1-1) If the AI art was good (consistent, no artifacts and weird stuff like odd fingers and skin and eyes, thematically-appropriate skin tones and features, no weird disconnect between subject and environment), and the art style could be whatever type you preferred (hyperrealism, 3D anime like, cartoonish, or whatever), would you be ok with it being used to make games?
I would say, at least for me personaly, it heavyly depends on the Artstyle used. If its thee "AI Slop" Look, than a NO. But if its a good Art stlye, closer to a good drawn art oder a good 3D art, why not.
AI art CAN be really good, but its not as easy as most ppl think to get really good art.
1-2) Would you prefer "good" AI to asset swap art? (Asset swap is when you pick a 2D or 3D hair and snap it to a base model. RPG Maker is an example of asset swaps for sprite creation. Some WebGL games use asset swap libraries for 3D characters.)
Same as above. I really wouldn't be supprised if thats already used in many modern RPG Maker Games/3D Model based Games, without ppl knowing that they are part AI Made. I mean there are even Big Game Studios who use AI assets like this and ppl don't know it.
1-3) Would you prefer "good" AI to porn inserts of real people?
No. I think, thats a line nobody should cross. But yeah probably most would want it.
1-4) Would you prefer "good" AI to DAZ 3D models?
I never worked with DAZ, so i can't say much about this.
1-5) Would you prefer "good" AI to hand-drawn art?
That heavily depends on the Model and trained Data thats used. I know there are Models out there, which can create really good looking Art. If ppl take the time and create a good looking, consistent and theme fitting art, without "AI Slop" look, sure why not?
2) Have you tried making AI art yourself?
Yes. And my oppinion is, most ppl think AI art is easy and fast. Thats not true. To create a few single images without need of consistency or theme fitting, thats tue. They are easy. But try to make a whole game with, lets say 500 Images/Assets. Than its not this easy anymore. First of all, it heavily depends on the Model u use. Than you need to refinine this Model with fitting LoRa. And probabaly you need to create many of the LoRa urself. And thats the Part where it really gets complicated. Ppl who already trainded Own LoRa probably know what i mean.
So yeah, AI Art can be good and usefull, but that requires a deep understanding of how the AI works, how it learns, and ofc how to Prompt it. Most models out there are TAG based, not clear text, wich makes it not an easy task to get what u want and need. I mean, to compare that, everyone (and i mean EVERYONE) can use a pen and draw things on a piece of paper. But if its art or just bad scribble, thats where time to learn and talent comes into play. Same you can say about AI art.
Now ppl can hate me if they want, but thats my oppinion
