Daydreaming and ideating in a human-machine collab.
I find artificial intelligence text-to-image models to be a great space for ideation and letting your imagination roam free. As a human, I provide free association – feeding the algorithm with prompts, ideas and unusual combinations. In return, the AI returns images – abstract collages based on my prompts.
As an artist, I can then look for brand new patterns within these images, find new connections and close the loop by feeding the machine with further queries.
It’s like looking for shapes in the clouds – but then being able to direct the wind.
The project came to being in an organic way. I was contemplating the pollution of our oceans while conversing with MidJourney AI. Over a number of exchanges I came to think of how a plastic bag floating in the water resembles a jellyfish. From there I begun imagining underwater civilizations and fantastical lunaparks in which the line blurs between living organisms and human pollutants.
The result was a series of artificial intelligence generations, looking for beauty in an environmental disaster.
Keeping with the aquatic, underwater theme, I decided to continue the exploration of this strange new world. Combining various sea shells, octopi and plush toys, I’ve arrived at the next series of images.
Cuteness overdose yet? Not quite, so let’s dial it to 11.
For about a month now, every evening a wild rabbit shows up in my garden and stares at my door until I feed it a carrot or a tomato. Naturally, the next animal pick was then: underwater cloud bunnies.
If you’ve followed my work over the past decades, you probably noticed my fondness for world’s most important pollinators – the bees. So, here are some bees. With dandelions.
It’s really kind of amazing how these images are in equal measure cute and sweet, while also being somewhat distorted and disturbing. More interestingly, the way AI diffusion algorithms look for meaning in seed noises is uncomfortably close to how our own eyes and brains process the world around us.
Observing, building, refining and generating these images has a certain recursive quality to the process – where the line between idea and image, the signifier and the signified, author, audience, human and machine – blurs into a beautiful goop of chaos.
Anyway, here are final five images of this series – filled with non specific cute little nightmares.
Daydreaming and ideating in a human-machine collab.
Creative DirectorQuba MichalskiPublishedAugust 2022FeaturingMidJourney AI