74. Year 1–2 — Specialization

Theme: converge (Ch. 61) — choose the spike, go deep, and produce one research-quality build in it.

Choose:

• Algorithms — Part VII/VIII depth, resource estimation craft (Ch. 28.9), the theory-mathematics track; target roles: Ch. 60.2/60.3; best if Ch. 61.2's evidence points you mathematical.
• Error correction — Part X depth, stim mastery, decoder engineering; target: Ch. 60.5 — the 2026-hot path, artifact-meritocratic, hiring now.
• Compilers — Part XII depth, Project 5 pushed to research grade, benchmark craft; target: Ch. 60.4 — the software-engineer-native path.
• Quantum simulation — Parts XIII's 47–48 depth, chemistry stack (Qiskit Nature, PySCF), Trotter/qubitization numerics; target: Ch. 60.9/60.2-chemistry — the application path.
• Cryptography — Ch. 51 depth plus classical crypto (the real prerequisite — a parallel study track); target: Ch. 60.10 — the immediately-monetizable path.

The choosing method is Ch. 61.6's evidence audit, not preference: artifact record, energy audit, reality feedback. The year's structure: two months of informed exploration (try each candidate's project-tier work for two weekends), then commit with Ch. 61.7's annual review date.

Deliverable: first research-quality project. In your chosen spike, one artifact that a practitioner in that subfield would take seriously: the QEC laboratory (Project 4) with threshold curves for compilers-turned-QEC; the compiler (Project 5) with its novel pass for the compiler path; a resource-estimation study for algorithms; a small-molecule benchmark suite for simulation; a PQC migration analysis of a real system for crypto. "Research-quality" = Ch. 54's discipline visible (seeds, error bars, baselines) + public + written up (Ch. 62.2). Year-2 exit: your specialization is legible to strangers from your GitHub alone.