Africa can benefit much more from artificial intelligence
- Having this in mind, Africa needs to develop AI to solve its problems instead of waiting for solutions from abroad. Such problems include traffic congestion, low literacy levels, fake news, speech recognition of native languages, agricultural information, diseases,
By Reuben Orinda
Artificial intelligence (AI) was once something we considered science fiction from movies like The Terminator, Robocop, The Matrix or Star Wars. I lived in this fiction until around 2010 to 2015, where personal assistants like Siri, Google Now, Alex and Cortana were put into use and made me realise AI is real and growing swiftly.
It was during the same period in my early tech enthusiasm years that NEIL was released, Tesla announced its Autopilot, Google launches its open source deep learning framework, TensorFlow was released and Open AI by Elon Musk was founded.
AI Expo Africa, now entering its fifth successful year, is the largest business-focused AI, robotic process automation (RPA) and 4IR trade event in Africa, uniting thousands of buyers, suppliers and innovators across the continent. When I saw the ad regarding this year’s event, its progress immediately captured my attention.
We have seen what tech giants have been doing in this industry, and the heavy investment they have made to improve their services. From search engines, ads recommendation, fraud, voice assistance, process automations, authenticity validations and a lot more. Having this in mind, Africa needs to develop AI to solve its problems instead of waiting for solutions from abroad. Such problems include traffic congestion, low literacy levels, fake news, speech recognition of native languages, agricultural information, diseases, etc.
For instance, I have seen two outstanding projects at the Nelson Mandela African Institution of Science and Technology (NM-AIST) in Arusha, which made use of AI to solve local problems. The first one identified three types of diseases from chicken droppings to determine if the cause if cocci, cecal, or whether it is a normal occurrence. Imagine what farmers could do with this innovation.
Another one was the use of X-ray images to determine if a person is suffering from tuberculosis. TB has been a big challenge in Africa for years, but with this information the result can be obtained quickly and the appropriate treatment prescribed.
Do we have enough data?
If there is something that makes AI projects undertaken by Google, Facebook and others successful it is data. The emergence of big data was a game changer in AI. I have given the above two uses, but the hustle involved was huge as researchers had to collect a lot of images to make their models smart and relevant. For instance, one had to have tens of thousands of images of normal droppings, those with cocci, as well as those with cecal just to make a smart model to understand the differences in order to determine chicken diseases from the droppings as mentioned earlier.
Africans have contributed a massive amount of data and information to tech giants, but we should also be able to help local researchers to get enough data to develop local relevant solutions in agriculture, education, transportation, health and other sectors.
AI applications in Africa have gained considerable media coverage. Several startups in Nigeria and Ghana are addressing shortages of doctors and low medical access in rural Africa. They have begun to use AI to empower doctors and leverage from growing mobile phone ownership as a vehicle for collecting data, improving administrative efficiency, and expanding treatment coverage.
Ilara Health in Kenya has been making health services more accessible and affordable to all people through embedding AI into diagnostics.
ITIKI Drought Prediction tool from South Africa is seeking to combine local knowledge and AI to provide a more effective and affordable drought prediction tool to local African farmers.
Others include the Egyptian startup Swivl that uses AI to coordinate a fleet of private buses, allowing commuters to bypass the country’s often overwhelmed public transit network. Hello Tractor app in Mozambique, which focuses on letting farmers share equipment, makes use of machine learning to predict crop yields. And again in Kenya, banking service M-Shwari relies on AI to review and validate online loan applications, helping it to field entreaties from customers who live far from bank branches.
Bringing together important actors and experts within the field of AI is also crucial, and initiatives such as AdaLabs are building AI labs and networks of young professionals, AI startups, and people involved in AI-driven positive impact. Also, collaborators such as Microsoft, IBM and Google, who have been supporting various African initiatives in hubs and research institutes, need to be recognised
Generally, I would urge tech enthusiasts to consider this as an opportunity. Governments and the people in general should help researchers to obtain data as these solutions will have a positive impact on the economic development of our nations through improved education and health systems, fraud detection in transport, tax, banks, and money laundering, while enhancing agriculture and processing industries without forgetting the disruption it could cause in distribution and supply chains.