X's automatic translation will bring disaster to Japan

There is no better way to describe it: the gates of hell have opened.

What I am referring to is X’s automatic translation feature, which has recently become a major topic of discussion.

As you know, advances in large language models (LLMs) have reduced translation costs to virtually zero. It is now possible to translate several hundred characters of foreign-language text almost instantly. Given this, it makes sense for platforms to translate posts into users’ native languages by default. As a result, X is rapidly transforming into a truly multilingual social media platform where language barriers are disappearing.

The “Suggested” section of the timeline is already filled with tweets automatically translated from English into Japanese, while English-speaking users are seeing Japanese tweets translated into English. At present, the system seems to focus mainly on exchanges between Japan and the United States, but it will likely expand to other languages such as Chinese and Arabic in the near future. Humanity may be on the verge of creating, for the first time, a genuinely pan-linguistic platform for public discourse. Many users currently welcome this change as a historic step forward.

However, I want to state my view clearly:

This reform is likely to produce deeply negative consequences.

At the very least, I am convinced of this. The history of machine translation services in Japan suggests a similar outcome. In this article, I examine the ethnic and political tensions that machine translation has previously intensified and consider what X’s multilingual translation system may bring in the future.

The Japan–Korea Divide and NAVER’s Translation Bulletin Boards

Although the accuracy of machine translation has improved dramatically in recent years, low-quality automated translation has existed for nearly two decades. The idea of “international exchange platforms powered by machine translation” dates back to the early days of the internet.

One representative example was Enjoy Korea (commonly known as “Enkori”), a Japanese–Korean translation bulletin board operated by NAVER, South Korea’s largest portal site. It attracted significant attention by allowing users in Japan and Korea to communicate through automatic translation. The service launched in 2002, just as anticipation was building for the Japan–Korea World Cup. NAVER’s intention was clearly to promote mutual understanding between the two countries.

What actually happened, however, was the opposite.

Rather than encouraging friendship, the platform accelerated hostility between users from the two countries. Although some participants used the site constructively, politically oriented boards—especially those focused on history—became centers of intense emotional conflict. The service has since been shut down, so direct records are difficult to cite, but later commentators often describe it as a turning point in online Japan–Korea tensions. Even the existence of a “Prohibited Terms” section on the site’s Nico Nico Encyclopedia entry gives some sense of its atmosphere.

During the 2000s, the number of so-called “net right-wing” users in Japan—people expressing hostility toward Korea and China online—increased dramatically. Many analysts have argued that the expansion of cross-border online communication, exemplified by Enkori, contributed significantly to this development.

Why did this happen?

Understanding Does Not Necessarily Produce Friendship

Dialogue is often assumed to promote mutual understanding.

This assumption underlies cultural exchange programs, international cooperation initiatives, and everyday communication between strangers. It was widely believed that machine translation and the internet would deepen understanding between nations and ultimately strengthen friendship and cooperation.

At the time, almost everyone believed this.

In the late 1990s and early 2000s, discussions about internet culture were overwhelmingly optimistic. Many people assumed that digital communication and machine translation would connect humanity so deeply that national boundaries themselves might eventually become irrelevant.

But from the perspective of 2026, we know that this prediction did not come true.

The internet did connect people—but it did not necessarily make them more cooperative or sympathetic toward one another. National borders remain firmly in place, and in some cases conflicts surrounding them have intensified rather than diminished.

Why?

One reason may be the mistaken assumption that understanding automatically produces friendship.