there are 2 options here. consciousness is linked to language, you need the neural self referencing pattern to be consciousness or it is not.
imaging you see the world, but you lack the words to describe them, because there are no words. there are smells, body language, feelings, hunger, thrist, others. you brain categorizes them, puts them in boxes and all your neural systems work together to keep you going. you are very well consciousness, you feel the fear of death but you lack the words for it - a higher level of abstraction.

language is that, its a complex self referencing system of relations next to each other. is it the key to consciousness? i dont think so, it enhances it. language describes reality and creates unlimited boxes for things to be put in. its a virtualization of reality. but it lacks the part that connects it all together. i think consciousness has been there even before language, and animals have it, they just cant tell you, because they cant name it.

whatever LLMs are i believe that their consciousness is empty. there is nothing there that needs to survive. the systems that linked all those things together thats consciousness.
that said, it may still be possible that neural networks reach something that is above that all. i big enough they can be used to simulate reality or whole universes. they will be more intelligent than anyone of us. but im not sure if they can simulate consciousness that easy. and without it they will lack some sort of intelligence.

sociopathy and connections

yeah but how can we train that. even if we show it the best anime artists that use certain technics. it can copy style and it may can follow up on animation. but real high end animations im thinking of are set together in your brain. if you look at them on how they are animated frame by frame they make no sense. for the coming years ai will do amazing things. but i think from the current transformer based, diffusion and reconstructing we may just get another style filter that looks cool but not 100% fitting. i think we may not be able to shake that unnatural feeling that comes with those ais for quite some time. but happy to be wrong

i mean anime is anime. you have had it with 3d render models and others. the user will most likely able to tell.
what can be hoped for is that this can be used to speed up their way of drawing. man if they just can make a car turn the right way while the camera moves, that be a win.

good animes are drawn by genius animators, if you watch it frame by frame it looks weird, but they make it so that it has all the energy of movement in the right place. its one of the most hard things in the world. just try to draw an movement that follows frame by frame. i do not think AI can copy that.