Problem:
GPT struggles to distinguish between different editions of frequently updated games, such as Warhammer 40K, Kill Team, and video games with balance patches. It often provides outdated rules or mechanics because it lacks clear tagging to differentiate between versions. This is especially problematic for games that receive regular updates, as responses can be misleading when based on obsolete information.
Proposed Solution:
Improve training data organization by prioritizing the identification of version or edition tags before processing game-related information. For example, when encountering an article or dataset about Warhammer or a video game patch, GPT should first determine which edition/patch it belongs to and associate that information with the response. This would ensure more accurate and contextually relevant answers, reducing confusion for users