Personal AI infrastructure · Active experimentation
Local AI as personal infrastructure.
Local AI Lab explores language and image models, speed and quality tradeoffs, GGUF workflows, memory limits, and the tooling choices that make local systems useful for real creative and operational work.
Approach
I configure hardware, test models, compare quality and speed, design workflows, and turn experiments into reusable notes and tools. The focus is practical fit and judgment rather than chasing every release.
Selected decisions
- Prefer practical GGUF workflows where possible.
- Separate reflex-fast models from deeper reasoning or creative models.
- Benchmark with repeated prompts so qualitative differences become easier to compare.