Blog

Africa’s population is growing and will continue to do so. Ghana alone has an estimated population of 30.4 million people and faces problems such as pollution and access to quality healthcare. As a result, it makes an interesting place to implement Artificial Intelligence (AI). Delivering AI has the potential to enhance quality of life for Ghanaians now, as well as for future generations. This is why Google has launched an AI lab in Accra, Ghana’s capital.

CURRENT USES 

To address issues of food insecurity, Dr Mercy Lung’aho from the International Centre for Tropical Agriculture in Nairobi, Kenya is working to develop an AI which predicts nutrition crises. Called NEWS, the system analyses satellite imagery, as well as data such as rainfall, temperature, and vegetation health to make predictions about crop health and nutritional value. By this information, this technology can help ensure that people have access to nutrient-rich food. This could lead to enhanced wealth and well-being, which has the potential to lift employment rates and impact economic growth positively.

Another AI devised by a small Nigerian start-up, Ubenwa, aims to fight the recurring problem of birth asphyxia. Nicknamed “Shazam for babies”, the product uses the baby’s cry to diagnose any problems. Detecting birth asphyxia early could save many children’s lives.

PROBLEMS WITH IMPLEMENTATION

Nothing ever comes without obstacles and AI is no different. Although AI has unmissable benefits, it also has the potential to be harmful to the citizens it serves.

A problem with expanding AI on the continent is that there is a paucity of technically skilled professionals on the continent who could be involved in the process of producing AI technology. This would in turn affect the product as expat workers may not fully comprehend the type of AI most needed in African countries. However, Cisse claims that the centre plans to combat this lack of diversity by granting scholarships to Ghanaians interested in AI.

Nyalleng Moorosi, a software developer at Google, also notes that a lack of diversity is a major risk in developing AI technology on the continent. As AI functions on data it has access to, it may be that the technology produced is not representative of its end users. This can also create unintentional biases and build discrimination into AI products. According to Moorosi, this is a human problem that can only be solved by diversifying the samples used to gather data. More local citizens need to be involved in data gathering and the creation of algorithms.

POTENTIAL RISKS

As AI tech relies on personal data, privacy is an important concern. Anti-poverty group, Global Justice Now, has expressed fears that major tech companies have free rein to create a global surveillance state under developmental pretences. However, Moustapha Cisse, the head of Google’s AI lab in Ghana, asserts that exploitation can be avoided if the company’s AI is aligned with the values of the society it serves. Tewodros Abebe, a doctoral student studying language technology, rejects the notion of ‘cyber colonialism’, and warns that “without regulation, people are vulnerable to exploitation.”

Another hurdle to integrating AI is enhancing access to fast broadband connections. Inconsistent distribution of broadband could increase disparities between lower and higher income areas, instead of boosting economies.

SOLUTIONS 

As with all new technology, AI comes with challenges, but there are solutions. Governments should embrace these challenges because the potential of AI is far too large to pass on. To erase fears regarding the misuse of private data, new laws and legal frameworks should be implemented.

Moorosi suggests that including local citizens in data gathering and algorithm could help address issues around diversity. Countries also need to embed digital skill acquisition into education systems. Cisse has consistently advocated for specialised AI education on the continent. He is working to persuade African governments to see AI as a key tool to combat leading development challenges.

Ensuring that everyone receives equal broadband coverage is another important issue, and will need to be addressed immediately. People cannot benefit from these new technological advancements if they have no access to it.

MOVING FORWARD

AI has real potential to help ensure sustainable and inclusive development on the African continent. It is vital that whilst we start to explore AI’s promise, we continue to think about the risks and solutions.

One possible risk is that the development of AI could boost wealthier African countries to the detriment of lower income countries. We need to conduct more research to ensure that everyone will be positively impacted by the integration of AI tech.

Aisha Patel is currently studying for a BA (English) at LIDC associate member, City, University of London.