MARS ART TODAY 1/21/23
RunwayML part 1
Last minute procrastination of runway submission .
So this animation is due in 2 days, luckily I have run across a new AI animation tool on Eden.
Already had a similar idea of how I might do future animations, and this was the perfect time to experiment, I went on to Giphy to find some animated gifts. I looked for old black and white movie type gifts and decided to work out a narrative just with the animated gifts.
I then took so the idea is to create AI animations to stitch the animated gifts together.
So with Eden I first got the animated gifts uploaded it to easy gift website, and I copied the first and last frame out of each animated GIF. I then took the last frame of one gift as the beginning image prompt in Eden, and then used the first frame of the another gift as an image prompt and Eden and let the AI create the imagery in between.
The Eden animation does not interpolate, it actually goes into latent space and generates animation images after detecting the image content, and works from both ends of the animation to the middle to create a semantic light connection between the two uploaded image prompts. Ugh I hope you understood that… If not sorry.
So I was able to start generations and trying to figure out how to imagine this whenever I import all of this and figure it out… Is very interesting so far here are some examples.
So being the weekend between being on call, and karaoke, I didn’t get much other artwork done. I did play around with a few image prompt or image blends in mid-Journey.
I did have a couple of good pop-up AI art today spaces that I do randomly, usually when I’m on the road. Some thoughts from one of those… Is that as artists it’s always easier to have a specific concept or artistic philosophy that we want to create art about, but a lot of times we are generating AI images and come across something very cool but it’s not what we were working on. So is AI artists we have to learn how to reverse engineer a story or a concept to match these brand new images that we’ve received. It’s way easier to control the concept.