· 3 min read
Hello World — What is Make and Modify?
Scattered notes, dead terminal sessions, and a head full of things I figured out the hard way. This site is the fix.
6 posts
· 5 min read
From Crashed GPU to 7-Second Images — Building a Local AI Art Pipeline
Flux melted my VRAM. Z-Image Turbo didn't. How a single RTX 4080 went from zero image generation to 100+ images in under a week — including the spectacular failures.
· 14 min read
Yes, I Named It JARVIS — A 48GB AI Server for Under $2K
Three Ethereum mining GPUs, a Facebook Marketplace deal, and a growing distrust of big tech. How frustration with telemetry and curiosity about open-source AI led to a dual 3090 home server — and a completely different way of working.
· 9 min read
Hitting the Ceiling — When the Model Is the Limit, Not Your Settings
Six YouTube tutorials, six written guides, every optimization applied. The honest conclusion about open-source AI video generation on home hardware — and why the future is closer than it looks.
· 8 min read
17 Videos Later — The Optimization Maze of Local AI Video
When better settings don't make better video. Two-stage pipelines, GPU memory wars, and the moment a manual prompt beat every setting tweak combined.
· 7 min read
AI Video Generation at Home — Can Two 3090s Keep Up?
Sora makes it look easy. Running LTX-Video 2 on a home server reveals how wide the gap really is — and where the open-source community is closing it.