African languages are getting left out of AI, but a new project is trying to fix that. The Masakhane African Languages Hub launched a major funding initiative to build datasets for fifty African languages. It invites researchers and technologists to create speech, text, and image data that reflects authentic cultural contexts. This effort tackles the near-total absence of African languages from the global digital landscape.
The project focuses on three core areas: collecting voice data, testing AI performance in real-world settings, and developing multimodal translation resources. Supported by several major foundations, it aims to prevent harmful biases in emerging technologies. The exclusion of over a billion speakers risks...