Summary

Doug O’Laughlin argues AI code agents (Claude Code) are now strong enough to one‑shot MVPs and materially boost analyst workflows, but still need expert supervision. Actionable insights: treat agents like junior analysts—use them to gather and structure data, but keep human judgment for synthesis; run internal benchmarks (case studies) to compare tool performance; and focus on model/tool routing by task. On semis, he reiterates that Moore’s Law’s slowdown shifts power to leading-edge manufacturers and creates pricing power; the current cycle is driven by AI demand, memory (HBM) constraints, and a potential CPU refresh squeeze. Investment references: NVIDIA as a prime beneficiary of the new compute paradigm; ASML as the “science‑fiction” capex backbone; broader semiconductor capex tied to AI scaling laws; memory suppliers positioned for pricing strength. Career advice: go all‑in on a high‑conviction wave early; build deep technical understanding (textbooks, primary sources); write weekly to sharpen synthesis; and periodic sabbaticals can reset perspective without derailing long‑term trajectory.

Chapter Summaries

• Chapter 1: AI as junior analyst. Agents accelerate research but don’t yet deliver reliable meta‑level judgment.
• Chapter 2: Claude Code’s step‑change. Agentic coding quality improved sharply, enabling one‑shot MVPs and faster iteration.
• Chapter 3: Benchmarking workflows. Use standardized case studies to test model performance in real analyst tasks.
• Chapter 4: Moore’s Law slowdown thesis. Shifts industry power toward advanced chipmakers and creates scarcity‑driven pricing power.
• Chapter 5: AI demand + memory bottleneck. HBM and memory shortages could force tradeoffs across GPUs, phones, and gaming.
• Chapter 6: CPU refresh + utilization squeeze. Aging cloud CPU fleets plus AI‑driven workloads may drive a new CPU cycle.
• Chapter 7: Accelerator skepticism. Many AI‑accelerator startups struggle; success hinges on real memory/throughput advantages.
• Chapter 8: Writing and learning process. Weekly writing, outline‑then‑sleep, and reading at scale build durable insights.
• Chapter 9: Personal reset. A long thru‑hike (CDT) helped sharpen self‑knowledge and long‑term motivation.