One Act Play

In this work-in-progress project, I built my own Stable Diffusion video generation system. One of the largest strengths of AI is its ability to learn and adapt based on the “food”(Data Inputs, Video, Scripts, anything you can contextualize) that you feed it. This project aims to teach an AI fear, one of the most carnal human emotions. Humans developed fear as a response to danger and potential future harm, a preventative measure to ward off impending negative stimuli. The end goal being a video generated off a single prompt, fear, leaving all other inference to the AI itself. In my opinion, fear is probably the farthest feeling from an AI’s natural state. They simply receive information or commands and act on it. Over the course of this project I have fed my AI with many things I consider scary or horrific, this was done in conjunction with forcing it to generate videos based around various different phobias. Included within this post are Agoraphobia, Nyctophobia, Thalasophobia, and a bench mark test of what it does with the prompt fear as it is now. As I continue to develop both this project and this AI, I hope to learn just what it is that machines fear.