This is impressive and inspiring. Not to detract at all from his dedication, but it really helps that there are so many content/resources in the target language: news, kids shows, anime, tutors, emigrant diaspora. As a side quest, look at another reputedly-hard language like Vietnamese, where there is not nearly so much. As an example, Google and Microsoft Translator apps speak different variants of Vietnamese (North…
You should try any of the SOTA LLMs (4o, Claude, Gemini) for translation rather than Google/Microsoft Translate. They are much better.
For instance, when prompted for the meaning of "da", it correctly provides skin, but also provides "already", which is a confusion with "đã". I don't believe I had any notable difference in error rate between Claude and 4o, and it's looking like I will create a system to have someone manually review what is generated.
My main motivation for working on this is that there is such a drought of materials. SOTA LLMs are certainly great and have a pretty good effectiveness, but I wouldn't say they make great teachers of more obscure languages yet. I recall the Llama models explicitly state what languages the model is proficient in.