AI art seems to be getting more popular as the programs become more advanced and the technology behind it continues to grow. People use it to generate new images and ideas from a prompt, new images from existing images or to create videos, in the form of latent space walks.
I wanted to create an AI-generated video. After a session with Joe Blair, I began experimenting with different data sets of images to create new photos I could blend into a short AI-generated video. I used the program Runway to create these short videos.
I began with a data set of clouds and started with a few images I liked, then added frames to the video which created further images based on the starred images. I chose the framerate and created a 40 second video. I ended up creating two videos with the data set, which consists of multiple images of clouds, but they turned out very differently. I also played around with the data set landscapes and undertook the previous process. I transferred the three videos over to Adobe Photoshop, where I layered the three videos and added a different blending mode to each; for the video I ended up with six layers, two of each.
After editing, I thought the video needed an audio element to make it more engaging. I looked at a few potential sound options and ended up using dream-like audio – like an acid trip to suit the aesthetic of the video.
