Session
Organizer 1: Linda Bonyo, 🔒UNECA
Organizer 2: Peter Addo, Agence Française de Développement (AFD)
Organizer 3: Risper Onyango, Lawyers Hub
Organizer 4: Jorge Clarke, 🔒
Speaker 1: Peter Addo, Intergovernmental Organization, Intergovernmental Organization
Speaker 2: Chenai Chair, Private Sector, African Group
Speaker 3: Hon. Neema Lugangira, Government, African Group
Peter Addo, Intergovernmental Organization, Intergovernmental Organization
Risper Onyango, Civil Society, African Group
Jorge Clarke, Private Sector, Latin American and Caribbean Group (GRULAC)
Round Table - 60 Min
1. The question on Language Preservation - How can policy support the preservation and promotion of African languages through generative AI models
2. The question on Data Collection - How do we ensure the availability and quality of data for training generative AI models in African languages?
3. The question of Bias and Ethical Consideration - How can policy address ethical concerns and mitigate biases in generative AI models for African languages to promote accessibility and inclusivity of generative AI models for African languages?
What will participants gain from attending this session? Participants will leave the session with a deepened understanding of the possibilities, challenges, and responsibilities associated with generative AI and African languages. They will be equipped with practical tools and insights to actively contribute to the advancement of this field, fostering linguistic diversity, cultural preservation, and technological innovation in Africa. This will include gaining a comprehensive understanding of the current landscape and potential of generative AI models for African languages, exploring latest advancements, research, and case studies in this domain, and gaining insights into the importance of preserving and promoting African languages through generative AI models, emphasizing the need for linguistic diversity and cultural representation.
Description:
The session “Generative AI and African languages" aims to address the issue of a lack of diversity as a result of many generative AI models being trained primarily on data from English and other Western languages, which limits their ability to generate content in other languages, especially under-resourced African languages. This leads to biases and reinforces existing inequalities in language representation and access. The session seeks to explore how generative AI models can be used to promote linguistic diversity, specifically in African languages. The session may cover topics such as developing generative AI models for African languages, the challenges of working with limited language data, and the potential benefits of promoting linguistic diversity in AI technology.
1 Policy brief on Generative AI and African Languages as a result of Participants collective development of policy recommendations or guidelines addressing the ethical, legal, and societal implications of generative AI in African languages.
Knowledge Exchange and Collaborative Initiatives among participants, which can lead to the formation of networks, partnerships, or working groups focused on generative AI and African languages.
Hybrid Format: To facilitate interaction between onsite and online participants in a session, we will - enable online attendees to ask questions and engage with speakers and onsite participants to ensure everyone has an opportunity to contribute.
The session will have an online moderator who will facilitate communication, manage questions, and ensure a smooth flow of the discussion. Clear instructions on how to join the virtual platform and participate will be provided, ensuring accessibility for all attendees. To enhance engagement and inclusivity, interactive elements such as polls, quizzes, or breakout sessions will be incorporated into the session. Complementary online tools and platforms will be utilized to augment the session i.e. Twitter and LinkedIn to share updates, insights, and encourage real-time discussions. Virtual whiteboards or collaborative note-taking tools like Miro or Google Docs will gather input and feedback from attendees, promoting collaboration.