Hello World — What is Make and Modify?

I’ve spent the last two months building things I never thought I’d build. A dual RTX 3090 AI server running local language models. Image generation. Video generation. Voice synthesis. An AI agent framework that lets my home server do more than answer questions. I went from not having a GitHub account to pushing code weekly.

None of that knowledge is written down anywhere useful. It’s in terminal history, scattered text files, and my own memory — which, if you’ve ever tried to remember a Docker command you ran three weeks ago at 2 AM, you know is not a reliable storage medium.

This site fixes that.

What Make and Modify Is

Every project gets documented. Not the polished “here’s how easy it was” version — the real version. What I was trying to do, what I tried first, what failed, what actually worked, and why. The dead ends matter as much as the solutions, because dead ends are where most of us spend our time before finding the thing that clicks.

The focus is on things you can build and run yourself. Local AI on your own GPUs. Self-hosted services that replace cloud subscriptions. Automation that makes machines do the boring work. No accounts, no API keys, no monthly bills.

Why I’m Doing This

I work in software and I’ve watched AI reshape the industry from the strategy-deck side for years. In late 2025, I decided to stop watching and start building. The full story is in the first post, but the short version is: I had some GPUs from an old Ethereum mining experiment, a growing frustration with big tech’s approach to privacy, and a realization that if the models are open-source and the hardware exists, there’s nothing stopping anyone from running this stuff at home.

Two months later, I have an AI server named JARVIS that runs around the clock, a workstation that handles everything else, and a growing list of projects I couldn’t have built alone. I’m not a software engineer — I’m a builder who figured out that the barriers to entry have mostly disappeared.

If you’re curious about local AI, thinking about building a homelab, or just want to see what’s possible without sending everything to the cloud, you’re in the right place.

What’s Here

Builds — Hardware, servers, the physical stuff. What to buy, how to put it together, what to watch out for.

Local AI — Running LLMs, image generators, video models, and music generators on consumer GPUs. Honest assessments of what works and what’s still rough.

Guides — Step-by-step walkthroughs when I solve something that took too long to figure out.

Troubleshooting — The problems that don’t have a single blog post with the answer. The ones where you’re reading a five-year-old forum thread and hoping the fix still applies.

How This Site Works

Astro and Tailwind CSS, deployed to Cloudflare Pages. Markdown in a git repo. No CMS, no database, no paywall. I write, I push, it deploys.

Let’s build something.