Do artists like it if the game is modded? It's mixed isn't it?
Yeah, personally i don't like modding all too much. I prefer the game in its original state. Visual 'improvements' made by modders often rape over original intentions.
But modding is optional. Anybody decides for himself, and every gamer first sees the original art.
If we get such AI post processing to alter artworks, the original art would be never seen by anybody else then the artists. That's a different situation.
Artists have less art control without doubt. They may even need to author their stuff mainly so it works well with AI, no longer so it creates the intended expression directly.
So the machine rules over the artist. Under such conditions, true art is not possible at all. And true art generated by machines isn't possible anyway.
As a consequence, we devalue our medium. It's nothing worth if it's generated by machine, so we can't request 80$ for that. Games would simply become worthless, eventually.
AI generated content isn't really inherently different than the non AI procedurally generated content used in games.
Yes. That's exactly where i see proper AI application. Procedural content. Stuff where quantity is the problem, not quality. Backgrounds, textures and materials, terrains, castles, etc.
I think AI can do great for that. It can look great, save costs, and everybody is happy.
But i would not use it for characters appearance, dialogue text, speech synthesis, or procedural stories. That's human territory, so the content should be created by humans.
That's at least how my ideology has shaped up til yet. Not sure about anything. This situation is so new and different, it's already hard to figure out what think about it personally.
For example, i think the limitation of dialogue and stories having to be just static content felt impossible to overcome. And this sucks, because i want to lift all limitations i can. So i can do new things, which are exciting because we did not have them before.
And now, out of a sudden, we could lift those limitations using AI. And we want to try it out. We can come closer to our dream of a simulated reality this way, so why not?
The reason maybe is that AI is not a proper simulation. It does not understand how things work. It only observes those things and then generates new but similar content out of those observations.
The devs also do not understand how AI precisely works, so they lack control. They do trial and error on picking from various ML algorithms, and they select the content used for training. That's all they can do.
It's all fuzzy and uncertain. Can we design good games under those conditions? Or does the game design as well become a process dictated and constrained by AI properties we neither understand nor can control precisely?
But what matters even more is end user experience.
For now that was a lot of issues. Characters clipping through walls, running in circles, or not moving at all although we shoot at them. Lot's of visual glitches on top. We totally failed at simulating natural characters.
What the user sees is labeled as 'computer bugs'. But that's not what it is. It is human error. The human player sees the inability of the human dev to create the illusion of intelligent life.
It's not great, but it is acceptable. Human error is unavoidable, and it is natural to all of us.
But it seems AI error is not as natural. It is alien, feels totally off, and unintendedly strange. It may be rare and subtle, but we may always spot it. Most humans might never accept AI characters. Neither visually, nor about their behavior.
Ongoing progress on the technical side to improve this may be compensated by increasing social resistance against the 'competing AI species' as well, making it a risky investment.
Personally i remain very curious but doubtful about real AI used in games. There surely is a lot of opportunities, but there also is a huge risk to just piss people off in the long run, it seems.
But i'm uncertain due to age. I realize i became a retro gamer already. I may be just too old to accept new things, and fail myself to figure out what people want.