Like other small languages, Faroese struggles to establish its presence in the digital world. Globalization and the development of large language models in major world languages place increasing pressure on smaller languages. This makes it essential to identify solutions that ensure Faroese not only survives but also thrives in this era of advanced artificial intelligence. Over the past year, various initiatives have explored these challenges and addressed how they should be managed. As the year draws to a close, we reflect on some of these efforts and activities, many of which revolve around Nordic collaboration. At the Centre for Language Technology, we have actively participated in several of these initiatives and look forward to seeing how this work will shape priorities and activities in 2025.
This year, the West Nordic Council recommended enhanced language technology collaboration between the Faroe Islands, Iceland, and Greenland. The languages of the West Nordic region face similar challenges: limited linguistic resources for training language models, a shortage of expertise in the field, and insufficient public investment. Collaboration between these countries could help advance solutions for all involved. By pooling investments and expertise, we can achieve more with fewer resources and collectively strengthen our languages. Staff from the Centre for Language Technology delivered a presentation to the West Nordic Council and provided advice on these recommendations, which can be accessed here and here.
In the same spirit, NORA organized a workshop aimed at producing recommendations on advancing artificial intelligence in the North Atlantic region. Experts from Denmark, Norway, Iceland, the Faroe Islands, and Greenland gathered to work intensively on this issue. The Centre for Language Technology participated in the working group, and the primary overarching recommendation was to prioritize language and culture in artificial intelligence. Language has become the interface for the average user to interact with AI, serving as the cornerstone for further development. The outcome of this workshop is the publication Aligning Efforts: Artificial Intelligence in the North Atlantic Region.
Íslendska tjóðin hevur í fleiri ár gingið á odda við at fíggja stórar máltøkniverkætlanir, íslendskum máli at gagni. Hetta framskygnið hevur m.a. havt við sær samstarv við OpenAI, soleiðis at íslendskt er væl umboðað í GPT-málmyndlunum, og lokala, fakliga umhvørvið er ment nógv. Hetta prógvar, at nógv kann fáast burturúr, tá ið ein tjóð setur sær fyri at verja málsligu framtíð sína. Íslendska Menta- og vinnumálaráðið gav í 2024 út eitt skjal, ið lýsir hesa góðu gongd og hevur uppskot til, hvussu onnur smærri mál kunnu nýta sama leist til frama fyri mentanarligum og málsligum fjølbroytni í vitlíki. Skjalið The Icelandic Approach: Preserving and Revitalizing Linguistic and Cultural Diversity in AI kann lesast here.
In 2024, the Nordic Council also issued recommendations on ethical and responsible approaches to AI and the development of Nordic languages. The recommendations emphasize ethical considerations, diversity, sovereignty, sustainability, and transparency. These topics are of great importance, and by working according to shared principles, Nordic countries can facilitate cross-border collaborations. The recommendations and the document Policy Recommendations for Ethical and Responsible AI can be accessed here.
In November, staff from the Centre for Language Technology attended a conference in Trondheim (Language technologies and AI in the Nordic Countries – NTNU), where various topics related to Nordic languages, technology, and AI were discussed. Researchers, developers, and language policymakers from across the Nordic region participated. Despite significant differences in the size of the countries and their languages, one key conclusion was that collaboration between countries can help address challenges and benefit all involved.
A good starting point would be to initiate Nordic collaborative projects that develop linguistic resources, create benchmarks, and secure joint funding. It is also crucial to collaborate on expertise, leveraging each other's knowledge and experiences. Through this approach, we can pool expertise, technology, and data to ensure that Faroese gains a stronger position in the digital world. The project The Faroese MegaWord Corpus is an example of such an initiative, where work on Faroese can build on the experiences and technology of the Icelandic Gigaword Corpus.
To underline the pressing need for prioritizing and advancing language technology and AI, we refer to the open letter sent to the Faroese government, calling for a political stance and commitment to AI. The Centre for Language Technology and other stakeholders initiated the letter, which was signed by leading figures from all public sectors. The open letter can be found here.
All these initiatives, documents, recommendations, and activities highlight the importance of what we are working on—something that everyone has a stake in. It is essential to continue developing and prioritizing Faroese language technology and AI in the coming year. Collaboration among countries, institutions, researchers, policymakers, businesses, and organizations can help ensure that Faroese continues to develop and benefit from the progress already evident in other small languages like Icelandic. Together, we can work to ensure that Faroese is not only preserved but also becomes an active part of the digital future.
With gratitude for fruitful collaborations in 2024 and hopes for continued opportunities and progress in 2025!
