Dynamic Subspace

Tag: cyberpunk

  • Working on the Site’s Cyberpunk Header Image with Stable Diffusion

    A cyberpunk scene of a floating AI in the center of a computer room and a woman standing to one side of it looking up. Image created with Stable Diffusion.

    I’ve been working on the site’s header image using different Stable Diffusion SDXL models, which give it a higher fidelity without needing as much post-processing outpainting and inpainting as the current header needed, which was made with an SD 1.5 model.

    The idea behind the header image is a cyberpunk scene within cyberspace. The scene takes place within a virtual room representing computers and terminals with an orb-like artificial intelligence in the center levitating above the floor. Within the orb, a shadowy figure can be seen. To the side of the AI is a woman standing before it and peering into the depths of its otherness.

  • Cyberpunk Help Desk Cat Made with Stable Diffusion

    A chubby anthropomorphic cat wearing a hoodie jacket is working at a cyberpunk help desk.

    When I saw this image of a cyberpunk computer technician anthropomorphic cat that I generated with Stable Diffusion, the first thing that came to mind was the Bastard Operator from Hell. Having worked at a help desk, I think it would be an interesting experience to be his co-worker. It certainly wouldn’t be boring!

  • Cyberpunk Innerspace Interface Image Made with Stable Diffusion

    Cyberpunk Innerspace Interface Image Made with Stable Diffusion.

    I call this Stable Diffusion crafted cyberpunk image, Innerspace Interface. With the introduction to multiple controlnet layers, I began creating images like the one above. I applied an image of a server farm aisle as the background, a 1790 portrait of Immanuel Kant attributed to Elisabeth von Stägemann (Anton Graff school) as the subject, and a screenshot of a 3D rotated brain model to match the orientation of Kant’s head in the portrait. In the prompt, I added the technical details, futuristic leather jacket, and feminine facial features.

  • Cyberpunk AI in Hell Made with Stable Diffusion

    Cyberpunk AI in Hell 1 created with Stable Diffusion.

    I call these two images generated during the same sequential run AI in Hell 1 (top) and 2 (bottom). Constructed in A1111 using controlnet paired with images of my fMRI brain scans, I felt like these could be concept artwork for a new Doom game. In both, my brain image conjures these demonic faces with glowing eyes. The former is in a more hell-like setting while the latter below is less a hellscape and more post-apocalyptic.

    Cyberpunk AI in Hell 2 created with Stable Diffusion.

  • Cyberpunk Cave of Wires and Light Image Created with Stable Diffusion

    Cyberpunk Cave of Wires and Light Image Created with Stable Diffusion.

    This was an unexpected result from using controlnet in A1111 with an image of my brain fMRI scan. Instead of a brain-like feature, it created this cave surrounded by cables and lights with a bright blue neon light in the center of the cave to which the human subject walks toward. Some of the best iterations in Stable Diffusion come from these unanticipated results.