Speaking is considered the pinnacle of learning a new language and a skill that traditional language learning apps like Duolingo or Babbel struggled to teach effectively. Recently, startups have shown how this issue might be addressed: with speaking trainers based on generative AI. This development has prompted major industry players to take action.
In late September, Duolingo introduced a feature in its educational app called “Video Call with Lilli,” an AI speaking trainer in the form of an animated character that helps users practice speaking in a foreign language. This feature is available as part of a paid “Max” subscription. Shortly after, competitor Babbel announced plans to introduce a similar feature by the end of 2024. Although Babbel has not yet implemented this plan, it has expanded the beta version available to all its customers.
The feature, originally named “Conversation Coach,” has been renamed “AI Conversation Partner.” It is now accessible not only on Android smartphones but also on iPhones. The beta is still available only for English as the base language. Initially, Spanish was the only target language, but now French, Italian, and German are also available. However, Babbel has significantly reduced the content: where users once had access to seven role-playing scenarios for English to Spanish, covering topics like greetings, experiences, plans, and shopping, now only two scenarios are available, both at the A1 level.
The free chat option, previously accessible only with Latin American Spanish as the target language, has been removed. Users can still have the AI alternate in continuing a story or guessing a term. Unlike Duolingo’s Lilli, Babbel’s Conversation Coach displays a transcription of its output. If users do not understand, they can have the sentence repeated and even display a translation. If still confused, Babbel’s AI can suggest possible responses, including translations, and read them aloud.
For voice input, users press a button and speak their sentences. A transcription is provided, and users can replay the recording, manually edit the input via keyboard, or re-record their response. However, in the beta version, the AI often corrected the input automatically, resulting in a correct transcription, which defeats the purpose of learning. Babbel should address this issue before the final release of the Conversation Coach.
Overall, the introduction of AI-based speaking trainers marks a significant advancement in language learning apps. These tools aim to provide users with a more interactive and engaging way to practice speaking, which is crucial for mastering a new language. As technology evolves, these features are expected to become more sophisticated and widely available, offering learners a more comprehensive language learning experience.
With the integration of AI, language learning apps are moving beyond traditional methods, offering personalized and adaptive learning experiences. The ability to practice speaking with AI trainers provides users with immediate feedback and a safe environment to make mistakes and improve. This approach not only enhances language proficiency but also boosts confidence in speaking a new language.
As Babbel and Duolingo continue to develop and refine these features, the future of language learning looks promising. The use of AI in education, particularly in language acquisition, is likely to become more prevalent, providing learners with innovative tools to achieve their language goals.