ABSTRACT
When discussing game artificial intelligence (AI), developers often get hyper-focused on advocating for a particular approach or a particular architecture, but many of the problems that make AI programming so hard appear regardless of the architecture use, and many of the most common solutions have been reinvented time after time in architecture after architecture. The difficulty of building an AI configuration generally scales worse than linearly with the size of the configuration. In other words, the more situations AI can handle, the more things it takes into account, the more options it contains, etc., the more complex it is to add another of one of those things. The reason for this should be fairly intuitive. Most reactive AIs function by evaluating the situation very frequently, and deciding what is the best thing to do right at that particular moment.
