The problem is that with everybody and their dog able to generate art, those "massive and fast" releases are gonna be garbage. There will barely be an uptick (if any) in quality products and instead we'll just have more trash to sift through. I'm not anti-AI per se, but I am anti low effort. Everyone's so interested in shortcuts and instant gratification these days that they skip the human input entirely. Sure, generate AI art if you must (I can't stand the way it looks), but at least learn how to edit the damn things and make sure that all the wonky shit gets fixed (seriously, hands are gruesome and RIP all foot-fetishists when the AI craze gets more popular). And don't get me started on unedited AI writing... sure, it can make a great foundation (and I've used it as such), but without heavy editing/monitoring the drivel it churns out is abominable.
Right now, AI's biggest potential is as a tool to aid artists/writers/programmers, but there's a rush to make it replace them. It's not there yet, and it's gonna be a while before it is. In the meantime it should be used for things like easing workflow, fighting writer's block, idea generation, art base generation, programming ideas, etc. But the amount of 'games' we're gonna see where the 'dev' is just jamming everything into an AI program and expecting a fully formed game to pop out without severe issues is gonna be depressing.
Not to mention the risk of a decrease in actual, human-made games. If the market becomes too saturated then there's a chance that 'proper' devs will be forced to abandon it to make a living elsewhere. And THAT is a dystopian future I don't want to see :-/
