parameters and outperforms larger models in sentiment analysis and displays remarkable consistency across multiple languages. We're also building collaborative AI communities.
Starting point is 00:05:40 We have Masahani, all building together, across over 30 African countries to strengthen natural language processing in our research. And this grassroots organizing approach has set out to demonstrate that low-resourceness of languages is not a data problem, but a societal one best solved through participation. In fact, Masahani have developed a non-traditional authorship model that acknowledges and includes all contributors in published papers, be it that you contributed data, lived experience, coded software,
Starting point is 00:06:15 or coordinated research participation. And this way, they've been able to publish translation results for over 38 African languages. These examples of Ubuntu tech are our way of charting an alternative path to developing and deploying AI solutions in Africa, by Africa, and beneficial for Africa. Ubuntu matters for a number of reasons. For one, we are a starting agency
Starting point is 00:06:42 to conceive and build AI futures beyond just the ambition of the metaphorical elephants and to contribute to a global commons. It also allows us to bring forward the indigenous wisdom of our ancestors so that Ubuntu is artificial intelligence powered by ancestral intelligence. And we also go to the indigenous wisdom.
Starting point is 00:07:09 forget to remind the elephants. They may fight, and will their might all they want, but they're also bound to suffer if they trample upon the grass to the point of irreparable damage or extinction. But when their power is exercised in relation to others, it makes them ecosystem engineers and redirects their energy
Starting point is 00:07:29 towards helping sustain a healthy and beneficial environment for everyone. And we see this in the savannah, when African elephants leverage their power to trample upon dense shrubs and acacia trees, they make room for smaller species to coexist. When they disperse seeds as they trek across the land, they help generate new growth and maintain the biodiversity of the savannah ecosystem.
Starting point is 00:07:54 So, in relationality, in coexistence, the power of the elephants is majestic, and it's a life force for themselves, are the wildlife and the savannah ecosystem. I believe, if we can reimagine, in humanity, beyond just the ambition of the metaphorical elephants, we can realize a world that benefits everyone. So I implore all of you. Take heed of the grass beneath your feet. Our collective future depends on it. Asanteen. That was Nangeras Sambuli, speaking at TED 2025.
Starting point is 00:08:40 If you're curious about Ted's curation, find out more at TED.com slash curation guidelines. And that's it for today. TED Talks Daily is part of the TED Audio Collective. This talk was fact-checked by the TED Research Team and produced and edited by our team, Martha Estefanos, Oliver Friedman, Brian Green, Lucy Little, and Tonica, Song Marnivong. This episode was mixed by Christopher Fazy Bogan.
Starting point is 00:09:03 Additional support from Emma Tobner and Daniela Balerozzo. I'm Elise Hu, I'll be back tomorrow with a fresh idea for your feed. Thanks for listening.

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.