Blog
Blog
Long-form essays on AI products, eval / benchmark methodology, and design.
-
Designing Loops Is Your Next Leverage
Loop Engineering turns repeated prompting into an automated feedback loop—but only when the task has a cheap, reliable verifier.
-
How Far Can You Trust an AI Agent?
An AI agent's safe autonomy depends less on raw intelligence than on whether you can check its work quickly, cheaply, and reliably.
-
What an AI-Native Person Is Actually Like
People who adapt fastest to AI are not always the most technical. What distinguishes them is a different set of default assumptions.
-
Why Agent Interfaces Run Backwards
A nine-year-old built a polished game with Claude Code. The model was capable; the interface determined whether a beginner could steer it.
-
What an AI-Era PM Actually Does
In AI products, PM work is shifting from requirement documents to benchmarks and evaluation systems that define how models should behave.