Executive / board reader
Start with AI economics and operating implications, then move into research and speaking topics.
Start here
This site spans strategy, research, code, market signal, teaching, and executive AI operations. Start with the route that matches your role, then follow the thread.
Best first move
The homepage and Writing hub now act as the front door. The newsletter keeps the operating signal coming back to you.
Reader paths
Start with AI economics and operating implications, then move into research and speaking topics.
Follow the operating layer: governance, adoption, token economics, data foundations, and controlled execution.
Go deep on context systems, agent architectures, RAG, memory, and production-grade AI software patterns.
Begin with the foundational Context Compilation paper, then read the trilogy and related systems essays.
Use the weekly industry publications for market movement, model shifts, agent techniques, and application-layer signals.
Current reading
Essay
Useful intelligence now runs on one consumer GPU. Advantage shifts to workload selection, evaluation, and operations because cheap AI is not automatically reliable.
qwen38-local-wow
Qwen3.8-27B can hold its native 262K context window on one RTX 5090. The reason is architectural, and the serving stack matters as much as the weights.
qwen38-local-wow
An independent lab investigation. The same open-weight model, served two ways, produced either an empty afternoon or a daily driver. What changed was the harness, not the weights. A newcomer-friendly entry to a four-part series.
Weekly signal
Cross-stack flywheel
For officers tracking AI market movement.
Model layer
For architects tracking model capability shifts.
Operating layer
For builders operationalizing agentic work.
Procurement surface
For buyers watching AI reshape software.