Technical
How I'm Building Software in August 2026
I threw out my prescriptive agent harness and rebuilt it around Anthropic's v5 guidance. The migration was miserable. The results are worth it.
Recent Entries
Technical
Is ASCII Art the True Turing Test?
Two frontier models couldn't draw an ASCII map of the US. Why LLMs read ASCII art fine but can't write it, and what that says about the jagged frontier.
Technical
LLMs Were Always Going to Be a Commodity
Chinese AI labs keep matching frontier models for a fraction of the cost. Here's why that was inevitable, and why US AI CEOs stayed calmer than the headlines.
Technical
Running a Wiki Whose Main Reader Is a Machine
I replaced a local RAG setup with an LLM wiki: a repo my marketing agent reads directly. The pattern that keeps a machine-first knowledge base true over time.
Technical
How You Do Something Matters
A horror show about a town with no rules cracked open something I've been chewing on: what belongs in an AI harness, and what should be left to the agent.