5
A major challenge is emerging as AI advances: most of Africa’s estimated 1,500 to 3,000 languages are being left behind because of insufficient digital data. While English has millions of online resources, many African languages have few or none, making it impossible to use them for training standard large language models. This data deficit risks locking millions out of economic and educational opportunities. To resolve this problem, pioneering researchers are reimagining AI development. Projects like African Next Voices are creating specialized, high-quality datasets by recording native speakers on vital topics like health and agriculture. The goal isn’t to replicate ChatGPT overnight, but to build accurate, culturally-aware tools for specific needs. However, experts stress that infrastructure gaps and funding remain major hurdles to developing AI that truly represents Africa’s linguistic diversity.
CNN