Do Africans Want Chatbots?

Over the past decade, Africa’s NLP community has made major strides in building language technologies for low-resource contexts. Yet, as large language models and generative AI take center stage, we must ask a deeper question: Do Africans actually want the kinds of chatbots and “intelligent” systems now being built for them?

This keynote traces the journey from early NLP and translation interfaces to today’s “everything machines,” exploring how human frustrations, design gaps, and cultural mismatches persist beneath the surface of technical progress. Language is not just data, it is deeply human, and our tools must reflect that.

I will argue that while global AI systems promise universality, they remain brittle, narrow, and shaped by traditions that often ignore local needs, values, and languages. To move forward, African HCI and AI researchers must take an activist stance: designing for people, not just benchmarks; embedding African philosophies and values into AI; and reclaiming the conversation about what a meaningful technological future looks like.

In asking “Do Africans want chatbots?”, we are really asking how Africans want to speak with machines, and to be heard.

‍