I'm still working out some kinks, but I keep hitting the GPT4 cap at the moment. I wouldn't mind some feedback while I wait to be able to access it again.
My idea is to generate video ideas about a specific topic, Choose one and have it generate a script, Create the dialogue for that script in a format that's fairly easy to copy into text to speech systems and finally generate all the images you'd need to create a video.
It works pretty well at the moment. It sometimes gets the orientation of the images wrong if you ask for portrait images.