Okay, after this one I am probably done with posting animations of my old arts.
First, I just wanted to complete the ensemble of GIFs for the rest of the ladies from The Real Superheroines of Hive Valley.




Using AI to try to generate video clips still has some weaknesses. Each of these took several attempts except for Mary Marvel.
Notice how a second dagger mysteriously appears in Sheena's hand? Although that is consistent with the character in my comics, I didn't ask it to do that.
Also, the "vortex of blue energy" is supposed to be on a vertical plane behind Fantomah and she is supposed to step backwards into it and disappear. Close.
Moon Girl. I tried and tried to ask the AI for her to do a loop de lop towards the camera. After several failed attempts I changed it to "fly up out of frame" which she still fails to do. Also, the AI does not know what a "superhero landing' is. I thought Deadpool explained this to everyone!
Mary Marvel is actually the very first try for Grok Imagine. I tried several times though to make her appear in a flash of lightning, but those all ended up being pretty lame. AI does not seem to understand that "lightning" is fast and not a gradually increasing flood lamp.
So, yeah, still problems. Mostly the same problems that AI image generators have. In that they don't follow all the instructions, they follow some of the instructions. They also can't make small changes. If you want to change a clip, you can't tell it to make a small change to the current version, it will simply start over from scratch with the original prompts and you get what you get.
For example, here is the cover of Java Jaguar Book One. I can't make the AI use this exact same video, but just remove the whiskers from the baby jaguar. Instead it will just create something brand new.

And don't get me started on trying to make fight sequences...

It will be a while before the technology is good enough to make exactly what the creator envisions. But I don't think it is that far away.
Few closing images:



Oh, and this was a Patrick Nagel print from the 80's! I had to do it. I just had to.

This stuff is likely to get better, but it's already impressive. A lot of jobbing animators must be nervous. I'm still waiting for the ironing robot.
!BEER
Right now it's good for making this cool little clips, but not for a coherent storyline. I tried actually making separate clips for each panel of my last Kickman page and it produces some bizarre results. A few more years of development though...
View or trade
BEER.Hey @sidekickmatt, here is a little bit of
BEERfrom @steevc for you. Enjoy it!Learn how to earn FREE BEER each day by staking your
BEER.View or trade
BEER.Hey @sidekickmatt, here is a little bit of
BEERfrom @steevc for you. Enjoy it!Learn how to earn FREE BEER each day by staking your
BEER.Impressive, but I wonder how well it will do making a comic strip into an animated clip AND if it can manage the comedic timing. At first it seems like it'd be very hard, given it's a tough task for even a comic's creator to transition from print to animation. But these tools are essentially pattern recognizing idiot savants, and if they have other animated series to reference as examples they could probably pull off the task pretty easily.
Extrapolate this out, and within the next decade or two we'll likely have AI entertainment companies going through decades of comics & books for new content ideas and trying to make cartoons or movies and what not from them.
At least the copyright laws for comics and comic strips is 70 years after the death of the creator... so we got that going for us, which is nice.