African Advantages in AI Discussed at ICT Editors Xchange
At the 7th ICT Editors Xchange held this week in Johannesburg, industry leaders emphasized that Africa’s edge in artificial intelligence (AI) will hinge on developing systems informed by African data, promoting education in indigenous languages, and securing ownership of relevant cloud infrastructure. This event, sponsored by Huawei, gathered various stakeholders from the tech giant alongside representatives from South Africa’s academic and business sectors to discuss the importance of local data, language accessibility, and cloud resources.
Data Disparity in AI Music Models
A significant highlight was the revelation from a Johannesburg-based music AI startup, which reported that a mere 0.04% of global data sets utilized in AI music models originate from Africa. Gift Rubele, the founder and CEO of AuraaAfrica, addressed attendees about this discrepancy, noting that an overwhelming 94% of AI music models are built on Western data. He warned that neglecting to incorporate African perspectives into these AI models could threaten the continent’s cultural representation in the digital age. Rubele urged Africa to transition from being mere consumers of technology to creators of its own datasets. His company recently earned the Grand Innovation Award in Huawei’s Code4Mzansi developer competition, standing out among 1,040 submissions.
Contradictions in South African Education
Dr. Olapeli Okugboyejo from the University of Pretoria shed light on unsettling contradictions within the South African education system. Despite achieving a record 88% pass rate in mathematics—the highest in history—the country is on the brink of a learning crisis, with 81% of fourth graders unable to read or comprehend any language. Okugboyejo emphasized that existing AI technologies often fall short in supporting African languages, directing her research toward indigenous language datasets and pan-African language models.
Redefining Intelligence Requirements
Opening the conference, Huawei South Africa’s Deputy CEO, Charles Chen, articulated a vision for Africa to define its own intelligence needs rather than adopting external models. He described “African intelligence, in African parlance,” as systems specifically structured to meet the continent’s unique challenges. Chen highlighted the necessity for innovation through collaboration among small and medium-sized enterprises, educational institutions, and public organizations, stating that Africa’s upcoming digital transformation will rely heavily on cloud computing, AI, secure networks, and a robust data infrastructure.
The Importance of Adaptability in AI
Digital futurist Nikki Vard delivered a keynote address titled “The future belongs to adaptability.” She stressed that the strength of AI in Africa lies in its capacity to respond swiftly without being hampered by outdated systems. Drawing lessons from the downfalls of companies like Kodak and Nokia, Vard cautioned against “titanic thinking” among corporations that rest on their laurels. She pointed out that the primary threat extends beyond job losses; it also encompasses a mental stagnation that can arise while still employed.
Opportunities Amidst Challenges in AI
Calvin Huang, Director of Cloud Solutions Sales at Huawei Cloud South Africa, discussed the swift evolution from generative AI to physical AI, echoing founder Ren Zhengfei’s belief that this technology may represent the final significant technological revolution for humanity. He indicated that investment in AI is poised to reach $810 billion by 2028, while also acknowledging the ongoing challenges South Africa faces in areas like implementation, costs, and data sovereignty. To navigate these obstacles, Huang highlighted Huawei’s commitment to developing open AI systems, establishing local infrastructure, and forging industry partnerships, with the company’s AI infrastructure currently set up in the Johannesburg cloud region.
