Traditionally, the global standard for translation has been converting foreign languages into one's native language. I understand this approach is predominantly due to the richer vocabulary available in one's native language for expressing ideas (with Japanese-English translations often considered an exception). However, the advancement of AI technology is significantly narrowing the gap in expressive capabilities between foreign languages and native languages. Could this development not heighten the importance of understanding the original text in translation? Ultimately, the best method for producing high-quality translations efficiently might involve:
- When the source language is native and the target language is foreign, a human who understands the original text could use AI to translate it into the target language, then fine-tune the translation while interpreting it in the target language, leaving the nuances of expression to AI, or
- When the source language is foreign and the target language is native, a human could understand the original text with or without AI's assistance before crafting the translation. While AI might be used to broaden the range of expressions, the fine-tuning of these expressions should remain a human task.
In this context, what becomes paramount when AI is leveraged to its full potential for unlimited expressions? It's likely the comprehension of the original text (source language).
Thus, if we presuppose the use of generative AI in translation, might the global standard shift towards translating into foreign languages rather than into native languages? The rationale being that understanding of the original text is inherently deeper when it's in one's native language than in a foreign language.
Applying this to English and Japanese, it suggests that Japanese translators would be better suited for Japanese-to-English translations, and native English speakers for English-to-Japanese translations.
Could we be on the cusp of a change in the global translation standard?
The above is merely my hypothesis and the predictions based on it.
