Society / Social Change

Track social change, shifting values, public sentiment and cultural transformation through structured summaries built from curated sources.
What Ancestral Intelligence Can Teach Us About AI | Nanjira Sambuli | TED
What Ancestral Intelligence Can Teach Us About AI | Nanjira Sambuli | TED
2026-03-04T16:00:25Z
Summary
Nanjira Sambuli emphasizes the importance of African proverbs in understanding the implications of AI development. The proverb 'when elephants fight, it's the grass that suffers' serves as a metaphor for the impact of power struggles on marginalized communities. Sambuli argues that great powers, represented by elephants, often overlook the needs and voices of the people, or grass, in the quest for dominance over resources. Sambuli introduces the concepts of Utu and Ubuntu, which highlight the interconnectedness of humanity and the environment. These philosophies advocate for the development of technologies, including AI, that benefit all people and respect ecological systems. By shifting the focus from data as a mere resource to a representation of lives and cultures, data governance can be redefined to prioritize community ownership and informed consent. Examples from Africa illustrate how local initiatives are embracing these philosophies to create AI solutions that are more inclusive and representative. Projects like Leila AI demonstrate that smaller, community-driven models can outperform larger ones by being more attuned to local needs. This approach challenges conventional wisdom that prioritizes scale and efficiency over relevance and cultural context. Masahani's authorship model promotes participation from all contributors, allowing for the publication of research in over 38 African languages. This model not only empowers local communities but also integrates indigenous knowledge into AI development. By fostering agency and collaboration, these initiatives aim to create AI solutions that are beneficial for Africa and informed by its rich cultural heritage.
Perspectives
short
Proponents of Ubuntu and Community-Centric AI
  • Emphasize the importance of indigenous knowledge in AI development
  • Advocate for community ownership and informed consent in data governance
  • Highlight successful examples of smaller AI models serving local needs
  • Promote participation from diverse contributors in research
  • Encourage a shift from viewing data as a resource to recognizing its cultural significance
Critics of Traditional AI Development Models
  • Argue that larger AI models are inherently better for efficiency
  • Claim that societal participation alone can resolve issues of language resource scarcity
  • Suggest that conventional power dynamics in AI development are necessary for progress
  • Maintain that traditional data paradigms are sufficient for current AI needs
Neutral / Shared
  • Acknowledge the ongoing power struggles in AI development
  • Recognize the potential for AI to impact various communities differently
  • Note the importance of biodiversity in ecological systems
Metrics
parameters
0.4 billion units
Inkuba small language model
This demonstrates the model's efficiency in processing language despite its smaller size.
Inkuba is trained on 0.4 billion parameters
weight_capacity
250 times its body weight times
Dung beetle's ability
This analogy illustrates the potential of small models to achieve significant outcomes.
Dung beetle, which can roll up to 250 times its body weight
other
over 38 African languages
number of languages for which translation results have been published
This highlights the inclusivity and diversity in research contributions.
publish translation results for over 38 African languages
Key entities
Companies
Leila AI
Countries / Locations
USA
Themes
#social_change • #ai_development • #ai_for_all • #community_ownership • #data_justice • #ubuntu • #ubuntuq
Timeline highlights
00:00–05:00
The proverb 'when elephants fight, it's the grass that suffers' highlights the impact of power struggles on marginalized communities in AI development. African philosophies like Utu and Ubuntu advocate for technology that benefits all of humanity and emphasizes community ownership of data.
  • The proverb when elephants fight, its the grass that suffers illustrates how power struggles impact marginalized communities, particularly in AI development. It highlights the need for local perspectives in a landscape dominated by powerful entities like corporations and nation-states
  • The African philosophies of Utu and Ubuntu emphasize interconnectedness and mutual responsibility. This philosophy advocates for the development of technologies, including AI, that benefit all of humanity and the environment
  • Data governance in Africa, informed by Ubuntu, views data as a representation of lives and cultures. This approach prioritizes meaningful participation, informed consent, and community ownership of data
  • Data justice ensures that marginalized groups, such as rural women with unique agricultural knowledge, are represented in data systems. This representation is essential for creating equitable agritech solutions that address their needs
  • Initiatives like Leila AI are developing lightweight African language models that effectively serve local communities. The Inkuba model shows that smaller models can outperform larger ones in sentiment analysis across multiple languages
05:00–10:00
Masahani's authorship model includes all contributors in research, enabling the publication of translation results for over 38 African languages. The concept of Ubuntuq promotes AI development that emphasizes agency and incorporates indigenous wisdom for the benefit of local communities.
  • Masahanis non-traditional authorship model recognizes all contributors in published papers, enabling the publication of translation results for over 38 African languages. This inclusive approach highlights the importance of diverse contributions in research
  • The concept of Ubuntuq represents an alternative path for AI development that emphasizes agency and benefits local communities. It integrates ancestral intelligence, allowing for the incorporation of indigenous wisdom into AI solutions
  • The ecological role of elephants in the savannah serves as a metaphor for how powerful entities can positively impact their surroundings. Their actions promote biodiversity, illustrating the potential for responsible power to benefit marginalized communities