How I made it

I suppose this project started after seeing Rebecca Leveille Guay's post on Instagram where she shared some AI art that she made, and explained why the images ARE NOT ART.

I had not spent any time looking into AI, although I had seen some really interesting images online, and I knew my students have used ChatGPT for papers and such. Folks flipped out when the camera was invented, and yet we still photograph paintings. I wanted to learn about AI.

After reading about several image creators, I decided to use Bing's free implementation of DALL-E. I spent a weekend exploring its guard rails, limitations and possibilities. EROS was the first work I made using its output.

I decided to use AI to make backgrounds for a story using characters I've been casually playing with in my comics and sketchbooks.

I was in Portugal when I truly started working on this. We visited Livraria Lello in Porto, and I bought a copy of Frankenstein. I read it on the trip, and found so many parallels to our current discussion around AI, that I decided to pay homage in my title.

At the time of this writing DALL-E outputs 1024 x 1024 pixel images. I wanted to work at 1920 x 1080 in the animation, so I used another AI tool to increase the size of my images to 8000 x 8000px. It worked pretty well, and of the ~3000 images I generated, I ended using about 125. I have enough left over to do several other films, I think.

A slightly different recap of this process can be read on my blog.