Delivering AI services in Africa
The race to roll out AI infrastructure and deliver AI services now reaches extraordinary proportions and geopolitical importance in the USA, Europe and China. Meanwhile we are developing ambitious but realistic delivery models to bring inference AI to Africa.
There is very little uncertainty about whether we will use AI services from now on. While individuals might not want to pay (directly) for it, those who have access to AI generative services can leverage their impressive ability to summarise, translate or generate new text and images. Businesses aim to embed them in their routines and processes, and while there are reports of various degrees of success (in terms of ROI) it is fair to say that AI is here to stay. Global executives dream of what the future could be if AI models turn out to be better at creating businesses, innovation, and value than we have been so far.
To power this demand, we see a rush to build infrastructure to train vast models (latest is about half a trillion parameters) and make them available via inference services. The challenge is there for most of countries, companies and individuals: how to control the data that is used in inference services. Most countries in Africa have now introduced GDPR-insprired data protection rules, meaning that models outside borders should not be provided with public or business data. This means that inference cannot happen outside borders, and AI inference services must be delivered locally.
Digital Africa Development Agency has launched a fast-track, standardised programme for public agencies and private businesses to gain access to AI services locally. This involves focused investments in data centres, computing power, connectivity and service capabilities, to be available in most African countries (over time).
ABOUT DIGITAL AFRICA DEVELOPMENT AGENCY
The Digital Africa Development Agency helps public and private organisations to maximise the impact of digital services on economic development and their stakeholders, by providing expert advice, market insights and operational support. More information at www.digitalafrica.ai