African artificial intelligence (AI) is about people. As the Stellenbosch University (SU) representative in AHEEN, I was struck by the story of a data science intern in a Kenyan refugee camp. She worked with a refugee-led student support organisation and applied her data science knowledge and skills in the organisation and to the problems at hand. The feedback on her work was incredible and showed just how much difference a skilled graduate can make in even just a short time. AI in Africa should focus on our talented and resourceful people who can use their AI-related data science skills to bring measurable change where it is needed most.
African AI is about language. The Deep Learning Indaba, initiated by Prof Vukosi Marivate of the University of Pretoria, and its grassroots AI projects, Masakhane and Lelapa AI, develop large language models for African languages. The project is described as “a grassroots natural language processing (NLP) community for Africa, by Africans". Here, we see the power of language, which is such a defining feature of our diverse, multilingual continent. While existing large language models (such as ChatGPT) lean towards English and Western languages, these projects help establish an African language, thought and philosophy approach to the training of AI systems. Locally developed AI tools will better serve our African thinking and doing, feeding our insights into the global community. In this way, we can create a more decolonised future in which AI systems 'think through' isiXhosa, isiZulu, Sesotho, Afrikaans, Arabic, Swahili and the like to provide equal and just information, learning, services and opportunities to all African communities.
African AI is about innovation and contextual entrepreneurship. In his contextual innovation and entrepreneurship course developed for the agrisciences industry, Dr Albert Strever (SU) intentionally incorporated AI tools such as ChatGPT and QuillBot in the students' curriculum and learning activities. In the process, he invested in the students' AI literacy and prepared them for a near future where the question will not be whether humans will be replaced by AI, but whether humans with AI will replace humans without AI (Prof Karim Lakhani). Lecturers should become leaders in incorporating the AI insights of their disciplines into their own teaching, learning and assessment. And, of course, lecturers themselves should also learn about AI in higher education!
AI tools are now available to all higher education stakeholders (students, lecturers and support staff). By focusing on people, language and contextual (Africa-oriented) innovation, African higher education can contribute immensely from a global south perspective. There are a plethora of new social entrepreneurial ideas just waiting to be turned into reality with the help of our computer, data and machine-learning scientists for the benefit of society.
I conclude with the underpinning philosophy of the Deep Learning Indaba, which strongly resonates with me:
“We work towards the goal of Africans being not only observers and receivers of the ongoing advances in AI, but active shapers and owners of these technological advances."