Your host in Osaka, Japan, puts on headphones and suddenly hears your words transformed into impeccable Kansai Japanese. Better yet, your response in your native language is perfectly understood.
Thanks to artificial intelligence (AI), neither of us suffers from misunderstandings. What once seemed like science fiction is now marketed as a quick fix for cross-cultural communication.
These AI-based tools will be useful to many people, especially for tourists or in any purely transactional situation, even if fluid automatic interpretation is still in the experimental phase.
Does this mean that learning another language will soon be a thing of the past?
As researchers of computer-assisted language learning and linguistics, we disagree and consider language learning to be vital in other ways. We have dedicated our careers to this field because we firmly believe in the transformative and lasting value of learning and speaking languages beyond the native tongue.
Lessons from past linguistic ‘disruptions’
This is not the first time that a new technology promises massive disruption in language learning.
In recent years, language learning startups like Duolingo have set out to make language acquisition as easy as possible, in part by gamifying learning. While these apps have made learning more accessible to a greater number of people, our research shows that most platforms and apps have failed to fully replicate the inherently social process of language learning.
One thing is clear: the huge popularity of language apps shows that there is still a high demand for language learning, despite the sharp decline in formal education. Duolingo alone had 113.1 million monthly active users worldwide at the end of 2024, 36% more than the previous year. This is approximately ten times the number of students taking languages other than English in American schools.
We recommend: Apple equips its AirPods with translation in English and other languages
The meaning of learning a language
Beyond the numbers, the gold standard of language learning is the ability to follow and participate in a live group conversation.
Since World War II, government departments and educational programs recognized that text-centered methods of grammar and translation were of little use for real-world interaction. Interpersonal conversational competence gradually became the main goal of language classes. While technologies that can be placed in the ear or worn on the face promise to revolutionize interpersonal interaction, their usefulness in such conversations falls on a spectrum.
On the one hand, there are the simple tasks that must be performed when visiting a city where another language is spoken, such as checking out at a hotel, buying a ticket at a kiosk, or finding your way around the city. That is, people from different backgrounds collaborating to achieve a goal: successfully check out, buy a ticket, or reach the famous museum you want to visit. Any combination of languages, gestures, or tools—even AI tools—can be useful in this context. In these cases, when the goal is clear and both parties are patient, shared English or automated interpretation can achieve the goal by avoiding the hard work of language learning.
At the other extreme, identity matters as much as content. Meeting your in-laws, showing up at work, welcoming a delegation, or making a presentation to a skeptical audience involves trust and social capital. Humor, idioms, levels of formality, tone, pacing and body language not only shape what you say, but also who you are.
The effort to learn a language communicates respect, confidence, and the willingness to see the world from another perspective. We believe that language learning is one of the most challenging and rewarding forms of deep work, developing cognitive resilience, empathy, identity and a sense of community in ways that technology struggles to replicate.
The 2003 film “Lost in Translation,” which tells the story of an older American man who falls in love with a much younger American woman, was not about getting lost in the language, but delved into issues of interculturality and self-discovery when interacting with the other.
In fact, increasing mobility due to climate migration, teleworking and retirement abroad increase the need to learn languages, not just translate them. Even those who remain in their place of residence often seek deeper connections through language, as learners with family and historical ties.
Also read: Meta develops an AI system that instantly translates from voice to voice in 36 languages
The limitations of AI
The latest AI technologies, like those used by Apple’s new AirPods to instantly interpret and translate, are undoubtedly powerful tools that will help many people interact with anyone who speaks a different language, in ways that were previously only possible for those who spent a year or two studying it. It’s like having a personal interpreter.
However, relying on interpretation carries hidden costs: distortion of meaning, loss of interactive nuances, and decreased interpersonal trust.
An ethnographic study of highly motivated American students with almost unlimited support revealed that turning to spoken English and using technology to facilitate translation may be easier in the short term, but this harms long-term linguistic and integration goals. Language learners constantly face this trade-off between short-term ease and long-term impact.
Some AI tools help accomplish immediate tasks, and generative AI applications can support acquisition, but they can eliminate the negotiation of meaning from which enduring skills emerge.
AI interpretation may be sufficient for one-on-one conversations, but language learners often aspire to participate in existing conversations between speakers of another language. Long-term language learning, although it involves certain difficulties, is beneficial in many ways.
At the interpersonal level, using another person’s language fosters both cultural and cognitive empathy.
Furthermore, the cognitive benefits of multilingualism are widely documented: resistance to dementia, divergent thinking, flexibility to change focus of attention, acceptance of multiple perspectives and explanations, and reduced bias in reasoning.
Precisely the attributes that companies seek in the age of AI—resilience, lifelong learning, analytical and creative thinking, active listening—are cultivated through language learning.
Rethinking language teaching in the age of AI
So why in the United Kingdom and the United States, increasingly multilingual countries, are fewer and fewer students choosing to learn another language in high school and university?
The reasons are complex.
Too often, institutions have struggled to demonstrate the relevance of language studies. However, innovative approaches abound, from integrating the language into the context of other subjects and linking it with service and volunteering, to connecting students with others through virtual exchanges or with community partners through project-based language learning, all while developing intercultural skills.
So what value is there in learning another language when AI can handle tourist phrases, casual conversations and urban orientation?
The answer, in our view, lies not in fleeting encounters, but in cultivating lasting capacities: curiosity, empathy, a deeper understanding of others, the reconfiguration of identity, and the promise of lifelong cognitive growth.
For educators, the call is clear. Generative AI can take on routine and transactional tasks, while excelling at error correction, information adaptation, and vocabulary support. This frees up classroom time for multi-party, culturally rich and nuanced conversations.
Pedagogical approaches based on interculturality, experiential communication, play and relationship development will thrive. Learning this way allows students to critically evaluate what AI headsets or chatbots create, engage in authentic conversations, and fully experience the benefits of long-term language learning.
*Gabriel Guillen He is a professor of Linguistic Studies and Thor Sawin He is a professor of Linguistics, both at Middlebury College
This text was originally published in The Conversation
Do you like photos and news? Follow us on our Instagram
