So, before you criticize me for using AI art, know a few things:
1 — If you want to make art for the site, and I like your art’s aesthetics, then I will gladly accept it. But I wouldn’t pay you for it, and ONLY because I am a broke lass. As for the aesthetics, well, look at the art that was generated — that’s what I am looking for, except I want it to be a little more true to the ideas herein (so skin tones,
2 — I need the art for myself and my players. It helps us to visualize the world as we play.
3 — it is just me. One person. A tool I already had for other purposes, added in a generative AI tool, and so I figured I would use it. I don’t even know what the basis for it is, and I am not that big a person on all the details for it.
4 — Have you used google images lately? It is all AI art.
5 — Ever try to find a picture of a mixed race person with mesoamerican, african, and southeast asian heritage who is also a dwarf, elf, gnome, spright, triton, iaran, etc when they are all very specific to the world described here and not your standard D&D basis?
6 — Ever tried to find a person of the above who is dark skinned?
7 — Ever tried to blend Mesoamerican motifs with Greco-Roman and Kushite motifs?
8 — All of the above in a photorealistic, anime influenced style?
9 — Are you willing to give me money to commission art for the site? The artists I do like and work in this style that I could hire would charge about 250 per image for the people images. There are about 4000 of them as of the time of the writing of this post, with around 5000 expected in the end.
10 — You are aware that right now, under federal Law in the US, there is no way to copyright or trademark generative AI images, so you can snag these yourself, are you not?
That all said, you should also understand a few things:
A) I am aware that all AI art at present is based on theft. I am also aware that the AI companies are likely to win the court cases — because I am in one of them, since my written work was used.
B) I did not go out and buy a tool — I already had one and it was added in, and I’m not paying anything for it at this time (no sub fee or other stuff). I didn’t get the tool I use in order to do this — it was something that was there.
C) if I ever do get enough money to commission art, I will absolutely change it all out — and then protect the fuck out of my rights to it because I am a complete bitch about IP. Plus, I design my own 3D stuff to print on my own 3D printer, and I have a blast doing so.
D) This site is for me and my friends in our game and unless you are interested in joining a game and I have slots, you aren’t one of them. That you are here is just a thing about the internet, and I could, other than the above, give a flying fuck about you or your opinions.
Now that you know and understand all of that, one final thing:
There is a microscopic chance that someone will want to license all of this stuff for sme purpose or other. I am not out there trying to sell it, nor am I really giving much thought or effort into the concept of promoting this place. That’s not what it is for.
But, should someone do so, well, then they get to deal with creating all new art for everything, and I will be more picky than I was in what I decided was “good enough” in terms of my using the AI tool.
So with all of that to think on, what have you to say about me using AI for the images that isn’t just being a fucking dick about it?
oh, and the above was generated off of a photo of me, lol.
So, now that I have that out of the way, let me explain an immense problem with Generative AI around Wyrlde, and why it will never be a substitute for actual art.
Wyrlde is a complex effort that pulls together disparate elements and reimagines things in ways that I haven’t seen before outside my own head.
This is a world where the basic, core concept for a “ship” is a giant fucking catamaran. A twin hulled beast of a water borne vehicle that no AI program I’ve seen or have access to (not that I have a lot of access or will even try for it) is capable to creating because generative AI does not imagine new things, it only creates versions of multiple existing ones — and it discards things which are outside the normative “bell curve” of that range.
Generative Ai is limited by what it learned on — and it is not the end user who is the issue thee around it, it is the company that creates the tool that the issue.
The goal of getting folks to turn away from AI art is to pressure the companies to pay for the art they have stolen. That’s it. To make AI so horrible a thing and so strongly avoided, that AI makers just automatically start to budget for the use of human created material to use for their LLMs.
To deame people who do use AI tools is to tell them that no Art is better than any art. Simply enough. Which, for someone who is trying very hard to imagine something that does not exist and lacks any real facility, is essentially telling them to fuck off and do something other than imagine shit on their own.
To make a point about someone else’s money (mine, in this case, as well as other people whose work was used without permission for profit).
I am perfectly fine with generative AI expressions never being able to use copyright. It is too late to try and put the genie in the bottle, and the odds of getting these companies to fork of any meaningful amount of coin that would truly pay for the crime they have committed is laughable at best to me (cynical as I am about them), but denying anything made with the use of AI the ability to be trademarked or copyrighted?
That guts the money issue. That undermines the entire *current vision and purpose and goal* of the AI creators. They want to build up huge companies worth billions, and they want other companies worth billions to pay for the use of their product, and the only reasons that people will do so in that situation is if they can own the final creation.
i, however, as is my wont, tend to be a little more seriously focused on it — and absolutist.
Let’s say that AI is used to create a “digital movie star”. Call it one quarter of all the work that goes into that digital movie star is based in LLM/ML AI stuff. Under my basis, that entire digital movie star is not eligible for copyright or trademark protection. As an IP, it is utterly and completely useless.
Let’s say that a film script is produced using AI, and then cleaned up by a human author. That script is not eligible for copyright protection. THe film produced from it is not.
The one exception would be if the LLM is one hundred percent validated and wholly paid for with a substantial value (and a royalty system from every generated image or response out of it) through the artists whose work contributed — and the company can prove it, with permissions on file and everything.
That would be fine — but that AI product would be expensive and costly, and out of the reach of the everyday person like me, because in order to pay those fees and royalties, the cost of such would be high.
For me, the approach that I am using towards generative AI art is to use it for what I want, and operate in a manner that follows this general practice. This is a subversive use of the potential, and does not contribute to the broader bottom line since I was already using the tool and would have used it outside the addition.
However, as an End User, I am not part of the problem. That’s like saying people who use toilet paper are responsible for destroying forests — it sure sounds good, but as an ethical and moral argument it is so full of shit that clogs are inevitable.
So, that’s my stance.