That's assuming that the only market is stuff people are hanging up. The games industry, one that already takes advantage of its workers, is going to love this to the detriment of really passionate artists who love their craft and industry.
So when an artist looks at AI, it's unlikely to be as a tool that will build a whole piece: Insufficient control, and currently nowhere near good enough to do more than occupy space, like a little painting in a hallway or in a hotel room. But it is something that can be used to better spend the budget in places where it'll be more impactful for the quality of the piece. Not unlike how CGI is often used today in places where it wouldn't have been 20 years ago, and it's aiming to be invisible. Not because the shot was impossible, but because it's cheaper.
Treating AI in art as a moral thing will end up being like the people being against synthesizers in the 80s: It's a viable creative choice for some things, but ultimately not a good expectation for industry direction. Ultimately the vast majority of art is commercial, and we'll see shortcuts being taken for budgetary reasons. Nobody is manually animating every detail of every mesh in a game like this was Toy Story. And even though doing that would produce more work for artists, it wouldn't make better games, really. And we'd sure have far fewer of them.
Expedition 33 nailed music, aesthetics, and narrative, and I am glad that they took a diffusion model for what it is, not for what marketing wants you to believe. although the game itself would benefit from one or two months dedicated exclusively to optimization, it is THE reference of how generative technology can be used - purely internally, to ideate and iterate at the pace of your taste and a bunch of H200s. we are aware of that process detail purely because they slipped in one place and got briefly "owned" by Twitter.
1: https://www.ign.com/articles/larian-ceo-responds-to-divinity...
2: https://www.dexerto.com/gaming/clair-obscur-expedition-33-ai...