Experts aim to close the language difference

Although Africa is home to a huge part of the world languages ​​- much over a quarter according to some estimates – many are missing when it comes to the development of AI.

This is both a matter of lack of investment and easily available data.

Most AI tools, such as the GPT chat used today, are trained in English as well as other European and Chinese languages.

They have huge quantities of online text to draw.

But since many African languages ​​are spoken in most, not recorded, there is no text to train AI to make it useful for speakers in these languages.

For millions of the whole continent, it means being abandoned.

Researchers who are trying to deal with this issue have recently released what is considered to be the largest known data set of African languages.

“We believe that in our own languages ​​we dream of them and interpret the world through them. If technology does not reflect this, the whole group risks being abandoned,” says Prof. Vukosi Mariva at the University of Pretoria, who worked on the project.

“We are going through this AI revolution, imagining everything that can be done with it. Now imagine that there is a part of the population that simply does not have this access because all the information is in English.”

The Africa Next Voices project has collected linguists and computer scientists to create AI-finished data sets in 18 African languages.

This can be just a small part of more than 2000 languages ​​that are calculated on the entire continent, but those involved in the project say they hope to expand in the future.

For two years, the team has registered 9,000 hours of speech in Kenya, Nigeria and South Africa, capturing daily scenarios in agriculture, health and education.

The recorded languages ​​included Kikuyu and Dhluo in Kenya, House and Joruba in Nigeria and Isizulu and Tshiven in South Africa, some of which are spoken by millions of people.

“You need some foundation to get started with and this is what Africa is the next votes and then people will build on everything from above and add their own innovations,” says Prof. Mariete, who runs research in South Africa.

His Kenyan counterpart, the calculation linguist Lillian Vanzare, says that recording the continent’s speech means creating data aimed at reflecting how people really live and talk.

“We have collected votes from different regions, ages and origin, so it is possible as possible. Great technologies cannot always see these nuances,” she says.

The project became possible with a grant of $ 2.2 million ($ 1.6 million).

The data will be open access that will allow developers to build tools that translate, transcribe and respond to African languages.

There are already small examples of how the root languages ​​used in AI can be used to solve real -life challenges in Africa, according to Prof. Mariva.

Farmer Kelebgil Mosime manages a place with 21 hectares in Rustenburg, the heart of the platinum region of South Africa.

The 45-year-old works with a small team for growing a row of vegetables-influential beans, spinach, cauliflower and tomatoes.

It started just three years ago, with a cabbage crop and to help use an application called AI-Farmer, which recognizes several South African languages, including Sesotho, Isizulu and Afrikaans, to help solve various problems.

“Because someone is still learning to farm, you are faced with many challenges,” says G -Jen Mosime.

“I see every day the benefits of using my homemade Setswana in the app when I encounter a farm problems, I ask something and get a useful answer.

“For someone in rural areas like me that is not exposed to technology, it is useful. I can ask for various insect control options, it is also useful in diagnosing diseased plants,” it is a broad manual sunhat.

Lelapa AI is a young South African company that builds AI tools in African languages ​​for banks and telecommunications companies.

For its CEO Pelonomi Moiloa, what is currently offering is very limited.

“English is the language of opportunities. For many South Africans who do not speak it, it is not just uncomfortable – it can mean missing out on basic services such as healthcare, banking or even state support,” she told the BBC.

“The language can be a huge barrier. We say it shouldn’t be.”

But this is more than doing business and convenience.

For Prof. Mariva, there is also a danger that without African initiatives, something else can be lost

“Language is access to imagination,” he says.

“These are not just words – it’s history, culture, knowledge. If the root languages ​​are not included, we lose more than data; we lose ways to see and understand the world.”