Look into design thinking and in particular ideation. There are lots of formal processes, exercises, activities, etc. that are used by individuals and teams in all sorts of contexts specifically for coming up with ideas. The process is usually one of throwing a bunch of things on the table, sorting through them, getting rid of most of them, elaborating on the ones that seem interesting, then following one to completion, or at least to some sort of first draft/prototype/mockup. You then decide whether or not you want to work on the draft further, or decide that it's a dead end and start from scratch. The thing with "ideas" is that all of them are terrible and only serve to help guide us towards doing something interesting. Creating things is an intensely iterative process, and what you start with is unlikely to look much like what you end up with after a number of iterations.

Ideas are also all derivative. There are no new ideas, just riffs on existing ones. Even most interesting and innovative works have been influenced by past works, or works from different disciplines, or inspired by nature. If you're looking to make a short comic, start by figuring out what works and artists and styles you like. Try recreating parts of them, or emulating them, or combining elements of them, and see if the results speak to you. That's one of the few actually useful applications of LLM AI. You can quickly test concepts, maintain some elements and discard others, do mashups, etc. When something grabs you, try to figure out what it is that resonates about it, then try to recreate it with your own spin.

Ultimately, ideas are just prompts for doing work, and having a good idea (to the extent that such a thing even exists) is far less important than being willing to test a number of ideas to find out what will motivate you to spend real time and effort on creating something.