As nations plot how best to unlock greater Artificial Intelligence (AI) potential to drive economic growth, African techies warn it may take up to 10 years to make the strides made so far by countries in the developed world.
African nations face a myriad of hurdles, preventing the new-age technology from making a significant traction.
In interviews with the Business Daily on the sidelines of the recent Gitex Africa 2024 expo hosted in Morocco, innovation executers from several countries painted a picture of an impending tough hassle as African businesses attempt to play catch-up with the concept that many have termed ‘futuristic’.
Among the factors prominently listed as key hurdles include data scarcity, skills shortage, insufficient tech infrastructure as well as a lacklustre attitude by the masses on the continent.
“A key challenge will be getting the right talent of people who have properly studied AI so that we ensure that the models are properly trained in a way that they are ably equipped to perform the intended task intelligently,” stated Ken Okolo, the head of commercial partnerships, Middle East and Africa at UK-based Raspberry Pi Foundation.
“There’s also the uphill task of ensuring AI is not used wrongly to pose new challenges on fronts such as security and democratic happenings.”
His sentiments were echoed by Essowe Bakoubolo, founder of Togolese tech startup Umbaji, who said that at the current rate of adoption, usage levels in Africa will mirror those of the developed world in at least 10 years.
“Technology is one of the domains that are rapidly evolving and AI is right at the heart of it. As such we have a lot to research on to achieve AI for African communities because we still are grappling with challenges around the assembly of the correct data in areas such as speech recognition and text-to-text translation,” stated Bakoubolo.
“These are fields that we really need to work closely with local communities to achieve tailored models because AI is a powerful tool for social development and we cannot afford to ignore it.”
The founder also flagged deficiencies such as poor internet penetration rates in many African nations, calling on governments across the continent to step up investments in local tech infrastructure.
For Yeabkalu Merkebe, the deputy project lead at Ethiopian health tech firm Adot, the prime pain point is the unavailability of locally-aggregated data to power the building of AI solutions for the African context.
“The key challenge we’re seeing is having data that is customised for Africa. For a start, we’re utilising ChatGPT but even that is not very satisfying, especially for local languages and so that’s a huge problem,” he noted.
“To train a model using data from other continents is creating a huge bias and we do believe that for us to adopt AI, we have to ensure that we digitalise every aspect of our lives. That’s the only way we can get enough data so that so that our models are not as biased as they are today.”
A number of countries, including Kenya, have put in place data legislation to regulate the processing and usage of personal information by processors, as well as establishing a framework for licensing those who handle data. Kenya enacted its Data Protection Act in 2019.
Other glitches in AI adoption hurdles cited in the past include the continent’s low capacity for virtual storage as well as increasingly outdated mobile technology.
But amid the teething challenges, corporate firms that have embraced AI have reported enhanced efficiency in their operations, a signal that has prompted players in the space to devise new strategies aimed at driving uptake.
Last month, tech giant Microsoft announced that it would support the development of an AI model in the Swahili language as part of its grand plan to invest up to $1 billion (Sh128.5 billion at current conversion rates) in Kenya’s digital ecosystem.
The multinational firm said the initiative will be geared towards supporting Kenya’s unique cultural and linguistic needs in a development that has been touted as the magic wand that will drive AI uptake among native communities.
Local firms have also stepped up their investments in virtual data storage facilities to enhance capacity as well as connectivity speeds in readiness for AI integration.
