Africa's AI Challenges: A Call to Action
On the first day of the 7th edition of AfricArena, held on Tuesday, October 15, at the Tunis City of Culture, Abderrahmane Chaoui, Advisor at Sendemo, highlighted the stakes of Artificial Intelligence (AI) in Africa, in a global context marked by titanic investments.
The Global AI Landscape
Giants like Alphabet, OpenAI, Meta, and Microsoft have raised astronomical funds: $50 billion for Alphabet, $55 billion for Microsoft, and $37 billion for Meta, he reminded the audience. Moreover, Dario Amodei, CEO of Anthropic, estimates that a model like GPT-6 could cost up to $100 billion by 2027. The Moody's agency predicts that global data center infrastructure will require trillions of dollars in the coming years.
Africa's AI Challenges
In the face of this race, Africa finds itself in a position of weakness. With only 145 data centers on the continent and a majority of the population (53%) lacking access to electricity, the continent risks becoming a mere consumer of infrastructures developed elsewhere.
In terms of financing, AI spending in Africa is expected to reach $3 billion in 2024, with forecasts of $6.6 billion by 2026. Despite some progress, such as the announcement of two new hyperscale data centers, one in South Africa and the other in Kenya, the infrastructure remains largely insufficient.
The lack of AI talent and weak links between the academic world and the private sector also pose a problem. However, promising initiatives are emerging: universities of excellence, such as CMU and AIMS, as well as Google's research center in Accra, are starting to train qualified professionals. Nevertheless, the continent must overcome obstacles such as unfavorable regulation and unequal access to the internet and education.
Two Types of AI
Chaoui proposes considering two types of AI in terms of transformative potential: incremental AI, which consists of acquiring basic skills to interact with AI, and transformative AI, focused on applied research and collaborative projects between universities and businesses.
AI Applications with High Potential
Several applications with high potential are developing in various fields, including:
- Agritech: crop monitoring, pricing, and distribution.
- Fintech: credit scoring and debt facilitation.
- Commerce: chatbots and customer assistance, sales forecasting.
- Logistics: robots and predictive algorithms.
- Health: genomics and cancer detection.
- Mobility: autonomous vehicles and large-scale automation.
- Smart Cities: energy production and distribution.
A Call to Action
For AI to become a true engine of development in Africa, Chaoui asserts that it is essential to establish a clear vision of what we want it to achieve. By strengthening our internal capacities and fostering synergies between the private sector and academic institutions, Africa can transform these challenges into opportunities.
