I started out in midgrade CGI before moving into proompting in 2023. The biggest challenge with generative AI is inconsistency; a minor needed change will result in a reroll of the dice and give you a brand new image/sequence. Yes, there's inpainting for still images but there isn't an equivalent for animated sequences. The studio's going with animals because they don't wear clothes and anatomy is covered in fur. This is a deliberate choice, as is imitating cartoons where no shot is longer than 10 seconds.
It's a fascinating technology but more for the things it nightmarishly misunderstands like physics (gravity, clothing, hair, wind forces), anatomy, materials, etc. that CGI already covers well. These guys are either scammers, deluded or both.
Expect the cat and the mouse to change relative size and appearance, background sets to change props, that sort of thing.
The underlying problems behind all of this have to do with intrinsic issues with how the tech works and no one is interested in rebuilding it from scratch.