Scripts are very intuitive, but can be difficult to disassociate from adjacent concepts and even harder to find a good formal correspondent. This book is responsible for a lot downstream, but perhaps in retrospect, it was not the best attempt to lay the foundations for the concept cleanly. However, Its been nearly 50 years, and the basic idea has survived yet, albeit through adjacent disciplines. I've been currently reading it very thoroughly in hopes of rationally reconstructing it, and to most who will find similar interest in event schemas and their potential, this is the canonical first step. I think a good moral from looking at this text is to not jump ahead, but to create good reliable foundations. Conceptual dependency was not a good theory and they knew this before it became undeniable; however, being able to gauge at more "interesting" questions was their main interest, and for that, it may have actually slowed its progress down the track significantly. There is a consequence for that kind of action.
A fascinating look at beginning efforts to codify scripts, plans, and goals. Ah, will humans ever consistently be this predictable? From my 2024 perspective, it’s easy to see the cultural assumptions built into these mental models. A decidedly good and necessary step into this field.
Amazing primer to how one might structure in-depth character AI. Comprehensive enough to help build your own systems, while not being too nitty-gritty as to be inflexible or a slog.