The update, powered by AI PaLM 2, covers languages spoken by 614 million people.
O Google Translator, the technology giant’s translation service, has just undergone the biggest expansion in its history, adding 110 new languages to its catalog. Before this update, the service supported 133 languages, which means the number of languages has almost doubled, now reaching a total of 243.
This significant expansion was enabled by the PaLM 2 AI language model, Gemini’s predecessor, which played a crucial role in learning these new languages. Google software engineer Isaac Caswell highlighted that the model was particularly effective in learning related languages, such as those close to Hindi (Awadhi and Marwadi) and French creoles (Seychellois Creole and Mauritian Creole).
Among the new supported languages, Cantonese stands out, one of the most requested languages by Google Translate users. Caswell explained that Cantonese presents unique challenges due to its overlap with Mandarin in writing, making it difficult to obtain data and train language models.
The addition of these 110 new languages significantly expands the reach of Google Translate, covering languages spoken by more than 614 million people, which corresponds to approximately 8% of the world’s population. This expansion includes both languages with millions of speakers and languages from small indigenous communities and languages in the process of revitalization.
Among the new languages, Afar stands out, spoken in Djibouti, Eritrea and Ethiopia, which received the largest voluntary contribution from the community for its inclusion. Another highlight is Manx, a Celtic language from the Isle of Man, which almost became extinct in 1974, but is being revitalized with thousands of new speakers. N’Ko, a standardized form of the Mandingo languages of West Africa, with a unique alphabet invented in 1949, was also included. In addition to these, there is Punjabi (Shahmukhi), a variation of Punjabi written in Perso-Arabic script, being the most spoken language in Pakistan, and Tamazight (Amazigh), a Berber language spoken throughout North Africa, written in Latin alphabets. and Tifinagh. Tok Pisin, an English-based creole and the lingua franca of Papua New Guinea, completes the list of highlighted languages.
*With information from Olhar Digital
Follow ADNews on Instagram e LinkedIn.
#ADDSomethingNew #ADNewsOficial #ADGroupBR
