A developer's quiet notebook — practical notes from experiment to production
McKinsey's 'Agents, robots, and us' Europe report shows EU and US automation potential are nearly identical. Contrary to popular belief, the gap is not in technology but in adoption pace and country-level variance. The real shape of skill reshuffling revealed by 5× AI fluency demand and the Skill Change Index.
AI products are fundamentally different from traditional software. This post distills 12 counterintuitive lessons from 20 builders, the CC/CD framework for AI development lifecycles, and a 3-phase eval system — all sourced from Lenny Rachitsky's newsletter.
Self-Tuning Loop 4 steps (Generate → Capture → Analyze → Evolve) extracted as a universal module. Supabase DDL, diff capture utilities, analysis/evolution prompts, email/blog examples, GitHub reference implementation.
The actual production system implementing Self-Tuning Loop from Part 1. Data collection (35 sources), AI curation, Telegram input pipeline, weekly auto-review, and Syncthing-based zero-deploy prompt evolution.
Six core patterns of multi-agent workflow (Supervisor / Sequential / Hierarchical / Network / Swarm / Map-Reduce), grounded in primary sources from LangGraph, CrewAI, OpenAI, and Anthropic. Each pattern's topology and fit, plus a decision framework for production.
A phased view of how a Korean financial marketing pipeline (briefing → research → planning → compliance → launch) could be moved onto multi-agents. Phase 1 RAG agent, Phase 2 collaborative multi-agent (CrewAI as one example) with HITL, Phase 3 macro orchestrator + memory layer + observability. Each phase is paired with a component sketch, the patterns worth watching, and where it tends to break — held loosely as one possible path among many.
Anthropic reports 90.2% gains from multi-agent. Cognition publishes 'Don't Build Multi-Agents'. A side-by-side reading of the two camps' primary sources, plus a decision framework for single vs multi.
Three forces (deregulation × AI build × SI inertia) producing three scenarios for Korea's software market — Korea Vindicated, Double Loss, Bifurcation. The real branching point.
Korean SaaS companies blocked by SI and on-prem walls at home, then taking #1 positions in Japan — a natural experiment in Korea's build lock-in.
How public SW pricing, Hancom/groupware NIH, and 30 years of SI outsourcing produced Korea's build lock-in.
Korea's software spending sits at 0.3% of GDP (rank 51) vs. R&D at 5.0% (rank 1). The series thesis reduces five 'peculiarities' of Korea's market to one root cause.
What surfaces on the second read of Anthropic's Advisor Tool. This isn't new architecture — it's a temporary fix shaped by 2026 pricing. A pattern that disappears once Opus prices drop, and eleven other papers from the same period are quietly moving the same way. The anchor of the series.
How do you organize the sliced agents. Four structures, paired with one skeptical paper that argues 'LLM swarms aren't really swarms.' The conclusion lands where it usually does — structure choice gets dragged along by pricing. Part 2 of the series.
The five ways 2026 papers slice an agent, side by side. Two things stand out by the end: Role, Skill, and Judge are different names for the same concept, and the time-axis literature is nearly empty. Part 1 of the series.
