Artificial intelligence can sort out most of Africa's problems
SEE ALSO :Team makes case for blockchain useThis has, in turn, disrupted lending from the traditional formal banking system. Mobile lending platforms and applications now account for the majority of loan accounts and value of new personal loans per capita in Kenya. Still, AI and the breadth of possibilities that could arise from machine learning technologies remain largely unknown to both experts and everyday users. Earlier this year, tech giant Google officially opened its AI Research Centre in Ghana’s capital Accra. The research lab is a first in Africa and brings together the continent’s leading researchers and engineers to explore potential uses of AI in Africa. The Financial Standard was recently invited to an extensive tour of the facility where researchers drawn from both the academia and giant tech firms are working to ensure Africa is not left behind the AI revolution.
SEE ALSO :Artificial Intelligence 101“Most of the people working in advancing the science and developing the field of AI are based mostly in western countries,” explained Moustapha Cisse, Staff Research Scientist, and head of Google AI Centre, Accra. “For such a crucial field, it is important the diversity of the problems and challenges the world faces are represented by the people who build this technology and advance the science,” he said. A study released earlier this month by the AI Now Institute found that more than 80 per cent of AI professors are male and that women comprise only 15 per cent of AI research staff at Facebook and 10 per cent at Google with the picture worse for black scientists. “For example, only 2.5 per cent of Google’s workforce is black, while Facebook and Microsoft are each at four per cent,” said the report in part. “Given decades of concern and investment to redress this imbalance, the current state of the field is alarming.” Scientists at the Google AI Lab in Accra say the facility works to ensure AI solutions developed are relevant to the realities and challenges unique to Africa and that the continent is not left behind in the global AI revolution. One of the projects that researchers at the lab are currently involved in, is the field of deep learning, the re-wiring of artificial neural networks that are central to developing autonomous systems used in driverless cars for example. The full exploration of deep learning requires the use of powerful machines that are inaccessible to much of the African populace. “Improving access to deep learning means designing learning machines that are as accurate and as compact as possible so we can deploy them on devices like the phones we have,” explains Moustapha. Nyalleng Moorosi, a software engineer from Lesotho who works on developing research at the lab says AI applications have already been successfully deployed in the African setting and now need concerted effort to scale the benefits. Ms Moorosi was part of a team that worked with park rangers and communities around South Africa’s Kruger National Park to apply AI to predict rhino poaching trends and help prevent them. “The park is vast and patrol resources are scarce, so the authorities approached us for a solution,” she explains. Ms Moorosi used several data sets including the park’s historical data, weather, and geographical data as well as behavioural patterns of the animals to build a probability map with a grid that showed where poachers would likely be found. The grid changed each day as new data was processed. Another innovative solution allows farmers to use their smartphones to identify diseases in cassava plants. The application called Nuru is built from training machines to identify diseases by feeding it thousands of images of cassava plant images. In one demonstration, farmers scan their phone on top of cassava leaves using the app and if a disease is detected, the app would identify it and give options on the best ways to manage it.
We are undertaking a survey to help us improve our content for you. This will only take 1 minute of your time, please give us your feedback by clicking HERE. All responses will be confidential.