Raw LLM Responses

I think that AI art on its own isn't anything to write home about. I think the thing artists SHOULD be doing is training their own models on their own art so they can do things like train it how to do animation inbetweening and generating assets for things like rocks and trees and background NPC's such in indie games and other time-consuming stuff that currently throws up a huge barrier between independent artists and more ambitious projects that can currently only be made by huge studios with a bajillions artists and other resources. Just using it to crank out _the whole art style_ is obviously a bad idea that looks bad. But using it to amplify your own art style might be the thing that allows the next indie studio to be able to make the next big open world game or fully animated feature length movie for a fraction of the cost to tell stories and create worlds that would NEVER survive for a second at a corporate board meeting. Human artists should still be making the important features of the project. Important characters, player models, important landmarks and scenery, etc... but then what you can do is create all these essential assets that establishes your style and then use that same work as well as concept art and even art that you never release to the public to train your own model. By training the model on art that does not get released to the public, you can make sure that nobody can truly copy the model, and you can use these pieces to nudge the model on certain directions. I think creatives are making the wrong call by going "all or nothing" with AI tools. I think the best way to move forward is to use them like the tools they to be able to tackle more ambitious projects that wouldn't be possible for one person alone to complete in their lifetime, but that now may be within reach. I also don't think that using bog standard AI generated art and shitting it onto the internet is a good thing either. I think AI should be used to fill gaps rather than being the whole product.