So my big issue for sometime is sourcing good/good enough images for the game and I've been fucking about with Stable diffusion over the past "X" amount of time and my thinking is broadly this;
What's been stopping me is finding appropriate images, AI solves that - so no excuses there.
Even if they're not _great_ they'll do for now.
If I want/find out how I can train a LoRA to make the characters consistent and make the faces (At least) match across scenes.
This is mostly programming which I am more comfortable with.
Maybe make a "The other girl" short story as a demo and see how/if it works.
Thoughts?