Ghana-based Google Artificial Intelligence (AI) Centre Accra – Google’s latest AI lab – has brought together top machine learning researchers and engineers dedicated to AI research and its applications.

The centre is also a central role-player in pursuing ‘AI by Africa, for the world’, which forms part of the vision to use AI to solve challenges in areas such as healthcare, agriculture and education, besides others, for “everyone, in every part of the world”.

Google AI Centre Accra forms part of the company’s structured efforts to explore and integrate more diverse experiences and learnings beyond present-day centres of innovation.

The team at Google AI has already started working on building AI-powered solutions to real-world problems, including helping communities in Africa and beyond to improve their lives.

The researchers of Google AI Centre Accra bring a fresh perspective and expertise to build new technologies in Africa that can contribute positively globally, Google says.

“In order to build technology that benefits people everywhere, it needs to be built by people with a diverse range of backgrounds, experiences and viewpoints.”

The new centre is headed up by Moustapha Cisse.

There is an increasing interest seen across the continent in machine learning (ML) research, along with the continuing growth of the computer science research community in Africa.

Google uses ML and AI in all its products, based on seven guiding principles for ethical use of AI and ML to help nurture an emerging technology.

“A strong focus area for Google is how AI and ML can be used for social good. We already see how ML is improving people’s lives – from protecting us all from spam and fraud to making devices more accessible through speech.”

Google’s AI for Social Good programme includes projects such as the development of a system that combines physics-based modelling with AI to produce earlier and more precise flood warnings and the application of AI to seismic data for the creation of a model that, while far from being fully accurate yet, is better than many other models at predicting where aftershocks will occur.