US-based internet search giant Google is committed to its digital infrastructure investment and inclusive connectivity priorities for the South African market.
So said Google South Africa country director Kabelo Makwane, in a wide-ranging interview with ITWeb TV this week.
The man at the helm of the search engine’s local operations joined the channel to reflect on his first year in the top job, what the artificial intelligence (AI) boom looks like for the continent, partnering with academia, and where the company is placing its biggest bets locally.
He also spoke about YouTube’s burgeoning local active user base, Equiano gains and whether another Africa-based cloud region is on the cards in the immediate future.
According to Makwane, South Africa remains a significant market as part of Google’s “build for Africa” commitment, with investments in a cloud region as well as submarine cable connectivity.
From the perspective of localisation of cloud technology, the Google Cloud region served as an important step, he stated. “This allows the opportunity to continue to localise a lot of the technologies that we have. Investing and building out local infrastructure is a strategic focus for us over the next three years-or-so.
“The second thing, which is supper important, is connectivity, which is an indispensable part of that infrastructure investment. Since building the subsea cable system – Equiano – and having landed it across the shoreline, we’ve effectively increased the connectivity bandwidth 20-fold and decreased the cost of internet connectivity four-fold. With that, the strategic target is to provide greater access to previously marginalised communities.
“The third element is making sure we move from just being consumers of technology to being participants and innovators in this space, and AI creates that opportunity. For us, particularly what I’m passionate about, is making sure we have innovators and people that can create businesses, with gainful endeavours on the back of these platforms that are also exportable to the rest of the world.”
Google’s first Africa-based cloud region goes head-on with other hyperscalers, such as Amazon Web Services and Microsoft Azure.
The local Google Cloud region directly supports African businesses and multinational corporations operating on the African continent by providing access to technologies such as AI, machine learning, data analytics and security.
Google owns and operates 29 data centres in 11 regions around the world.
Asked if Google has plans for further local data centre or cloud region expansion, Makwane noted the first one has just been launched, asking to give the current cloud region some time.
“As it progresses, we’ll look at opportunities of what that expansion looks like. In the short-term, we have no immediate plans to launch a second region or anything of that sort. But with growth, who knows.”
Google officially unveiled its Equiano submarine internet cable in Cape Town in 2022, in which the tech giant has invested a total of $1 billion (R17 billion). Equiano runs from Portugal, along the west coast of Africa to South Africa. At the time, the company said the Equiano cable will boost South Africa’s gross domestic product by R120 billion.
In 2024, Google announcedUmoja, which it said is the first fibre-optic route to directly connect Africa with Australia.Anchored in Kenya, the Umoja cable route will pass through Uganda, Rwanda, Democratic Republic of the Congo, Zambia, Zimbabwe and South Africa, including the Google Cloud region, before crossing the Indian Ocean to Australia.
Inclusive AI
Google parent company Alphabet is projected to spend between $175 billion to $185 billion on AI in 2026, as it looks to secure dominance against rivals.
The company also previously announced a $2.85 million (R49 million) funding commitment through its philanthropic arm Google.org, to support AI skills development, education, online safety and cyber security in South Africa.
As far as making sure Africa doesn’t get left behind during the AI boom, Makwane said there are a couple of fundamentals that need doubling down to make AI inclusive.
First among these, he stated, is infrastructure, particularly given its importance to support the emergence of these technologies. “We’re talking in terms of connectivity infrastructure, the GPUs [graphics processing units] that run large language models, making sure they are available to the markets that we serve.
“Secondly is the part of localisation. Once we deploy the infrastructure, that which works for international markets like the US, Europe, etc, may not necessarily be immediately replicable, so there is a big portion of it that’s about localisation.
“With AI, this has become front and centre. This is mainly because you are training a conversational language model to understand local context, dialect and culture. Therefore, part of the journey with localisation of AI is making sure we’re localising languages as well.”
Makwane added that Google’s Waxal language project, which combines machine learning, linguistic research and community collaboration to improve how AI systems understand and generate African languages, now incorporates 21 African languages.
“We’ve got over 2 500 different languages and dialects…the task is a mammoth one, but we had to choose the more ‘popularly’ spoken languages like isiZulu, Sesotho, Afrikaans and isiXhosa, for South Africa. Added to this, is Swahili, Yoruba and Ibo, which are some of the languages popularly spoken in the rest of Africa.”
Beyond languages, he noted, is the element of capacity creation to digest and make gainful use of the technology.
“If people don’t understand how to interact with this tech, they obviously won’t be able to use it. So, you’ve essentially created a barrier to access in a different kind of way. It’s important for us to create awareness, training and avenues for people to find those skills.”
Share
