I'd like to make this the start of a thread for long-form AI videos -- that is, videos assembled from multiple shots, videos that are something more than the output of a single prompt.

Semi-realistic loli warning! Don't click below if you don't want to see such things!

And, before you report this post for that, keep in mind Gurochan allows photorealistic content, even with underage characters, so long as it's ultimately fake with no real people in it, and so long as the files are hosted elsewhere.

https://files.catbox.moe/cxo9mz.mp4
https://file.garden/aecQ3YtN8glaOMdJ/gc-posts/snuffing-sandra.mp4

I'd like this to be a thread for videos that aren't just "money shots", but that try to tell at least a teeny tiny story, that do a little set-up work before the major action occurs.

And let's say videos that are at least 45 seconds long.

I'd also like this to be a thread where people talk about how they assembled their videos, with tips, tricks, resources, etc.

My video to start this thread is based on a drawing by the artist Mad.Aye. I took the liberty of changing the teased death by gunshot in the drawing into an explicit death by gutting and stabbing. (Since only one image can be included per post, I'll put the drawing in a reply.)

I wish I could say I'm trying to be a perfectionist. This video, however, is definitely well short of perfect. It has some wonky transitions, not-so-great sound, and bits that are definitely rough and fake looking.

What I will say is that I've been willing to do lots of edits and Photoshop work and rewording of prompts and multiple iterations of every clip until I at least raised the bar a bit on what imperfections I'm willing to live with, compromises so that I don't get stuck in a rut fighting every little thing that doesn't work quite right.