This was one of the first pieces I created with Stable Diffusion in August of 2022. I was trying to create something resembling many of the repetitive and overused 3D art tropes on r/vaporwaveaesthetics. The “person wearing a hooded robe looking at big thing away from the viewer” thing was something Caleb Worcester was churning out without fail for months. In a way, my use of AI in this piece was merely to satirize many karma-farming assembly-line pieces people like Caleb were creating at the time. Caleb was clearly only creating these pieces to prop up sales on his print shop and to sell NFTs, which is something diametrically opposed to my public domain licensing and open-source distribution.
Of course, as I used Stable Diffusion more, I realized that it was an invaluable addition to my toolbox. I ended up building a PC to house a 3090 so I could experiment further with machine learning. I’ve since created a lot of cool things with it, including Love Taste. All of my work is open-source and licensed into the public domain. The link to my art portfolio is essentially a single-click download to grab all of my art at once.
Oh, and more recently, I did try and train a quick Dreambooth model based on some examples of 3D art that I thought represented the bulk of the popular stuff on Reddit. It’s a bit weak, but you can sometimes get some interesting results.
You must log in or # to comment.