The Continued Exploration in AI for Language Learning

In my second blog about user experience with AI for personalized language learning, I discussed some issues about the feedback or lack thereof that I have received. Since that blog, I have been using Duolingo Max, which I got a free trial for. Since starting this free trial, I have received better feedback and more personalized lessons. For example, I am now able to get explanations for my mistakes, and it explains the grammar rules so I understand not to make them again. In some of my research on using Duolingo, some users expressed frustration with the cost of either Super Duolingo or Duolingo Max (Anderson & Charaf, 2020). If I did not have a free trial, Duolingo would cost around $20 a month, and I am unsure if it would be worth it. While there is more personalization in that I get feedback and can roleplay and practice previous mistakes, $20 a month can be costly, depending on various financial situations. I like using ChatGPT, as I can get more personalization and feedback for free.

As I continued to use ChatGPT, I added a microphone extension. Similarly to role-playing through Duolingo Max, I can decide on a theme, prompt the bot to converse with it, and then receive feedback based on my pronunciation. As of yet, I have not had issues with speech recognition that some articles have discussed their students having experienced (Godwin-Jones, 2024). This could be due to some inherent bias that the machine has for countries that have different accents than North America.

If I compare these two AI-powered tools for personalized language education, I am still unsure which one is more advantageous. On one hand, you have Duolingo, which is more gamified and constantly sends out reminders, so you keep practising (Shortt et al., 2021). On the other hand, you have ChatGPT, where you are more in charge of your learning. Thus, to answer my previous question on how effective either one of these tools is, it depends on what one is looking for.


References

Anderson, P., & Charaf, A. (2020). The reviews of users of the duolingo application: Usability and objectivity in the learning process. International Journal of Research – Granthaalayah. https://doi.org/10.29121/granthaalayah.v8.i9.2020.1326

Godwin-Jones, R. (2024). Distributed agency in second language learning and teaching through generative ai. ArXiv (Cornell University). https://doi.org/10.48550/arxiv.2403.20216

Shortt, M., Tilak, S., Kuznetcova, I., Martens, B., & Akinkuolie, B. (2021). Gamification in mobile-assisted language learning: A systematic review of duolingo literature from public release of 2012 to early 2020. Computer Assisted Language Learning, 36(3), 1–38. https://doi.org/10.1080/09588221.2021.1933540

2 thoughts to “The Continued Exploration in AI for Language Learning”

  1. Interesting thoughts on these personalized language learning options Catherine! I am a Duolingo user (although the Owl is mad at me right now, I haven’t been back in a couple of weeks) but I have never used Chat GPT for language learning before. That’s really interesting!

    What do you think the teacher’s role should be in helping students to choose different AI-enabled technologies? Should the teacher be the curator of these technologies, presenting options or comparing the pros-and-cons of each for students? Or, should the teacher be responsible for conducting more needs assessment and recommending options to the individual students? As a third option is this the opportunity for inquiry-based learning, letting students investigate options on their own and present back their findings? I’m sure your answer would also depend on the age group the teacher is supporting. I’m interested to know your thoughts.

    1. Hi Andrea,

      Thanks for your comment and I too am familiar with the Owl being mad at me (I am currently replying while receiving notifications from the Owl not to lose my streak). To answer your question, I think that the teacher’s role for using AI technologies should be as an additional learning aid and not the full lesson. As Nushi and Eqbali (2017) pointed out, even Duolingo promotes using peers as way to collaborate with each other through helping with corrections and completing challenges together. As for ChatGPT, Goodwin-Jones (2024) also discussed how with the way the AI machines are being built they too can be considered ‘peers’ to help the learning process.

      Keeping the idea of language learning being a collaborative process, a teacher may use these tools to help with students to have someone to bounce their knowledge off of. My experience with using ChatGPT, I think with young students (or even adults) it would be helpful for the teachers to provide a guide with prompts that may be useful for students to get the best results to help with their language studies. The teacher should definitely be presenting the pros and cons of each technology and especially keeping in mind the data that is being collected and being used, and the biases that may be seen when using these technologies (Goodwin-Jones, 2024). With younger students, I think the teacher should be more heavily involved in the learning process and conducting the needs assessment.

      For the third option, I have enjoyed my time with experimenting language learning through both these tools. I think this option would be best suited for secondary school students to higher education and it would also be helpful to have a basic understanding of the language you want to learn if using ChatGPT. If using Duolingo then it would be fine if there is no prior knowledge of the language as it starts with the basics.

      While these are all good options, you are correct in saying that it depends on the age group!

      Cheers,
      Catherine

      References
      Godwin-Jones, R. (2024). Distributed agency in second language learning and teaching through generative ai. ArXiv (Cornell University). https://doi.org/10.48550/arxiv.2403.20216

      Nushi, M., & Eqbali, M. H. (2017). Duolingo: A mobile application to assist second language learning. Teaching English with Technology, 17(1), 89–98. https://eric.ed.gov/?id=EJ1135889

Leave a Reply

Your email address will not be published. Required fields are marked *