Honest, real-world field tests of the hardware, operating systems, and LLMs I use in my home lab. No confusing benchmark charts or marketing jargon—just a straightforward look at what actually works for a local AI enthusiast and where the friction points are.
UPDATE April 29: Yesterday, I encountered the issue described below and initially believed it was resolved. However, the problem reoccurred today—the laptop screen remained off after boot. Currently, the only…
There is a growing disconnect in the AI space. If you spend any time on YouTube or tech blogs, you are bombarded with “game-changing” agents and “revolutionary” workflows. However, as…
In this post, I share a candid look at the current state of local AI agents. I attempted to use a local LLM setup—running on my Mac via Ollama—to handle…
I just ran Qwen3.6 — Alibaba’s latest local-LLM offering — on my Mac Studio M4 Max and my first impression is honest: this is a monster of a model. Not…
The AI world is obsessed with leaderboards. But for builders and “vibe coders,” MMLU scores mean nothing compared to how a model actually feels on a mid-range laptop or a…
In the world of AI development, we are obsessed with the “engine.” We argue over VRAM, we obsess over Apple’s Unified Memory, and we track tokens-per-second with religious fervor. But…
I spent 25 years in the Linux ecosystem until kernel regressions pushed me to the limit. I bought a Mac Studio M4 Max to escape the chaos. Was it worth…
After 25 years of navigating the Linux ecosystem, I’ve developed a very specific set of requirements for my daily driver. For the last decade, Gnome was my undisputed choice. Coming…
The “Travel” Experiment For years, I had a peculiar habit. Whenever I traveled to Europe, I would “nuke” my laptop and install a temporary, minimal Linux distribution. The goal was…