Facebook’s parent company Meta’s “No Language Left Behind” Al project, can now translate 200 different languages in a single mode. The newly engineered Al model is part of the ambitious “Everyone in the World” project.
The “No Language Left Behind” project expects to eliminate language barriers by creating a universal speech creator. The speech creator can edit the information it receives to translate to preferred languages with the help of artificial intelligence to help users understand the information. The project aims to promote connectivity among people across the world. Meta (Was it a spokesperson? The CEO? Be more specific with this quote if at all possible.) says, “The Al tool is not just about giving people access to content on the web in their language, it’s about making the transfer of knowledge and online contributions possible too.”
Mark Zuckerberg, the founder of Meta, said in a Facebook post, “Communicating across languages is one superpower that AI provides, but as we keep advancing our AI work improving everything we do — from showing the most interesting content on Facebook and Instagram, to recommending more relevant ads, to keeping our services safe for everyone.” Meta expects to democratize the virtual world with rare and low-resource languages to add to the list. Meta’s CEO (Again, a spokesperson? The CEO? Who said this?) said the project is expected soon; however, the time frame is not clear yet.
Why Al over Machine Technology?
Machine Translation (MT) relies on textual data, which works for languages with words. Many MT systems are bilingual with separate models for each pair of languages, such as English to Russian or English to French. People with native non-standardized languages that do not have written words are left out of the internet that dominates the information exchange of the world. Meta expects to fill this gap with AI automated direct speech-to-speech translation. If this succeeds, people who speak different languages may communicate in real-time using any device to make a direct speech-to-speech translation inclusive.