Even as AI translation continues to make progress in replacing countless Japanese-to-English translators, two falacious arguments continue to be used to support the mistaken prediction that AI won’t replace professional human translators.
- Translation by AI cannot capture cultural nuance and emotion.
This is arguably true, but for only a tiny portion of translation that is paid for. Only a very small portion of the translation that is paid for has anything to do with culture or emotion. In contrast to the relatively much smaller fields such as literature, games, other entertainment, or marketing, translation in fields such as finance, business, technical, legal, patent, and the like, which have supported the livelihoods of the overwhelming majority of Japanese-to-English translators do not involve culture. There’s just not much cultural nuance or emotion in sight in most of the mainstream translation fields, which can be handled satisfactorily—the client decides what is satisfactory—by AI combined with low-paid post editing.
- Mistranslation will result in injuries, deaths, and litigation.
However attractive such outcomes might seem to translators hoping for a schadenfreude hit, these warnings, almost always presented in the abstract without factual basis, are not supported by verifiable incidents. These warnings of incredible risks need credible facts to support them. Those facts haven’t been forthcoming.
No amount of chanting of the cultural-nuance mantra or the deaths-and-litigation mantra will change what has already progressed in translation to an extent not much talked about, perhaps because the people affected are gone.
The replacement of professional human translators by AI with post-editing is well underway and coming soon to your optimistic neighborhood.