Long-form writing on runtime, distributed systems, kernel internals, Go concurrency, memory models, eBPF, and assembly walkthroughs.
When the monthly LLM bill jumps several times over, the first instinct is that the model got more expensive or usage simply grew. It is almost always something else: a distributed systems failure mode, retry storms, fanout amplification, cache misses, unbounded conversation growth, that happens to be denominated in tokens instead of network calls. Debug the call graph, not the model price.
The most common production AI agent failure is treating validation as an assertion, a one-shot pass/fail check, instead of a loop that validates, scores, and decides whether to accept, retry, or escalate. Deterministic code can assert. Non-deterministic model output needs a closed loop, or your agent will report success while doing the wrong thing.
In a rules-first AI system, the deterministic checks get authority. They gate output and overrule the model judge. But a wrong check with authority is worse than no check at all. A first-principles standard for verifying the ruler before you trust it.
Every part of an AI system has exactly one right home: a rule, a state machine, or a bounded judge. Control belongs to the orchestrator, not to an autonomous loop. A first-principles boundary for deciding which technique owns which part.
The reliability of a probabilistic system is set by how much of it you let the model decide. A first-principles standard for shrinking the stochastic surface, on the generation side and the evaluation side.
Generative AI nails plausible shapes but misses constraints, composition, and verification. A first-principles read of the gap, and the architecture that closes it.
Anthropic named the advisor strategy in April. Tobi Lutke made it viral in May with Qwen plus GPT-5.5. Stanford's HazyResearch formalized the same shape earlier. One cost-curve frame unifies all three: a cheap executor runs the loop, an expensive advisor weighs in only at hard decisions. The third recursion.
CodeGraph is the LLM-symbol-graph my prior retrieval post argued should exist. Read against its own SQLite index: why its architectural choices are right, and where the abstraction leaks.
Lossless curated notes vs lossy auto-compression with vector recall: two AI-memory designs that fail differently. One fails like a cache — classical systems already mapped it.
Why Claude Code uses grep instead of RAG: a cost-curve argument, audited against the source — with the Explore vs Fork A/B test nobody mentions.
© 2026 HarrisonSec. All rights reserved.