Speaking at the CITO East Africa Pitch Parade at Capital Club East Africa in Nairobi, Dr Mark Nasila, chief data and analytics officer at FNB South Africa and author of two books on artificial intelligence, made the case that Africa's relationship with technology needs to change, from adoption to ownership.
Dr Mark Nasila, chief data and analytics officer at FNB South Africa has been tracking big data and AI since around 2010, when it became clear that companies like Google and Facebook were not simply building platforms. By 2018 and 2019, one thing stood out to him about AI specifically.
“Other technologies came and basically expanded, or they came and scaled. This technology specifically reimagined things. At the centre of that reimagination is the future of work and organisations looking at what they can offer,” he said
His first book, African AI, grew from that observation. He argues that Africa's default response to new technology has been to wait for it to arrive from elsewhere and then adopt it.
“We in Africa wait for technology to be discovered, and it's actually a good way of how innovation works, we call it innovation by iteration or by analogy. But the gap between the potential we have and the potential of what has been pre-developed is what we are going to miss if we just adopt this technology as is,” he added.
His second book, Mastering Sovereign Artificial Intelligence, takes that argument further. Mark frames AI as the new form of industrialisation, drawing a direct parallel to how raw material exports historically left Africa with the by product rather than the value.
“The equivalent we are talking about with AI is data. Manufacturing intelligence and using it to drive decisions, the process of it has got much more benefit than the output itself. The person who owns the factories will create intelligent output that will help drive efficiency and create jobs,” he explained.
For the CIOs and CITOs in the room, his message was about where the conversation needs to move after four or five years spent debating model sizes and generating use cases. An estimated 47 percent of AI projects are expected to be dropped by the end of next year.
“Have clarity on your customer needs so that you can pick the right technology. Do not send the question of value outside of your organisation. Let that question sit with the inefficiency, ineffectiveness and opportunities for innovation that lie within,” he said.
His closing point brought the argument back to people. Only six percent of organisations are currently realising the value they expected from AI, and Mark’s explanation is consistent with everything else he said.
“Start by asking yourself the strategy and the objective. What is it that AI cannot do? And start sufficiently training people to move there, so that when AI implements the task, they are able to be productive with what AI cannot do. The AI strategy becomes a people strategy,” he said.
