Large Language AI Models
Posted: Thu Jul 10, 2025 3:45 am
And although the language is very much alive, very resilient, and widely spoken, it is not commercially viable or interesting for international markets. Such is (or was) the case for Papiamento in a digital sense as well: the smaller the language, the longer it takes for a language to be supported or included in software or online products.
But the tide seems to be changing: Launched in 2019, the National Library of Aruba’s online collection (hosted by the Internet Archive), has grown into a veritable National Collection effort called Coleccion Aruba with over a dozen partner institutions, from Aruba and beyond, providing access to handwritten, printed and audiovisual works in seven languages, including
the largest online text corpus for the whatsapp number database language, spanning over a million digitized and digital-born pages. Using this growing Papiamento text corpus, (LLMs) like ChatGTP can now converse and answer in Papiamento/u, and Papiamento/u is now a supported language in both Meta’s AI-assisted “No Language Left Behind” initiative and Google Translate. And just recently in January 2024, the Council of Europe recognized Papiamento as an official European minority language, after having been officialized in Aruba in 2003 and in Curacao and Bonaire in 2007.
Download the complete Vanishing Culture report.
The advent of artificial intelligence has made quite an impact in the world of documentary heritage, with one of the newest developments being handwritten text recognition (HTR).With new technologies like the AI-supported Transkribus, HTR technologies are becoming available not only to the bigger institutions in wealthy nations, but also to small island institutions such as Biblioteca Nacional Aruba (the Aruba National Library) and Archivo Nacional Aruba (the Aruba National Archives), which adopted this new technology at a relatively early stage in 2019. The ability to add text layers to scanned manuscripts unlocked documentary treasure chests containing centuries of written records, correspondence, and prose—all within the reach of the Aruba institutions.
But the tide seems to be changing: Launched in 2019, the National Library of Aruba’s online collection (hosted by the Internet Archive), has grown into a veritable National Collection effort called Coleccion Aruba with over a dozen partner institutions, from Aruba and beyond, providing access to handwritten, printed and audiovisual works in seven languages, including
the largest online text corpus for the whatsapp number database language, spanning over a million digitized and digital-born pages. Using this growing Papiamento text corpus, (LLMs) like ChatGTP can now converse and answer in Papiamento/u, and Papiamento/u is now a supported language in both Meta’s AI-assisted “No Language Left Behind” initiative and Google Translate. And just recently in January 2024, the Council of Europe recognized Papiamento as an official European minority language, after having been officialized in Aruba in 2003 and in Curacao and Bonaire in 2007.
Download the complete Vanishing Culture report.
The advent of artificial intelligence has made quite an impact in the world of documentary heritage, with one of the newest developments being handwritten text recognition (HTR).With new technologies like the AI-supported Transkribus, HTR technologies are becoming available not only to the bigger institutions in wealthy nations, but also to small island institutions such as Biblioteca Nacional Aruba (the Aruba National Library) and Archivo Nacional Aruba (the Aruba National Archives), which adopted this new technology at a relatively early stage in 2019. The ability to add text layers to scanned manuscripts unlocked documentary treasure chests containing centuries of written records, correspondence, and prose—all within the reach of the Aruba institutions.