As has been pointed out exhaustively, using LLMs for translation is totally fine. But to satisfy OP and folks like @Matija_Nalis who have been arguing about the point, below the cut is a translation of the original post, in my LLM of choice. My translation tool also helpfully points out that the Korean contains a likely typo, 여러 방명에, which it interpreted as “in a variety of ways.” I would ask that @apm-wa run this translation through the AI detector tool that the forum moderators have been directed to use in order to show what happens when LLM-translated text is tested in this way.
The forum administrators have announced a rule prohibiting the use of AI in posts made on the official forum. I am calling it a “rule” for now because I do not think it contains enough specific detail to qualify as a “guideline.”
As I understand it, the OSM forum decided some time ago to prohibit and penalize the use of AI. Whenever I have seen discussions about this, I have pointed out that I also use LLMs—large language models—when communicating with English speakers, and I have asked, above all, for the permitted scope of LLM use to be clearly defined. Unfortunately, I could not find any such clarification in this announcement either.
Some people simply say that AI use is not allowed, without offering any specific details. Others say that using it for translation is acceptable, or that receiving some assistance from it is acceptable. Still others say that the forum already has a translation feature and that people should use that instead. As a result, I do not know where the administrators intend to draw the line.
In particular, some people use expressions that imply “automatically generated posts.” Honestly, although it may technically be possible to write a post entirely with an LLM, I have difficulty imagining a situation in which someone would do something so nearly pointless.
My assumption is that they are referring to cases in which someone gives an LLM a topic and a general direction, then posts the generated text unchanged. Even in such cases, however, there is still debate over who—or what—should be regarded as the original author of the text.
More importantly, even experts have reached no agreement at all about who should be considered the author when a person receives assistance from an LLM in any form. In other words, the “permitted scope of AI use is unclear.”
I understand that, even though AI is now widely used, some people still have negative views about how it should be understood or what attitude should be taken toward it. Others may be using it in a wide variety of ways.
The question of the “permitted scope of AI use” may not be resolved cleanly until these different positions have been sorted out to some degree. However, if the forum administrators intend to establish this as a formal policy, I believe users need to know exactly what that policy means.
Ideally, this should be discussed in the original thread. If that is difficult, I believe there should at least be some discussion among the members of the Korean community beneath this post.