This chapter discusses a number of issues relating to the future of AI and games that do not fit squarely into the preceding chapters, and which are likely to become (or already are) important research topics in the field. The first half of this chapter discusses the challenge of generalization: developing general game players, general content generators, and general player models. A key argument is that the age of foundation models has not quite hit games yet, and that creating such general methods is a huge challenge with a potential huge payoff. There is also a discussion of multiple use cases for AI in games and AI research problems spurred by games that do not fit neatly into the classification in the rest of the book, such as virtual cinematography, chat monitoring, and hyper-formalist game studies. Finally, there is a short discussion of the limitations on game AI brought by existing game engines and the prospects for building AI-enabling game engines.