Generative AI art. Everybody has heard about it now, right? Programs like Dall-E, Stable Diffusion and Midjourney using some pretty impressive technology (and questionable attitudes towards an artist’s ownership of their own work) to create some, let’s face it, amazing images.
The fact that you can tell a computer what you would like an image of and how you would like it rendered, and it is actually able to do it, is still able to stir a sense of awe and astonishment in me. There are ethical issues, environmental issues relating to the massive amount of computing power required to achieve these outcomes, and there is definitely cause for concern from commercial artists losing work to the machines, but if you put all that to one side for a minute, the technology itself is pretty cool.
So, of course I wanted to play with it.
What I did was use a number of photographs that I had taken of actual real world walls covered in ripped and weathered advertising posters, and asked Midjourney to create completely new iterations of these images from that source material. It took some practice. Learning how to describe what I wanted the program to do, but I think I got there in the end.
These images were created using Midjourney v4. I understand that, as I type this, the current release is v6, so I can only imagine how the quality of these types of creations have improved.