AI will have a huge impact on Africa. But we must avoid the hype

Kathleen Siminyu is an AI Researcher who has focused on Natural Language Processing for African Language. She is a 2025 Mozilla Rise25 honouree. Mozilla’s Rise25 awards celebrate the people leading the next wave of AI - using philanthropy, collective power, and the principles of open source to make sure the future of AI is responsible, trustworthy, inclusive and centred around human dignity. 

I have worked in artificial intelligence (AI) for eight years and the current hype cycle around generative AI has made this easily the most exhausting year of my time in the field. 

It is both fascinating - and unnerving – to see just how much the term AI has entered into everyday conversation.

I now shy away in social settings from saying outright that I work in AI when I introduce myself, because the follow up conversation is likely going to be more about what people have heard, learned, or are doing with AI tools - and these will most often be the generative AI tools that have percolated through every fibre of our existence. 

Generative AI is now, for the public, AI. 

But generative AI is just a sub-field of a much wider field, yet it is understood to be the beginning and end of that field. It is believed to be a silver bullet for everything. 

In development funding, Artificial Intelligence for Development (AI4D) programs are getting as much as five times more funding than was allocated to them just five years ago. 

While this is exciting for researchers and practitioners from the developing world who are eligible for this funding, the knowledge that some of this funding has been re-allocated from poverty alleviation programmes is sobering.

Do we expect that AI is going to alleviate poverty on the African continent? Generative AI no less?

The excitement around a new technology and the expectation that it will solve all developmental challenges is not new. About 20 years ago, there was a lot of focus on what was known as Information and Communications Technology for Development (ICT4D), precipitated by personal computers becoming household items. 

It was believed that ICT skills would improve the lives of the less fortunate in Africa. I am sure that many people benefitted from the programmes run in that time, but clearly not enough for us to not need yet another silver bullet. 

Tech is a tool

Narrowing in on a single sector, let’s look at education. With the belief that ICT skills would be game-changing, many countries implemented schemes such as the One Laptop per Child (OLPC) programme. OLPC was a non-profit initiative with the goal of bringing low-cost computers to children in the developing world.   

In Kenya, it was well meaning and looked good on paper, but the feasibility of implementation was something that lagged behind. 

The government promised to issue laptops to all pupils joining primary school in 2013. In 2019, this was suspended and the government instead committed to the construction of computer laboratories for ICT integration in each public school - a cheaper and therefore more attainable goal.   

Now, with the excitement around AI, there are proposals to take this one step further. Devices are once again proposed to bridge gaps in education and this time they come with AI applications. 

At the level of formative education, AI tools are proposed to improve child literacy rates. Now don’t get me wrong, this is an exciting prospect, but learning more about the education sector in Kenya makes me aware of so many gaps that AI would not be able to fill. 

We could improve the education sector by building more schools, training more teachers and providing bursaries. These ideas would do more for the system. Additionally, children are the most vulnerable amongst us and while these tools are still experimental, it does not seem fair to deploy solutions targeting them. 

AI will have a huge impact on many sectors. But my belief is that a lot of what we hear right now about its possibilities is just hype. 

We need to ground our expectations in the fact that tech is a tool. It is wielded by us and has limitations. It will not be a silver bullet for anything, but applied to targeted problems can bring change.

I am excited to put AI tools, and language tools in particular, in the hands of content creators.

Crucially, I would like to enable the creation of more content in less dominant languages and, for me, the novelty of having equal support for African language speakers on digital platforms is something I can truly get behind.