AI in Africa: Less Leapfrogging, More Sequencing
My four-point response to the question: “Can AI Take Off in Africa?”
I recently joined VoxDev for a podcast conversation on AI in Africa, now out alongside a thoughtful write-up on their Substack, Ideas in Development. The question at the center of our discussion was a simple one: Can AI take off in Africa?
We cut through much of the surrounding hype and focused instead on the less glamorous, but far more consequential, conditions that will shape whether AI actually delivers value on the continent.
Here are four points I argued are essential to answering the question:
1. AI runs on basics, not buzz.
If African countries are to meaningfully participate in the AI era, they need the same enabling conditions that underpin modern digital economies everywhere else: affordable and reliable power, robust digital infrastructure (from data centers to fibre networks), widespread connectivity (including affordable access to cloud services), and high-quality local data. There’s no AI shortcut around these fundamentals.
2. Embrace sequencing (don’t let perfection be the enemy of AI deployment)
African countries do not need to complete electrification or achieve universal connectivity everywhere before AI can be useful anywhere. The more realistic approach is sequencing: start in the places and sectors where adoption is feasible now, while steadily building the foundations for broader access over time. Progress doesn’t require a fully “finished” system, just smart prioritisation.
3. AI is not a classic leapfrogging technology.
Unlike mobile phones or digital payments, AI doesn’t bypass infrastructure—it depends on it. Compute, energy, data systems, and institutions matter enormously. Treating AI as a frictionless leapfrogging opportunity risks repeating familiar mistakes from past tech-for-development cycles, where promising technologies were layered onto weak systems and expected to compensate for deeper structural constraints. Two decades of excitement over mobile-based tools “revolutionizing” agriculture in poor countries, for example, have delivered far more pilot projects than durable transformation.
4. Africa needs more of its own AI thinkers.
We closed the episode by previewing work underway at the African Tech Futures Lab. The aim is to cultivate deeper, more grounded African thinking on AI—by Africans, on African terms. There’s no shortage of Substacks and commentary on the future of AI, and a cadre of influential Global North thinkers (think Ethan Mollick or Gary Marcus, for exampl) are shaping the mainstream conversation. There are still far too few African voices shaping how these debates land on the continent.
This piece was originally published on Kibao, a Substack newsletter, and is reposted here with permission. The full episode can be found here, or wherever you get your podcasts.
