Thoughts
on building software, reasoning about systems, and the occasional philosophical detour. Written to think out loud. RSS Feed
Featured
-
I tagged every blog post by cognitive mode. Most of mine are knowledge-telling — and that's fine.
Adding one optional frontmatter field to my blog produced a taxonomy I couldn't have written without doing the work. The categories come from Bereiter & Scardamalia (1987), and applying them to my own corpus was uncomfortable in useful ways.
-
Scanners produce reports. Insurance needs witnesses.
The AI-supply-chain cohort is already crowded: Invariant, Snyk Agent Scan, JFrog, Klaimee, a dozen more. They all ship the same artifact, a point-in-time report. That is not what underwriters can price against. The primitive insurance needs is a falsifiable attestation with a TTL that re-verifies on drift. Nobody ships that yet.
-
The Firmbyte Gap: Why the Most Valuable Connections Never Happen
The biggest breakthroughs come from connecting unrelated fields — but the mechanism that produces them is broken. Hidalgo's firmbyte concept explains why.
-
Inversion of Control at the Agentic Boundary
Applying IoC and dependency injection patterns to agentic service boundaries — and why technical plumbing alone won't prevent control inversion.
-
Building Smarter AI Agents With Ideas From Philosophy
Updated:Philosophically informed AI agent design — modeling belief, knowledge, and epistemic norms — makes agent behavior more reliable and inspectable.
Recent Posts
-
xAI shipped the names of every knob. They withheld the values. Which mistake is that?
When xAI open-sourced their For You ranker on 2026-05-15, they redacted every numeric weight, every threshold, every Kafka topic, every config string — but they left every symbol name in source. For a serious competitor, the list of names is the more useful artifact. So either someone messed up, or that was always the plan.
-
MeMo prompting, deconstructed: it's just self-routing
MeMo claims near-SOTA prompting via Munger's mental models. Strip the framing: what's left is one trick — ask the LLM to name its reasoning strategy first.
-
Genetic programming, deconstructed: four of five assumptions are 1990s artifacts
Classical GP bundles five independent design choices behind one biology metaphor. Four of them are compute-era assumptions that don't survive contact with frontier LLMs. What's left is a different algorithm.
-
Verifier-native search: the 2026 shape
LLM generation costs roughly 1000× more than verification. The serious program-search systems already exploit this. The agent-scaffold mainstream doesn't — and nobody yet compounds search traces into a domain-specific mutator.
-
Claude Code's Bash tool returns exit 1 with no output? Check /tmp.
If Claude Code's Bash tool returns 'Exit code 1' with empty stdout and empty stderr for every command, including `true` and `echo ok`, your /tmp is almost certainly full. Here is the fix.
-
Building a Behavioral Health Monitor and Feedback Loop for AI Agents
Applying signal processing concepts to coding agent actions — from FFT and Markov models to a pragmatic health scoring algorithm for Claude Code hooks.
