Yes, I use AI.

The front facing image of this website, featuring a canine in a green jump suit with red flowers is, indeed, AI generated, and a placeholder. I dig the vibes.

I have a simple setup using ComfyUI, the PonyDiffusionV6XL checkpoint, and many LoRas trained on CivitAI.com using my own art as dataset images.

Because I sometimes make use of the img2img function with images I did not create (for example, the rabbit illustration in the pipeline below was not created by me) I should justify it's inclusion: transferring 'style' and compositional elements to the output, to a degree I can manually control. This feels sufficiently transformative that I do not consider it plagiarism.

Further, here is a dataset image from the active LoRa I trained to reproduce my husky character, which is certainly contributing even more to the final output example above as the sampled rabbit illustration.

In short, I'm not interested in criticism for using AI unless the critic has used the tech themselves, and is familiar with what it can actually do through their own experimentation. I am never advertising myself as an illustrator capable of producing images of the same fidelity presented in any promotional material for my projects.

While I do have many arguments against "being against AI" I'll save them for future installments when they come up. I see generative AI images as another tool for the artist that should remain completely unhampered by regulatory bodies.