Red Riding Hood - Characters

Fairytales x Iron/Steam/Dieselpunk

Like everyone, been playing around with AI lately. Thought i'd try to put together a little batch of concepts for a little story/world. Went with something simple as a base, red riding hood with a twist. Not all are matching up 100% thematically, as I was getting excited by different ideas while working, but more or less in line. My general observation is that currently it's quite easy to get something beautiful very quickly, but the more accurate you want to be in your vision, the more work it takes. The craziest thing for me is how, once you have 1 good version, spitting out infinite similar variations is essentially free. Definitely some difference in difficulty though. Fantasy warrior? Free. Steampunk spider scissor witch? Much more involved process, due to the uniqueness of the idea.

For what it's worth, I find the workflow to be super fun, going back and forth and back and forth between manual paintover and AI refinement in a multi-multi-layer sandwich.

It's pretty fucking ridiculous and 50/50 exciting/terrifying. On the positive side, I think this will enable smaller and smaller teams to make bigger and bigger projects. On the negative side, it's dizzying to think what it means for everyone who's spent years and thousands of hours honing skills only to have that gap closed in a matter of a few months. In the end though, I think you either do your best to adapt, or you don't.

I don't think the comparisons to previous advances in technology are completely accurate and fair, due to the sheer exponential curve of the progress. Even within the last couple of months, the models have improved so much that whatever I had generated then is a joke compared to what's easily possible now. Why bother with spending manual time fixing hands etc, when in a month or so a new version will come out that will fix it even better. We're kinda at the stage where the software is improving faster than most people can adapt.

I think the logical conclusion to this is that within a decade or two, a single person will be able to make some truly huge projects, with zero traditional skills as we see them now. I think a lot of unmet niches of content will be filled in. Even with the idea of all of us creating our own IP's and graphic novels etc, I'm not sure if I can learn to be a good writer faster than GPT-3 (also, if the roles were reversed and writers were losing jobs, it'd be a joke to think that they could conversely all become good artists).

The most positive angle I have on this is comparing it to books. There are millions of amazing books out there, and I will never be able to read them all, but there is still huge demand, there are bestsellers etc. Currently the huge companies have monopolies on high visual fidelity movies/tv-shows/etc, but thinking about it, there aren't THAT many good series or films, and a lot of the main shows available now don't meet my preferred niche exactly, but I watch them more due to the lack of selection. So in that sense I'm super hyped! The question is then, what kind of projects will be possible by 100-1000 people teams?

Lastly is the question of copyrights etc. Personally the rule I set for myself is to not use living artists names in the prompts, that feels icky to me, as these are people still trying to make a living. I am all for the idea of greater regulation within the datasets, but I'm not sure how much is possible with the speed of this tech and the much slower pace of legislation. I think once the music industry is hit with this, there will be a lot more public discussion and clarification - sadly artists don't have the legal strength and public support. In terms of the general opinions on this from non-artists (i.e. people on the subreddits etc), they are pretty appaling and heartless to me, people truly don't give a shit, but being honest I don't think I can blame them. This is meaningful to me, because I'm an artist, but I'd be equally excited to mash up my favorite bands to create new music with little remorse, so I think I gotta acknowledge my hipocrisy.

So yeah, in the end, very mixed emotions.
That's the end of this Unabomber manifesto.