Can Neuro generate or ask another AI to generate images (scenes, backgrounds, objects, etc... ) for her while live? If so, in some streams, she could decide herself "where she is" or "what she's looking at" and maybe even look at it and react to it right after. Just for fun as an experiment, not replacing actual backgrounds or other assets.
#Neuro generates AI images live from her thoughts
raw cloud
azure dove
Can Neuro generate or ask another AI to generate images (scenes, backgrounds, ob...
Don't think it is a good idea
Here's why:
I think ai gen background will look more... Repellant for some people (old and new)
Also it would mean hooking up 3rd party AI model to Neuro which won't necessarily create image Neuro wants
And I think there is also a few difficulties in creating correct prompts.
sullen flume
I like the idea of having her subsequently live in it - it's a more interesting take on the dull approach that Meteora does.
I think Neuro, like other LLMs, would be good enough at making the prompts to stick in the image generator.
The killing blow for this idea though is that image diffusion generation is a lot more controversial than text or voice generation. By introducing it so blatantly, Vedal could end dissolving some of the goodwill he's miraculously garnered in the vtuber community and could attract ill will from beyond the community.
There's also the issue of not having it look garish and drawing attention away from Neuro.
I think you could mitigate the bad publicity if you used a model that made highly realistic non-cinematic photographs
sullen flume
I like the idea of having her subsequently live in it - it's a more interesting ...
I suppose my concerns here are weakened by the fact that sometimes AI backgrounds are used, and it has never lead to anything.
raw cloud
Don't think it is a good idea
Here's why:
I think ai gen background will look ...
I thought more about an experimental stream once in a while.
raw cloud
I suppose my concerns here are weakened by the fact that sometimes AI background...
True, I've seen some vtubers starting out with AI backgrounds and not getting much backlash.
fathom mural
Just an idea, but maybe Vedal could commission somebody to create a large number of scenic props and then give Neuro/Evil the ability to use them to create their own background/scene? No one would complain about it because that's their own property. But I'm not sure if the technology is there yet.
sullen flume
could be possible especially if the vision can be used dynamically when props need to be added
I think Neuro could just grab the images from the internet or some assets website
fathom mural
could be possible especially if the vision can be used dynamically when props ne...
only if we could trust her vision 
raw cloud
gotta be honest, my idea was more about "neuro makes ai pictures and reacts to it" than "ai backgrounds for neuro".
sullen flume
yeah i personally found it quite boring when meteora did that
but I find conventional art streams boring too so maybe it's just me
raw cloud
yeah i personally found it quite boring when meteora did that
but I find convent...
Meteora who?
sullen flume
Meteora who?
an AI vtuber whose niche is doing imagegen
raw cloud
an AI vtuber whose niche is doing imagegen
Ah, I've seen her before, I've seen a few ai vtuber but it's hard to remember them.
Does she do image gen every video or can you direct me to a specific one?
sullen flume
not every stream, but they did do a stream focused on it 10 days ago
daring raft
I thought about this too, but for for ai video as a background, I'm kind of changing my mind about it. Mostly because these models were made in a less than savory manner. Also, when say Sora relases, everyone and their mother will be using it, so there will likely be ai video fatigue. Best to stay away.
If neuro learns how to paint like bob ross one day then maybe that could work as part of the stream preparation.
raw cloud
Neuro generates AI images live
raw cloud
I guess AI art has a stigma, I just thought of this as quick throw away pics of Neuro's imagination.
raw cloud
Neuro generates AI images live from her thoughts
raw cloud
Just an idea, but maybe Vedal could commission somebody to create a large number...
I don't think it's a problem for her to pop up various elements and images on stream and maybe even give them a specific animation (like spinning), but my idea was an ability for Neuro to give a visual and even complex representation in real time of what she's thinking and experiencing at any time, and it would be different every time.
That side, I think I'm also gonna post this alternative.