What is Against Entropy?
Against Entropy is a living notebook on AI systems, retrieval, agents, and the economies of intelligence. I write about search infrastructure because I build it at scale. I write about agents because they're changing how we think about software. I write about product strategy because that's the lens I see through every day.
This isn't a content business. It's a long-term intellectual project — a place to publish observations, experiments, and essays before they scatter.
Core themes
Search and retrieval. How information retrieval works at scale — vector search, hybrid ranking, caching, latency budgets, and why recall fails in ways that have nothing to do with your embedding model.
AI systems and agents. Harnesses vs skills, orchestration patterns, tool use, eval reliability, and the operational reality of running intelligent systems.
Product strategy. Tradeoff thinking, adoption patterns, PM craft, and the intersection of scaling laws with what customers actually need.
Economics of intelligence. Model pricing, token efficiency, inference costs, and the caching decisions that compound over time.
Where to start
If you're new here, these are the best entry points:
- How Storage Innovations Split Search Infrastructure in Two. On blob storage, consistency, and why search split into two problems.
- Where the Moat Lives in AI Coding Tools. Models are shared. The moat is in the tooling around them.
- Closed-Loop vs Open-Loop Benchmarking. Two fundamentally different ways to measure performance, and why mixing them breaks everything.
How to follow
- Essays. Long-form pieces every couple of weeks. Add the RSS feed to your reader.
- Notes. Shorter observations, usually once or twice a week. Same feed.
- Newsletter. Subscribe. One email every Friday with the week's writing and a note on what I'm thinking about.
- Now. The /now page for what I'm currently working on.