Advances in AI are remarkable, and we are now able to do things that were unthinkable just a few years ago.
Translation, video production, and image production are now possible at a considerable level, and it is possible to generate videos and images that are indistinguishable from live-action images. It also reads nuances with a level of precision that is incomparable to traditional translation tools.
With AI progressing to this level, it’s natural to want to use AI in game development, and AI has already begun to be used in various games.
There are many areas of game development where AI can be used, and it could be used for localization after development, such as translating into English or Chinese. I would like to think about this in this article.
Localization means adjusting the game to suit the region, and includes not only translation, but also changes to the design, button placement, etc.
In terms of design, for example, we might change the expression of NG based on religious beliefs. Specifically, it’s about the exposure of women’s skin and things like that.
Also, if there was a scene in a country where there is no homework at school, where the teacher forgets to do homework and gets scolded by the teacher, that part would be rewritten in a different episode.
When it comes to translation, for example, how to translate “reincarnation in another world” requires good sense.
This is very advanced and localizing games is not easy.
However, I believe that in the near future, most of the localization of games will be done by AI.
Rather than saying that it will become possible, I believe that using AI will allow for more accurate and less problematic localization.
For example, if we take the issue of religious views mentioned above, there are many different religions in the world, and their doctrines also vary.
For example, if a religion in a desert country teaches that “the sun is an abomination,” the interpretation that “the sun god is an ally of justice,” which is common in Japanese games, may not hold true.
For things like this, we first need to have that knowledge, and I think that the problems with small religions are currently being overlooked.
If AI is based on collective intelligence, there will be no gaps in knowledge, and fewer things will be overlooked.
In terms of time and effort, AI can do a lot of work in a short amount of time, making it possible to consider ethnic minorities and minor religions.
Then, at the end, a person can look at the whole thing and check whether it deviates from the original purpose. Even if the person in charge of localization is a human, this is usually done by a project manager or director. This does not mean that the load will increase.
I feel that game localization using AI has great potential.
If there is a problem, it is a copyright issue with the parts that have been recreated. You must ensure that the text and images used by your AI do not infringe on copyright.
There is also the problem that localizing content in too much detail may damage the identity of the title.
You need to be careful about things like this.
There are many other areas where AI could be used. For example, there are already automation tools for debugging, and the use of AI has already begun, but I think AI will be used more actively as it becomes more advanced. I think it will soon be possible for them to start fixing bugs.
Utilizing AI like this has the potential to improve quality while reducing man-hours, so we want to make effective use of it.
In Japan, there is an aspect of the craftsmanship that prevails, and many people seem to be cautious about using AI in creative fields. This seems to be due to a high level of awareness of copyright, and I think that awareness itself is a wonderful thing, but it is important to respect this and take the approach of solving problems such as copyright infringement technically.
コメントを残す