"Adapting" to AI, preserving culture.

An alarming warning. Once real-time translation devices like AI-powered headphones or smart glasses allow users to communicate directly in their native language by instantly converting both speech and text, a question will inevitably arise: Is it really necessary to spend money and time learning foreign languages?

Furthermore, experts warn that this convenience could diminish the motivation to learn foreign languages, thereby accelerating the simplification of grammar and shrinking vocabulary in many languages. They also predict that the ability to connect the masses, like English, will become dominant, and forecast that by 2100, the number of languages still in use could decrease to around 600, compared to over 7,000 today.

According to UNESCO, nearly 3,000 languages worldwide are at risk of disappearing, while more than 200 languages have not been spoken since 1950 due to a lack of speakers.

Besides the worry that thousands of languages are gradually disappearing from this Earth, there is another, even more worrying, concern: the loss of language " sovereignty ".

As a component of culture, language is not only a tool for communication but also the "soul," containing knowledge, culture, a mirror reflecting and preserving the thinking, identity, and indigenous knowledge of a nation. Language embodies values, symbols, social norms, and expresses all the activities of a community and a nation.

However, since AI is primarily trained using English data, there is a high risk of missing or misinterpreting local contexts. This directly impacts efficiency in fields such as customer service, finance, and public services – where accuracy and cultural appropriateness are crucial.

Therefore, the concept of "sovereign AI" is emerging as a significant trend. Countries are increasingly focusing on developing and controlling AI systems that are appropriate to their legal, linguistic, cultural, and data contexts. This is especially important when AI handles large amounts of sensitive data in sectors such as culture, healthcare , finance, and public services.

A prime example is India, a country that has successfully built a digital ecosystem with platforms such as Aadhaar (digital identity) and UPI (electronic payments). By adopting open standards and emphasizing multilingualism, India has created a flexible technological environment that effectively serves a society with diverse languages and cultures.

Lessons from India show that the future of AI lies not in focusing on a few global models, but in its ability to adapt to specific regions. A sustainable AI ecosystem must ensure transparency, protect data, and build user trust.

In the era of globalization 4.0, the world is flat, but unfortunately, language, writing, and more broadly, culture, cannot be "flat." Multilingualism is the foundation of cultural diversity, a vital policy that UNESCO has long pursued. But what would happen if countries did the opposite, adjusting their language, culture, and data to fit the universal AI ecosystem?

Drawing from India's success, the challenge facing many other countries remains: How to adopt AI systems that are both global and reflect national characteristics? In other words, how to transform AI from a universal technology into a flexible, multilingual, and multicultural system?

Apparently, the problem is not easy, but it must be tackled.

Source: https://baovanhoa.vn/the-gioi/tiep-bien-ai-bao-ton-van-hoa-225867.html