Can AI create a Moho character?

[rocky53204]

Not yet it seems, but, I invite you to submit the following prompt to chatgpt.com for a quite surprising result.
It seems chatgpt is quite familiar with Moho! Once you get an answer it will ask for other prompts... follow that rabbit hole as far as you like.

prompt submission: design a production ready, fully rigged, Moho animation, male character with the appearance of a gnome.

[SimplSam]

I took the bait.

It presented this as a visualisation idea for a Dinosaur character called Rexy.


Then, later provided this as an .svg file that could be imported into Moho.


... nuff said

[Greenlaw]

Well, I give up. Thanks, guys.

[dcr]

FWIW, Corel has a product called Vector FX which uses AI to create vector images that can be exported to SVG. That would save the step of trying to vectorize the bitmap image that most AI image generators produce. Can't say how well it works as it is Windows-only and I use Macs.

But, when you consider the licensing and ethical issues, I think you're further ahead to just draw your own characters. I think it's okay to use AI to generate character ideas and relatively rapidly generate different options and ideas that you can use for inspiration in the creation of your own characters but I'm not keen on using AI to generate the final image or character to use.

[Greenlaw]

I don't think anyone here is being serious about the topic.

Personally, I don't see how using generative AI to make art for rigging saves time, at least when the goal is to produce anything of quality or originality. And, yeah, that's not even considering ethical and other problems related to using generative AI.

A while back, I dabbled with generative AI to see if it was feasible for creating background art. In each case, I spent so much time repainting and correcting the paintings that I felt I could have done a much better job more quickly if I had designed and painted the work from scratch. Another problem was maintaining scene consistency between camera angles for what were supposedly the same locations...this was pretty much impossible. Besides that, none of the art looked anything like what I would have created myself, even after all the repainting.

So, while this was an interesting experiment, it was also a big waste of time for me.

For actual production art, I'll stick with creating it myself without gen AI, or hire a fellow human artist who knows what they're doing.