Bryan Ramos

Some notes from maintaining a TurboQuant llama.cpp fork and testing whether hot MoE experts on the GPU could make a huge local model practical.

The machine I built for local AI work, why I built it, and why good enough hardware changes how I use these tools.

Cloud AI is useful, but I want more of my day-to-day AI workflow on infrastructure I control.

I was an early Nostr adopter. Two and a half years later, I find myself removing the link from my website after inactivity.

Most Linux systems are good enough. Some systems have strict timing deadlines.

Yesterday Twitter became X. It doesn't really matter, and that's the whole problem.

Nostr is rough around the edges and the user base is tiny. I'm watching it anyway.

A practical look at PGP, why I finally got around to setting up my own key, and why you probably should too.