Illustrations for FastFud

illustration • midjourney

During my studies at FUD UJEP I got an opportunity to create a series of illustrations for the 10th episode of the FASTFUD podcast. It was about the Interactive media studio in our faculty, which often explores the boundaries of media, information, art and even reality itself. I used Midjourney to help me combine photos from the episode's recording with my drawings and created a place where two otherwise incompatible worlds meet.

One of the illustrations featuring AI generated interpretation of my drawing combined with a photo

One of the final illustrations for the podcast episode.

Looking at the photos taken during the recording of this FASTFUD episode, I immediately had a vision of different creatures having fun and enjoying themselves, but I didn't want to just make a simple illustration, because I thought it would result in a shallow comparison between the artists and some cute monsters. I proceeded to read the annotation about the studio and their artistic approach and found inspiration in their dedication to testing the limits of artistic mediums and the boundaries of pretty much everything in existence. Why couldn't the people in the image also be the monsters?

Image generated by using the blend command in Midjourney

Another image generated by Midjourney based on my visual inputs

Some unused drawings (1, 3, 4) and Midjourney generated outputs (2, 5).

This is when I started experimenting using Midjourney's blend option. I was always vary of big generative AI models because of how opaque their datasets are. As an artist, you have no control over the visual inputs. The blend command gives the user way more control by allowing her to combine several visual inputs together. I used the above mentioned photos and my sketches of creatures, experimenting with different color options, textures and brushes and enjoying the workings of chance in the form of AI.

Photo overlooking a quarry with some mining machinery in the foreground

A sketch inspired by the quarry, crayons

1: Photo from the episode's recording. • 2: Drawing of some creatures inspired by the photo. • 3: Final result after several layers of AI and manual editing.

I carefully picked the best generated images and further edited them in Photoshop, using the content-aware fill tool or sometimes even Dalle2 to add the missing parts to the composition. I tried to stay as true to my initial idea and continue working with the element of chance as much as possible. I saw it almost as a conversation between me and some imaginary representative of the digital world.

1: Final cover for the podcast episode. • 2–4: Countdown for Instargam stories.

In the end, I used a similar workflow to create a background and then combine it with the previously generated image, which resulted in a series of fake-drawn monsters and fake-(and for some reason black and white)-photo people throwing a party on the border between reality and virtuality, algorithm and imagination.