NEW YORK, June 29: Google has unveiled its largest-ever expansion of the Translate platform, adding 110 new languages made possible by advanced artificial intelligence (AI) models, the tech giant announced on Thursday. This significant enhancement leverages Google's PaLM 2 large language model (LLM) to extend translation capabilities to a broader global audience.

Isaac Caswell, senior software engineer for Google Translate, highlighted the diversity and scope of the new additions. "From Cantonese to Q'eqchi', these new languages represent more than 614 million speakers, opening up translations for around 8% of the world's population," Caswell noted in a release.

The expansion encompasses a wide range of languages, including several major world languages with over 100 million speakers, as well as languages spoken by smaller indigenous communities. Approximately a quarter of the newly added languages originate from Africa, marking Google's most extensive inclusion of African languages to date. Among these are Fon, Kikongo, Luo, Ga, Swati, Venda, and Wolof.

Caswell emphasized the significance of adding Cantonese to the platform, a highly requested language that posed challenges due to its overlap with Mandarin in written form. Similarly, Shahmukhi, a variant of Punjabi widely spoken in Pakistan, and Afar, a tonal language spoken in parts of East Africa, were also notable inclusions due to their community contributions and cultural importance.

The integration of Manx, a Celtic language once nearing extinction, reflects a successful revitalization effort on the Isle of Man, where thousands of speakers now keep the language alive. Additionally, Tok Pisin, the creole language of Papua New Guinea, has been added to Google Translate, providing English speakers a unique opportunity to engage with its English-based structure.

Powered by the PaLM 2 LLM, these AI-driven translations mark a significant advancement in language accessibility and connectivity globally. Google aims to continue enhancing Translate's capabilities to better serve diverse linguistic needs worldwide.