What’s the more sustainable path to Africa’s AI participation?

During the closing ceremony of the National Artificial Intelligence Strategy Workshop in April, Nigeria’s minister of technology, Bosun Tijani, announced that the project has received $3.5 million in funding from interested partners. This AI strategy project has been one of the principal agendas of the minister since his appointment in 2023.

Tijani, who has been vocal about his plans to shape the direction of AI in the country, believes that AI is at the helm of the digital conversation now, and Nigeria needs to be a part of that, or risk being left behind.

He’s not wrong.

AI has seen robust investment across the world with AI-related startups raising about $50 billion globally in 2023 alone. Investors and tech companies are putting unprecedented amounts of money into the sector as they have recognised the huge financial returns, as well as the access and innovation that AI is unlocking. Africa alone could see its economy expand by $1.5 trillion if it can capture 10% of the growing AI investment, according to a report. Currently, Africa holds less than 1% of the AI market, and the majority of AI-related projects are surface-level, with investors being unsure if the continent has what it takes to build deeply-rooted AI solutions that can move the needle enough.

According to Olu Oyinsan, managing partner of Oui Capital, Nigeria is in the stage of using AI rather than building AI startups like the rest of the world.

“There are the fundamental models which are capable of taking large amounts of data, and there are AI models that include just using AI as a part of its tech stack—which is where we’re at,” he said.

Oui Capital has five startups in their portfolio that deploy AI, including Duplo, Ndovu, and Maad; and Oyinsan believes that is the more sustainable model of AI introduction on the continent.

What chance does Africa have in the AI race?

According to Oyinsan, the jury is still out on whether or not Africa can catch up to the rest of the world on the foundational level of AI. The major reason being that the continent doesn’t seem to have the resources—like technical talent—required to build AI startups.

“Because tech talent is now more borderless than it used to be, most of the brilliant AI engineers in Africa are not working for African companies,” he said. “They’re developing foundational models for companies outside Africa, which is more lucrative for them, and that is the real problem we have to tackle before we can have an actual African AI movement.”

Between 2022 and 2023, African AI-focused startups in general received a total of $641 million in investments from VCs. In Asia, this figure was as high as $3 billion, with a single startup, Moonshot AI, securing $1 billion in a funding round.

“I wouldn’t say that as investors, we’re looking for AI startups in Africa, but we recognize that startups that can successfully use AI into their existing stack will probably outdo their competitors,” he said. “We’re already behind on the AI arms race, but we’re not behind on the application level. Companies here can use already developed models to plug and play into a tech stack or further develop and customize it for what they need it to do.”

Ayobamigbe Teriba, a venture partner at the HoaQ syndicate community, is more optimistic. He believes that beyond solutions that use AI, building successful AI startups in Africa is achievable and crucial. According to him, innovators on the continent have to find a way to build out these fundamentals that ensure representation and inclusion in the “next innovation paradigm”.

“Building foundational data models in Africa using African data is extremely important,” he said. “We may lose out on major corporate development activities (acquisitions) from foreign corporations in the future if we cannot build AI products that reflect the Nigerian/African reality.”

Teriba believes that heavy investment has to be made into ensuring that African companies can accomplish this, and he is optimistic that, for a country like Nigeria, this won’t take too long.

“The AI groundwork, such as building Large Language Models (LLMs) is underway, catalysed—interestingly—by the public sector [ministry of communications, innovation and digital economy]. I believe that there will be increased capital flows into this vertical, taking a cue from investor enthusiasm in the public markets,” he said.

In 2023, Africa’s pioneering AI startup, InstaDeep, was acquired by a German biotech company BioNTech in a deal worth $684 million. InstaDeep was one of the only AI-focused startups from the continent that scaled to a global level, and Teriba is convinced that the continent could see more of this if the initial investment is made to cultivate them.

According to InstaDeep’s cofounder, Karim Beguir, one of the major reasons why they founded the startup was to show that Africa had the potential to build AI and deep tech solutions.

“The large players like Open AI and Microsoft are not necessarily going to be the ones to solve the problems of Africa,” he said.

One of the biggest challenges that InstaDeep has had to navigate is finding and retaining talent and while the solution for them was to offer stock options alongside mouth-watering financial benefits, this is not realistic for other AI companies on the continent amidst stiff competition from AI giants in other parts of the world.

In Nairobi, the Artificial Intelligence Centre for Excellence (AICE) led by John Kamara is working to find a way around this by training AI engineers for the continent.

Kevin Simmons, a venture capitalist at LoftyInc Capital, shared that more organisations and startups like AICE are building in the fundamental AI space, but aren’t visible due to the sheer time it takes for these larger models to be built and tested.

“It’s too early to conclude whether or not African countries are left behind in the AI race,” he said.