Artificial intelligence is increasingly used in Africa for tasks such as writing letters, translating documents, preparing lessons, and developing businesses. However, despite this visible presence, Africa remains largely excluded from the manufacturing and decision-making processes surrounding these technologies, where access and pricing are determined, and where the value generated is concentrated. This means that while AI tools are becoming more common, the decisions about how they are created, sold, and used are often made outside of Africa. Africa's continued trajectory may result in missing an opportunity to redistribute power in the technological sphere, instead becoming a consumer dependent on external decisions and resources. This situation is not attributed to an inherent African curse but rather to material, scientific, and political conditions that enthusiasm for technology alone cannot change. Africa possesses well-trained minds but lacks the capacity to equip its population to become a collective force, with the primary responsibility resting with those in power. The deployment of artificial intelligence requires significant infrastructure, including processors, data centers, cooling systems, and a reliable electrical supply. However, sub-Saharan Africa still had around 600 million people without electricity in 2024. In 2025, 36% of the population in the "Africa" region used the Internet. The notion of a revolution accessible to all is challenged by these conditions, which highlight the gap between the existence of a tool and its actual usability. In 2023, sub-Saharan Africa allocated 0.38% of its GDP to research and development, compared to a global average of 1.92%. The region has 88 researchers per million inhabitants, compared to 1,486 globally. This disparity underscores the challenge of building a scientific capacity in the field of artificial intelligence. The World Bank reported that high-income countries hosted 77% of the world's colocation data center capacity in June 2025 and received 91% of the venture capital invested in artificial intelligence. African startups often rely on models designed elsewhere, host data elsewhere, and pay prices set elsewhere. This dependency can be precarious if suppliers increase prices or discontinue services. Despite these challenges, there are examples of progress. The South African Lelapa AI has developed InkubaLM, a small language model dedicated to several African languages, and has made its work public. The World Bank acknowledges that developing countries can benefit from specialized and inexpensive models without building large systems themselves. Africa must choose its dependencies, master strategic skills, retain value, and ensure benefits reach education, agriculture, and public services. The political difficulty remains in translating strategies into action, with limited resources often leading to the launch of nearly identical projects in every country, which can be more costly than sharing computing power and research programs. While the African Union adopted a continental strategy on artificial intelligence in July 2024, the effectiveness of this strategy is measured by its implementation, including access to computing power and the development of local solutions beyond pilot projects.