░░░ ░░░
Where the monochrome world gets its one color — and the machines do the talking.
Local LLM — Hermes
Open-weight LLMs running locally via Ollama on a Windows home server. Full privacy, zero API cost, 24/7 inference. Hermes is the daily driver for writing, Q&A, and ideation.
Claude as Co-architect
AI-assisted development as a daily practice. This entire site — design language, animation system, architecture — was built in conversation with Claude Fable 5. No IDE, no boilerplate.
Self-Hosted AI Stack
Home server running AI workloads around the clock, reachable from anywhere over a Tailscale mesh VPN. Remote inference from phone, tablet, or any device — no cloud middleman.
AI × Learning Design
LLMs as curriculum accelerators — rapid content prototyping, scenario generation, and assessment design at speed. The next frontier for L&D professionals who ship.
“The best time to learn how these machines think was five years ago. The second best time is tonight, on your own server.”
— field note, hermes uptime day 214