October 17, 2019
Not only is Africa’s population expected to double to 2.5 billion people by 2050, but the continent will also be home to 40% of the world’s working population. While the projected growth holds promise for the continent’s labor force, leveraging emerging technologies can also significantly impact Africa’s contribution to global development.
“Being in Africa comes with its challenges; accessibility, limited infrastructure and exposure and the like, yet we like to see these challenges as opportunities,” shared Dr. Moustapha Cisse, Head of Google AI in Ghana, in a talk at Ashesi. Visiting campus to share his organization’s work in Africa, Moustapha spoke to students and faculty about the possibilities that lie ahead for Africa.
“We are a mobile-first people, and possibly mobile-only continent – the first time most Africans interact with a computer is through using the cell phone,” he explained. “In many ways, this phenomenon has forced us to change the way we think about deploying technology in answering some of Africa’s most pressing problems. In our work, we deal with research and design that involves creating tools that enable people to build solutions to their own problems.”
Here is a snapshot of insights from his conversation on campus.
Making Technology Multilingual
“One of the exciting opportunities in machine learning is in aiding in understanding languages, including sign languages. Even though a lot of people may speak French, English or German, Africa is by far the most linguistically diverse part of the world. So how do we make sure that the technology we build also accounts for all these languages? It is important because the languages also reflect part of our identity; the way we say things, the way we stress things. We need to create algorithms that should be able to learn these the same way they have learned French, German and English.”
“In sub-Saharan Africa where we have 1 doctor to over 1,000 patients, chances are that you won’t have direct access to a medical doctor to give you immediate and quality treatment if you’re not living in a major city. Can we use the current state of technology to complement the existing infrastructure, so that people who don’t have access to reliable diagnosis can be helped? When diseases are diagnosed early on, we can learn how to better treat them.”
On Fairness and Transparency
It’s good to look at all the possibilities offered by the breakthroughs in machine learning, but it’s even better to pay specific attention to fairness and transparency. What are we building for? What is the intended impact? What is the impact that technology might have? Is the technology we’ve built going to discriminate? Does the tech contain certain biases, or amplify certain biases that already exist in society? These are very important questions to pay attention to – designing algorithms that are fair and transparent. Building this way is value-driven, and it’s important that all the voices are heard, and presented at the table. One of the ways this can be done is to ensure that not only do we get researchers from everywhere, but then also, research is done everywhere.”
In early 2019, Google established the Artificial Intelligence Research Centre in Accra. It was the first on the continent, aimed at meeting the increasing interest in machine learning research across the continent.
The center will collaborate with local universities, research organizations and policymakers in Ghana and across Africa to deploy AI in solving challenges in the healthcare, agriculture, education, and other sectors.