Certificazione DIN EN ISO 17100:2015-05 · TÜV DIN CERTCO Reg. 7U385
Membro Certificato ProZ.com Translation Network
Home › Blog e Analisi › Google Translate Adds 110 New Languages Using the PaLM 2 Large Language Model

Google Translate Adds 110 New Languages Using the PaLM 2 Large Language Model

On June 27, 2024, Google announced its largest-ever expansion of Google Translate, adding 110 new languages spoken by more than 614 million people worldwide—including Cantonese, Sicilian, Lombard, Venetan, Afar, and Tok Pisin—powered by the PaLM 2 LLM.

614 Million Speakers Gain Digital Translation Access

On June 27, 2024, Google unveiled the largest single expansion in the history of Google Translate, adding support for 110 new languages as part of its '1,000 Languages Initiative.' Collectively spoken by more than 614 million people—approximately 8% of the world's population—the new additions range from major global languages like Cantonese (long requested due to its distinct vocabulary and writing conventions from Mandarin) to European regional languages including Sicilian, Lombard, Venetan, Ligurian, Friulian, Occitan, and Breton, as well as African and Pacific languages such as Afar, Wolof, Tamazight, and Tok Pisin.

Google Translate 110 New Languages PaLM 2 – LANGUAGE INCLUSION
LANGUAGE INCLUSION — Professional Localization & Regulatory Quality Assurance (June 27, 2024)

How PaLM 2 Learned Closely Related Regional Languages

Research scientists at Google explained that the PaLM 2 Large Language Model played a pivotal role in training low-resource languages by leveraging cross-lingual transfer from closely related linguistic families—such as Romance macro-varieties or Indo-Aryan dialects—while partnering with native speakers and linguists to handle regional orthographies and dialect variation.

Executive Takeaways for Global Enterprises

  • Key Insight: Google Translate added 110 new languages on June 27, 2024, reaching over 614 million speakers worldwide.
  • Key Insight: Notable additions include Cantonese, Sicilian, Lombard, Venetan, Ligurian, Afar, Tamazight, and Tok Pisin.
  • Key Insight: Cross-lingual transfer in PaLM 2 enabled rapid learning across related regional and low-resource language families.

Nota Editoriale e di Riferimento: Questa analisi tecnica è stata elaborata in modo indipendente dal dipartimento di ricerca di Alphatrad Worldwide per fini di approfondimento professionale e conformità settoriale. Per la documentazione ufficiale e gli atti di riferimento, consultare:

Fonte e Documentazione Ufficiale: Google DeepMind & AI Research (https://blog.google/technology/ai/google-gemini-ai/)

Avete bisogno di traduzioni multilingue certificate, asseverazioni legali o localizzazione conforme agli standard ISO per i vostri progetti internazionali? Richiedete un preventivo gratuito personalizzato ad Alphatrad e riceverete un piano operativo dettagliato entro 24 ore.

← Ritorna alla lettura di tutti gli articoli Richiedi un Preventivo Gratuito →