REFLECTIONS
Looking back, what could have been done differently?
Originally, I set out to understand if AI could effectively and efficiently teach a foreign language to adults.
Through my research, I think the answer to that is not on its own, but maybe it can be used as a tool to supplement other types of instruction or learning. My reason for believing this is because, through the applications that are available on the market right now, none of them effectively give users all of the elements they need to learn a language (input, output, and feedback).
Although my design proposal for Lingua Lore attempts to fix the issues identified in other language learning technologies, it is probably not a viable solution at the moment. Before this idea could move forward, more exploration around the design concept itself and the technology needed to make it work would need to happen. However, there was promise for this idea due to the high amount of interest expressed by many people for this idea.
Limitations
Limitations in technology made it difficult to take this idea beyond the prototype phase. If someone were to try and build this in the future, they would need to solve for how to create a generative AI narrative game that gives users freedom of choice and immerses them in the learning environment with custom imagery. This is on top of other technologies like speech recognition in other languages, accurate translation, and storing user progress.
Input from linguistic professionals would also be necessary to ensure that whatever solution is created helps users to learn.
In short, this is an ambitious project that would require experts in many fields to bring to life.
The closest I got to creating something similar to Ligua Lore was with ChatGPT’s “Create Your Own GPT”. There are many issues with the gameplay, but you can play it here.