Notes
Short pieces about the methodology and architecture decisions behind the AI systems I ship — specs, evals, multi-agent orchestration, LLM integration, and the discipline of directing coding agents.
July 1, 2026
'RAG is dead' is a category error
Every time context windows get bigger, the headline returns: 'RAG is dead, just put everything in the prompt.' It's wrong every time, and it's wrong for an instructive reason — it confuses a technique with a problem. Retrieval didn't die; naive vector-database RAG did, and it's being replaced by smarter retrieval, not by no retrieval. 'X is dead' is almost always the sound of someone mistaking the current implementation for the permanent need underneath it.
- methodology
- architecture
July 1, 2026
The best agent is the most boring one
The industry sells autonomy as the goal: hand the agent a fuzzy objective and let it figure out the how. But the systems that actually survive production do the opposite — constrained tools, deterministic workflows, bounded decisions, human gates. Autonomy isn't a virtue you maximize; it's a budget you spend down, and every dollar you spend buys you a new way to fail. Put the intelligence in the narrow decision, and make everything around it dumb, predictable scaffolding.
- agents
- architecture
July 1, 2026
The internet went dark. Build for a web you can't trust.
'Dead internet theory' used to be a conspiracy meme. Now that a majority of new web pages contain AI-generated content, it's an engineering constraint. Your agents retrieve from a web where you can no longer know who — or what — produced anything. The danger isn't that everything's fake; it's that provenance became unknowable. Which means 'it's on the internet' is dead as a trust signal, and trust has to move to the data layer: signed, allowlisted, provenance-tracked sources.
- security
- architecture
July 1, 2026
Your million-token context window is lying to you
Vendors sell context length like RAM: bigger is strictly better, just stuff everything in. But attention isn't uniform. Studies keep finding the same U-shape — a model reliably uses the start and end of its window and quietly ignores the middle, with accuracy dropping 30%+ once the important thing is buried in there, sometimes after just 10k tokens. Context isn't a bucket you fill. It's a scarce, positional resource you engineer. 'Put it all in the prompt' is the new premature optimization.
- architecture
- methodology
July 1, 2026
Your model is a vendor, not infrastructure
Founders treat the frontier model like electricity — a utility that will always be in the wall. It isn't. It's a vendor with a shrinking lifecycle and a kill switch you don't control, and in 2026 we watched a top model get pulled worldwide days after launch for reasons that had nothing to do with anyone building on it. Model portability isn't over-engineering. It's the minimum posture for a business that runs on an API it doesn't own.
- business
- architecture
June 23, 2026
The scarce resource is the people who design the architecture
Noam Shazeer — co-author of 'Attention Is All You Need,' the paper behind basically every modern AI model — just left Google for OpenAI, less than two years after Google paid $2.7 billion to bring him back. Google lost two AI heavyweights in three days. Strip away the drama and there's a clear signal: the bottleneck in AI isn't compute or data. It's the handful of people who design the architecture.
- architecture
- business