Meta’s new AI model brings translated content to African language speakers

Sub-Saharan Africa is the world’s fastest growing mobile economy, but millions of internet users are still unable to access content in their first language. Meta’s new "No Language Left Behind" project seeks to change that.

By

Image : Confidence / Adobe Stock

Facebook’s parent company, Meta, has announced the launch of “No Language Left Behind” (NLLB), an AI program which will make digital content more accessible across 55 African languages spoken by more than half of the continent’s internet users.

Meta claims that the technology will provide state-of-the-art translations between over 200 languages worldwide – more than double the current number.

Meta CEO Mark Zuckerberg announced that the new software “will enable more than 25 billion translations every day across our apps”, bringing content from Facebook, Instagram and Wikipedia to millions of people in their native languages for the first time.

“In the future, imagine visiting your favourite Facebook group, coming across a post in Igbo or Luganda, and being able to understand it in your own language with just a click of a button”, said Balkissa Ide Siddo, Facebook’s public policy director for Africa, at the unveiling of NLLB.

Going forward, Siddo said that “accurate translations in more languages could also help to spot harmful content and misinformation, protect election integrity, and curb instances of online sexual exploitation and human trafficking”.

Moreover, Meta says that it aims to set a “new standard of inclusion” as it helps to build the metaverse, holding out the prospect of everyone having “access to virtual-world content, devices and experiences, with the ability to communicate with anyone, in any language” there.

Boost for “low-resource” languages

Language barriers prevent many Africans from either accessing or producing their own digital content – while 130m, 115m and 35m Africans speak English, French and Portuguese respectively, those who solely speak “low-resource” languages are unable to use social media to expand businesses into new markets.

Meta’s report defines a “low-resource” language as one which has produced negligible amounts of published content on the internet. Of the 55 African languages surveyed, just seven escaped this categorisation: Afrikaans, Modern Standard Arabic, Sotho, Swahili, Xhosa, Tswana and Zulu.

Widely spoken languages like Yoruba (45m), Igbo (30m) and Fula (35m) are defined as “low-resource”, yielding poor accuracy scores in existing automated translations.

Facebook boosts African reach

Meta, formerly Facebook, has long had a close association with the rollout of internet services in Africa. Its Free Basics app, launched in 2013 as Internet.org, brings basic internet services to mobile phone users in a simplified, highly-accessible format and now boasts 100m users worldwide, many of them in Africa.

Language barriers heighten the disparity between urban and rural internet penetration in sub-Saharan Africa. In West Africa, where 38.1% of urban residents have access to mobile broadband but just 7.1% in rural areas are connected, the prevalence of traditional languages outside urban centres means that rural citizens are less likely to engage with digital content even if they are able to connect to the internet.

Want to continue reading? Subscribe today.

You've read all your free articles for this month! Subscribe now to enjoy full access to our content.

Digital Monthly

£8.00 / month

Receive full unlimited access to our articles, opinions, podcasts and more.

Digital Yearly

£70.00 / year

Our best value offer - save £26 and gain access to all of our digital content for an entire year!