The topic of Artificial Intelligence (AI) is hard to ignore in 2023. There seems to be new tools and applications introduced almost daily using text generators like ChatGPT or image creators such as DALL-E, Midjourney and Leonardo AI. Specialist tools like Vizcom AI are targeting designers with a sketch to rendering application.
I’ve been experimenting with Midjourney both out of personal curiosity and for potential uses in my professional work. It also occurred to me that with the already large and growing collection of images on this site, it might be interesting to see what results came about from mixing images together.
At this point it’s probably worth commenting on some of the controversy surrounding AI particularly involving image generation. I do believe stealing and profiting off someone’s personal artistic style is wrong. In the design world however, referencing (inspiration) happens universally. And that includes competitor analysis studies, teardown labs, and side-by-side testing. Designing in isolation simply does not happen.
What is intriguing about the current state of AI image creators is the ability to randomly experiment and quickly test ideas as a starting point. It’s a way to break free of the familiar shapes or aesthetic habits we all use.
The images that follow were all culled from Seen.Today and combined using MidJourney. The only additional information I added to each prompt was the aspect ratio. None of the output images were further edited other than some up-scaling. To improve the hit ratio, I selected matchups that included similar perspective views and had the focal point of each image fairly isolated.
There’s some interesting results in this batch of 15. The Saab/VW I.D. Buzz combination is shockingly good. As AI image creation evolves, I hope to revisit this exercise and test new ways to creatively use the content on this site.
Photos and Text: Dave Pinter
Additional Images: Midjourney AI