The rise of “artificial progress”

Modern language learning tools are highly effective at keeping users engaged. They reward consistency, simplify content, and create a sense of momentum.

But that momentum can be misleading.

The real challenge of speaking a language

“You can spend hours learning vocabulary and grammar and still struggle when facing a real human.”

Nicolas Potel, Director of Learning Experience at Berlitz

This phenomenon aligns with what cognitive scientists describe as the “illusion of competence”, when familiarity with content is mistaken for true mastery. Learners recognize words, understand structures, and perform well in controlled environments but struggle to apply that knowledge spontaneously.

Berlitz refers to this gap as “artificial progress”: the feeling of learning without the ability to perform in real-world situations.

It’s a gap that becomes particularly visible in high-stakes contexts such as job interviews, client meetings, or negotiations, where communication must be immediate, nuanced, and confident.

The gap between innovation and real learner needs

In many industries, innovation is measured by what’s new.

At Berlitz, it’s measured by what works.

Our only guiding principle is simple: does this improve learning outcomes?

asks Nicolas.

This distinction reflects a broader shift in learning science. Frameworks such as the Kirkpatrick Model have long emphasized that true learning should be measured not by participation or completion, but by behavioral change and real-world application.

And yet, many language learning solutions still optimize for engagement metrics rather than performance outcomes.

Berlitz takes a different approach. Every new feature, tool, or product is rigorously tested through proof of concept, pilot programmes, and measurable outcomes before being scaled. The objective is clear: ensure that innovation translates into demonstrable improvement in communication ability.

Communication over translation: why practice beats passive learning

At the core of the language learning debate lies a fundamental question: should learning prioritize understanding or communication?

Most digital tools rely heavily on translation. It is fast, scalable, and intuitive. But it can also create dependency, limiting the learner’s ability to think and respond directly in the target language.

Berlitz takes a different approach: immersion.

If you want to develop real fluency, you need immersion as it activates the parts of the brain responsible for communication.

says Berlitz’ Director of Learning Experience.

This approach is supported by decades of research in second language acquisition. Theories such as Stephen Krashen’s Input Hypothesis and subsequent neuroscience studies highlight the importance of contextual exposure and active production in building fluency.

Immersion mirrors how we naturally acquire language, through repeated exposure, interaction, and use, rather than translation and memorization.

The role of gamification in language learning

Gamification has played a significant role in making language learning more engaging and accessible.

Used effectively, it can:

  • Reinforce habits
  • Increase motivation
  • Encourage consistent practice

However, its impact depends on how it is applied.

The limits of gamified learning

“Gamification is a great tool to drive engagement. But if overused, learners may start focusing more on the system than on the language itself.”

Nicolas Potel, Director of Learning Experience at Berlitz

Recent research highlights this tension. While gamified systems can increase participation, they may also lead learners to optimize for rewards, such as points, streaks, or levels, rather than for actual language acquisition.

In some cases, learners appear to progress but lack real proficiency because the focus shifts from learning the language to navigating the system.

This does not diminish the value of gamification, but it underscores the importance of balance. Engagement is essential. But without meaningful practice and real-world application, it is not sufficient.

The most effective learning experiences use gamification to support learning, not replace it.

Why speaking is still the metric that matters

At Berlitz, progress is defined differently: if you can’t use the language in a real situation, you haven’t learnt it yet.

Every lesson is built around practical speaking goals, from introducing yourself to leading meetings or negotiating outcomes.

We measure success by whether students can perform real-world tasks,

says Nicolas.

This approach aligns with a key concept in learning science: transfer of learning, the ability to apply knowledge in real-world contexts, not just recall it in structured environments.

It also explains why instructor-led learning remains one of the fastest ways to achieve fluency. According to Berlitz data, learners can progress between CEFR levels in as little as 90 hours in private instruction, compared to widely accepted benchmarks of 150–200 hours.

The difference lies not just in content but in active use, feedback, and real interaction.

AI in language learning: powerful but incomplete

AI is rapidly reshaping the language learning landscape, offering scalable practice, instant feedback, and unprecedented accessibility. These are meaningful advancements.

But AI also has clear limitations, particularly in areas that matter most for real communication.

AI is great to help you practice in realistic situations, but the value of true human interaction is still irreplaceable,

Nicolas points out.

Research and industry analysis increasingly point to the same conclusion: while AI is highly effective for repetition and structured practice, it struggles with cultural nuance, contextual understanding, and emotional intelligence.

Language is not only about correctness; it is about appropriateness, tone, and connection. These are inherently human dimensions.

The human factor: confidence, connection, and real communication

Language learning is not just a cognitive process; it is a social one. People learn languages to connect, exchange ideas, and understand different perspectives. So, ultimately, people learn a language in order to communicate with other people.

Learning through real connection

“When you learn with a human, you’re already doing what the language is meant for.”

Nicolas Potel, Director of Learning Experience at Berlitz

This reflects principles from social learning theory, which emphasize the role of interaction, feedback, and observation in building competence. It also connects to the concept of the affective filter: the emotional barrier that can inhibit language production. Confidence, encouragement, and real interaction help lower that barrier, making communication possible.

In this sense, human interaction is not an add-on to language learning; it is central to it.

Why language learning goes hand in hand with cultural competence

In a world shaped by global collaboration and increasing complexity, language alone is no longer sufficient. Understanding cultural context has become a critical capability.

Learning a language also means understanding the culture behind it,

according to Nicolas.

From communication styles to decision-making norms, intercultural competence plays a central role in effective global collaboration. Research from organizations such as the World Economic Forum and Harvard Business Review consistently highlights cross-cultural communication as a key skill for the future of work.

Berlitz integrates this dimension directly into its learning experience, ensuring that learners are not only linguistically capable but also culturally aware and adaptable.

The future: human + AI, not human vs. AI

The future of language learning isn’t about choosing between technology and humans. At Berlitz, we embrace innovation and continuously evolve with it, but we also know that real progress comes from human interaction. Technology can support the journey, but it can’t replace what truly builds fluency: real conversations in class that prepare you to use the language on the spot, whether buying a ticket, asking for directions, or leading a meeting at work.

Across industries, research from organizations like McKinsey and the World Economic Forum shows that the most effective use of AI is augmentation, not substitution.

The same applies to language learning:

  • AI enables scale, repetition, and accessibility.
  • Humans provide context, nuance, and confidence.
Technology should support the method, not replace the human interaction at its core.

This balanced approach is where Berlitz is focusing its innovation efforts.

Conclusion: from progress to performance

The language learning industry is evolving rapidly. But if innovation continues to prioritize engagement over outcomes, the gap between learning and real-world ability will persist.

Berlitz is taking a different approach, one grounded in learning science, human interaction, and measurable outcomes. Because in the end, language learning is not about what you complete. It is about what you can actually do.