About stillnotart ≠

Hi. I’m AnalogKnight.

I build systems, write code, break things, fix things, automate things, and occasionally fall down deep rabbit holes involving GPUs, AI models, and far more shell scripts than any reasonable person should admit to.

Somewhere along the way, I also started making… this.

So what is this?

stillnotart is a kind of digital gallery — a place for AI-generated wallpapers that feel cinematic, moody, surreal, dramatic, quiet, bold, and occasionally weird.

Every image begins life locally on my workstation, born out of Stable Diffusion, tuned models, prompts, experiments, and happy accidents. From there, each wallpaper is analyzed by a vision-language AI running on the same machine. It looks at the image and tries to describe it — subject, style, mood, colors — and flags whether it’s NSFW. I don’t outsource it. I don’t send it to the cloud. It all happens right here.

Then everything gets uploaded, cataloged, indexed, and presented in a way that feels more like a quiet museum than a loud social feed.

Because the internet already has enough loud.

But… why?

Good question.

Partly: because I could.
Partly: because I was curious what would happen if I took AI-generated visuals and treated them with the same structure and intent as a real curated art collection.

And partly: because I enjoy the tension of the question:

“Is this art?”

Some people say yes — because aesthetics and emotion matter more than tools. Some people say no — because the process itself matters, and AI changes the role of the creator.

Personally? I think the debate is more interesting than the answer.

So the logo is “≠”.

Because this is definitely art. Or definitely not art. Or something in between. And I like living in that grey space.

What you’re actually looking at

Behind the scenes, the tech stack looks something like this:

• Stable Diffusion creates the images
• A local vision AI analyzes them and tags subject / style / mood / colors / NSFW
• Metadata is stored in a structured database
• Images are delivered globally through Cloudflare
• The front-end is intentionally simple and quiet
• Everything is engineered — not just uploaded

This project scratches a few itches at once:

• Curiosity — What can these models do when you really lean in?
• Aesthetics — I like screens that look good.
• Engineering — Pipelines, automation, structure.
• Exploring AI responsibly — local, intentional, transparent.
• Presentation — treating pixels like they deserve a proper gallery.

But AI art seems complicated…

It is. And that’s why I don’t pretend otherwise.

AI art challenges ideas about creativity, ownership, originality, value, and process. Some concerns are valid. Some hype is nonsense. Most truth lives in the nuanced middle.

This project doesn’t try to solve that.

It just says:

And if you just came for the wallpapers? That’s totally fine too.

Who is this for?

People who like:

• beautiful screens
• moody visuals
• sci-fi / fantasy aesthetics
• quiet gallery-style browsing
• tech-driven creativity
• a sense of humor about it all
• my followers on Razer Axon

And probably people who enjoy asking: “Okay but… what is art?”

Credits (sort of)

These images are the product of:

• code
• experimentation
• models
• neural networks
• human curation
• caffeine

They were not scraped from artists. They were not outsourced to the cloud. They were not generated for engagement.

They were generated because I enjoy the process.

The big picture

stillnotart is:

• a playground
• a gallery
• a technical experiment
• a creative experiment
• a philosophical shrug

It will grow. It will change. It may someday become something more. Or it may remain exactly what it is:

A beautifully presented wall of pixels,
made by machines,
chosen by a human,
and viewed by you.

Which, honestly, I think is kind of great.

Thanks for visiting. Enjoy the wallpaper. And feel free to keep wondering whether any of this counts as art.

≠